This year, my newlywed wife and I have had the great fortune of welcoming our first child into the world, a beautiful baby girl named Ava.

This September, we dove head first into the 24/7 care, love and attention that parents give to their babies as they require help feeding, sleeping, pooping, and feeling safe, and learned just how much energy goes into raising a healthy and well adjusted child.

Along the way, we heard stories about the many babies who are not so lucky, that get left behind at hospitals with no one to care for them. Sometimes when our daughter cries out after waking, looking for the comfort of her parents, and we think about all the babies who have no one to care closely for them.

Here in Vietnam, we were introduced to a country-wide network of volunteers who take in these babies and care for them. Sometimes these volunteers are only available for 2 weeks at a time, and are then passed around to different parts of the country based on their availability.

These babies are often left behind as newborns and don’t have any clothes, milk, diapers or other resources they need as a fast growing children. As our daughter has been quickly growing out of her clothes, we have been donating them along with extra breast milk, to folks in these networks who have been organizing resources for the babies.

Coming up on the end of the year, our company, Mosendo, has decided to help the caregivers get more of the resources they need, such as diapers, milk, and clean clothes and blankets.

We’ve already organized a clothing drive with some of the local families in Vietnam, with over 20 families contacting us to donate their old baby clothes.

Next up is finding some funds to purchase other consumables such as diapers, milk, baby formula and other resources needed by the orphanage.

Crypto gives back

At Mosendo, we’ve spent the better part of the last 2 years focused on building decentralized technologies.

Along the way we’ve come across more than our fair share of disagreements, greed, and fraud, which can make us all lose sight of the end goal of all of this — to make the world more equitable place.

Sometimes we need a quick win to remind us of why we are doing this, and being able to get high leverage on your resources to help those who start life with the ultimate of unfair positions is a great way to get that quick win on the longer road of building infrastructure.

While orphans in Vietnam don’t have many resources, the good news is that a little money goes a long way.

For instance:

• Bag of 76 Huggies diapers — $10.23
• 1kg powdered formula — $12.50
• 100 pack of wet wipes — $1.68

Mosendo will be contributing 250 DAI to buying resources to donate to the orphanage when we deliver the clothes on Dec. 22.

We welcome the rest of the crypto community to donate DAI as well, of which 100% will go towards purchasing goods to deliver to the orphanage. Mosendo will film and document the trip so you can see the results on the ground in Vietnam.

Donation address: xmas.mosendo.eth
0x969dEC1169A0377a1764bD4ee342f9f4CD196BA4

(Have a wallet with DAI but no ETH? Send with Gasless)

Merry Christmas and Happy Holidays,

May you and your family have the good fortune of being together, sheltered, fed and clothed,
Tom and the Mosendo team.

PS. Want to donate but don’t have crypto? Get in touch by emailing tom@mosendo.com

Get abandoned babies some milk and clothes with DAI for Christmas in Vietnam was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

This week, we’re excited to announce that Nour Haridy has joined Mosendo as Blockchain Architect!

Nour has been building on Ethereum for over 3 years, as co-founder of Ethereum R&D firm Lamarkaz and their DAI wallet, Metacash.

Nour’s research into EIP1014 led to the discovery of CREATE2 as an on-boarding mechanism for smart contract wallets.

Now Nour has joined forces with the Mosendo team to build towards the shared vision of usable p2p electronic cash.

Mosendo Co-founder and CSO, Tom Howard said “we’ve really admired the work Nour was doing with Metacash and are excited to have him on board!”

It’s great timing for Nour to join the team as we are in the final stage of preparing for our alpha release and getting early testers involved in submitting feedback to the team.

Our waitlist has been growing and we’re receiving strong interest from the community every week so we are excited to continue development on Monsendo with Nour as part of the team.

Nour Haridy of Metacash Joins Mosendo was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

Letter to Mosendo investors — Libra update

Publicly releasing our letter to investors which we sent out as part of last week’s investor update.

This week, Facebook officially announced the release of their digital currency initiative, Libra.

Facebook and all of the partners involved in the Libra project are some of the largest organizations on earth, and are widely seen as legitimate and trusted organizations. It is highly likely that they will be able to quickly get more users using their digital currency than all of cryptocurrencies put together have been able to achieve so far.

As Mosendo aims to provide a very similar service to what the Libra project aims to do, we’ve been tracking this very carefully and planning our strategy in light of this unprecedented announcement.

While full details are not yet available, we’ve been able to map out a few different scenarios, and the good news is that overall, each scenario bodes well for Mosendo and our mission to bring open money and finance to everyone on Earth.

The Beginning of the Currency Wars

The Libra announcement marked the first time in history that a non-governmental organization creates an international currency of a significant size.

With 2.3 billion Facebook users and countless other users in the Libra partner network, this could be the largest captive audience gaining access to a new currency, ever. No COUNTRY has ever launched a currency with such massive potential on day 1.

Such an unprecedented move will not go unnoticed by not only governments worldwide, but also other mega-corporations who may feel inspired or threatened by the Libra project, and we are already seeing moves by world governments such as France and the US to halt the launch of Libra.

We expect that we will see more similar coalition currencies announced in coming years, from small countries working together to create a regional currency, to mega corporations wanting to cash in on the additional monetization potential of having their own currency. In South East Asia, we are already aware of early talks to exploring what a regional currency might look like.

We expect that major US tech companies such as Apple, Amazon, Google, and Microsoft will seriously consider the threat or benefit of Libra and possibly conclude they should create their own currency. In Asia, we expect the mega-corporations such as Kakao, Samsung, Alibaba, Hanwha, and many others to collaborate with each other and sovereign nations, to form their own coalition currencies.

There is even a possibility that world superpowers start leveraging digital currencies as a form of outright neo-colonialism, and a digital or physical war emerges to make their fiat currency the official currency of the world. As you may know, China is already doing this indirectly through predatory lending practices with African and Asian countries.

As the second and third order effects of the Libra project become clear to those in power, expect to see more Libra-type projects come onto the market and politics start to heat up.

What does this all mean for Mosendo?

We started Mosendo as a way to sustainably solve some of the major injustices we see in the world.

The money system as it is, a system of control for the ruling political class, keeps billions of people in poverty and excludes them from financial opportunities that could be available to them.

The change we want to see in the world is one where all people have access to stable and secure money, and safe and beneficial financial products, services and tools.

This means that people in a country like Venezuela do not suffer from the decisions of a ruling dictator, and are instead enabled to replace the despot.

This means that people all over the world can freely send money to anyone else, across borders, without fear of censorship or loss.

The means that the greedy and powerful are less able to leverage their control of monetary systems to extract wealth from the less powerful.

For us to fix these problems, we realized needed to use a decentralized and neutral platform so there is no possibility for bias or corruption.

We also realized that for this new platform to gain usage, we needed a way to incentivize people to use and encourage usage on this platform. The best way to do this quickly and consistently is by leveraging natural human greed, and design ways for people to either earn or save money using our system.

Thus, Mosendo was born, as a for-profit business that makes using decentralized money incredibly simple, while leveraging token incentive systems for user growth.

We always knew that at some point, major corporations and governments would realize the massive advantages of digital currencies.

We’ve also known that depending on the architecture of the system, digital currencies could be used as a force for good, to reduce and distribute absolute power and make things a bit more equitable, or as a force for evil, a tool for consolidating power and increasing inequality in the world.

We’ve always felt that we are in a race to make the more equitable version of digital currency widely available before more tyrannical versions monopolize the market, and the announcement of Libra certainly turns up the pressure on this timeline.

Now is the time to prepare ourselves for the coming battles, and choose our alliances carefully.

It remains to be seen if the Libra project will turn out to be a force for good or an attempt at creating a future ruled by corporate sovereignty over democratic sovereignty.

Either way, the Libra project actually accelerates our mission in multiple ways, either as an ally contributing resources to the cause, or as a Goliath whose resources can be redirected towards a benevolent cause.

We can use our agility to our benefit and leverage guerrilla strategies and asymmetric tactics enabled by decentralized technologies.

The war is about to begin, and now is the time to defend the future we want.

Libra, friend or foe?

Libra’s marketing materials make many claims that resonate with us, they say they are a cryptocurrency, they say they will finally reach the unbanked and underbanked, they say they are operating an open network that anyone can participate in.

However, when digging into the details, we’ve become skeptical that these claims are true.

Unlike a cryptocurrency, the Libra token is backed by fiat currencies in a bank account instead of provably scarce assets on a cryptographically secure blockchain.

From a consensus level, their network is closed, and only trusted entities are allowed to validate transactions and earn money from doing so.

On the user level, they have stated that all wallets and exchanges must comply with money transmission laws, which immediately excludes the majority of the population who are unable to comply with those laws such as citizens in the US sanctioned Venezuela or the 1.1 billion people without government ID.

So seemingly, they will not be able to provide access to many markets, particularly not the underbanked.

It remains to be seen if they will evolve over time towards a fully open and decentralized network, or if they are simply using these terms to take advantage of regulatory loopholes to run an extractive centralized network.

For now, we will continue keeping ourselves informed on their developments and stay involved with the growing ecosystem emerging around Libra.

We assume that they will be very centralized and conservative in the first few years, which means we will be able to provide access to regions they can not.

Even as a centralized service, they actually provide a few key resources for us to leverage:

1. They are directing massive amounts of attention to the cryptocurrency and digital money space which will increase awareness and help legitimize true cryptocurrencies.
2. They will build much more seamless fiat on and off ramps in developed markets such as USA and EU. Creating trust-less p2p OTC swaps on Libra is much easier to accomplish since their ledger is open for Ethereum smart contracts to read. This enables instant fiat/crypto settlement.
3. They are opening up a massive digital identity system that can be leveraged to seed a decentralized identity protocol and enable decentralized reputation, credit ratings, and cheap unsecured lending.
4. They will attract the attention of all regulators worldwide, paving the way for other cryptocurrencies to hide in their wake.
5. While we currently think it is unlikely, should they prove to be an ally for a decentralized and equitable future down the road, we can more deeply integrate their currency into the Mosendo ecosystem.

Moving forward

We continue to build on the Ethereum ecosystem and using DAI in partnership with the Maker Foundation.

Ethereum and DAI continue to be the most usable, decentralized, censorship resistant cryptocurrency available today and the best way to deliver on the mission of Mosendo.

We are discussing with other Ethereum teams different ways of building Libra to DAI bridges or swaps, and since the testnet of Libra is available today, the technology to bring Libra liquidity to the DAI ecosystem should be ready whenever Libra goes live in 2020.

The next 6 months are critical for the Ethereum and decentralized technologies ecosystem as we simply must get the user experience problem solved in order to stay a step ahead of Libra.

We are looking forward to releasing Mosendo this summer with a seamless experience which allows people to send money as easily as they can send emails.

Thank you for your ongoing support in the coming battle for the people’s money.

Sincerely,
The Mosendo Team

PS. If you’d like to share your thoughts on this letter, we’d love to hear them. Please email us at rossco@mosendo.com & tom@mosendo.com this might just be one of the most important discussions of our time.

PPS. Want to get involved with Mosendo?

Letter to Depfi investors — Libra update was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

Drinks paid in DAI — what happens when the bar only accepts crypto?

DAI Hard Drinks Recap — From NYC Blockchain Week 2019

For the team building Mosendo, our primary focus is making cryptocurrencies usable as a daily currency, so it’s always struck us as a bit ironic that most cryptocurrency events don’t even accept cryptocurrency as payment!

So when we were planning an event with MakerDAO to kick off NYC Blockchain Week 2019, we knew that we absolutely HAD to make the event focused on real usage of the cryptocurrency DAI, not just talk and speculation.

Thus, DAI Hard Drinks was conceived of, where attendees could ONLY pay for their drinks using DAI, no other payment forms accepted at the bar!

Maker + Mosendo Presents - DAI Hard Drinks

The days leading up to the event was a whirlwind of preparation as our event sold out within 28 hours, at 225 person capacity. Our waitlist grew to 390 people — clearly there is an interest for stable decentralized cryptocurrency!

We had been working on a proof of concept in collaboration with Connext network that demonstrates some of the key features we are building in to the Mosendo wallet (read our Hackernoon article for the full product vision).

The Mosendo demo featured

• instant, browser-based wallet. No fumbling with private keys
• no ETH or GAS required to make a transaction, in fact no fees at all
• instant transactions, no need to wait for block confirmations

The experience worked a bit like this:

At the rooftop bar, we setup all the bartenders with the Mosendo demo wallet and walked them through accepting DAI as payment. Faster than a credit card and just like holding cash!

Then things kicked off! People started showing up, and we handed out QR codes loaded up with 40 DAI for buying drinks!

The QR codes were actually links that contained a one time redeemable hash. When the user opened the link on their phone via the QR, the Mosendo web wallet was created in their browser with the $40 redeemed.

This “send link” functionality is possible due to our layer 2 state channel technology partner, Connext Network.

Sempo on Twitter

What a view! @MakerDAO and @GetMosendo drinks for #BlockchainWeekNYC. You can donate to @UNICEF @LupusOrg and @OxfamAustralia!

In addition to paying for drinks, we had QRs posted up that people could scan and instantly donate to charities!

As people started spending the DAI at the bar with instant transactions, the party kicked off to DJ Tweed’s beats.

Despite the rainy day, people were having a great time, chatting about the state of crypto and Ethereum as well as learning about DAI and the latest tech enabling easy to use crypto experiences!

Multiplied on Twitter

Out here at @MakerDAO + @GetMosendo meetup where you can redeem #DAI for drinks or charity 🥳 #blockchain #crypto #Consensus2019 #NYBlockchainWeek

User Feedback

Overall, the feedback from the crowd was incredibly positive. Several people had ‘eyes wide’ WOW moments when they realized that crypto payments can be fast and easy while still remaining decentralized and self custodial!

We were able to capture some of the excitement from the event on audio:

Crypto needs to be used as a payment method and if DAI enables that through Mosendo, its pretty critical to the industry.

If there’s any app that make sending payments easier without using a private key thats a huge plus for the industry

DAI is a great onboarding experience to crypto and will further the use of cryptocurrencies if we can get it right

User stats

The next day, we pulled some statistics from the layer 2 network:

• 289 payments total totalling over $5,300
• $18.34 average payment
• $14 median payment
• $11.50 most common payment
• 130 payments /hr in busiest hour
• 10 payments /min in busiest minute
• Someone kept the party going, there were 3 payments after 2am (bravo)

Learnings and next steps

We learned a lot from this experiment!

It was great to have our thesis validated that non-custodial cryptocurrency is 100% viable as a spending currency, if the user experience is designed correctly!

An entire bar full of 225 people was able to order and pay for drinks in a fast-paced chaotic space, just like they would with credit card or cash. The previous barriers to crypto currency usage in bar-like environments have been eliminated!

At one point we had an issue with payments going through due the network being undercollateralized due to underestimation of usage. In the future we’ll need to be sure to have sufficient collateral on hand for optimistic cases.

Crypto users in particular are more advanced, so the simple interface hid some of the details they are accustomed to. We had people trying to do layer 1 activities that did not work with our layer 2 wallet, so we need to improve the cross-layer compatibility 🤣

Also browser based wallets are a bit risky when it comes to private key management; more robust fund recovery mechanisms need to be in place to ensure users don’t lose their funds.

Next up for us is launching the full native experience!

Sign up to get notified of the full Mosendo release.

Drinks paid in DAI — what happens when the bar only accepts crypto? was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

My wedding in Vietnam was 3 days after Christmas, and the wedding savings I transferred from the US to Vietnam to pay for the expenses was gone.

Thanks to the USA banking system, sometimes things randomly don’t work, and you have to wait 3–5 business days for the bank to resolve it. This means you can’t access that money at all, it's in limbo.

If something happens to go wrong on a Friday night, well the bank closes for the weekend, so add 2 days to that 3–5 days. And if the Christmas holidays happen to fall right after that weekend, you can add another 2 days.

So there I was, one day before our wedding, without the money I needed to pay for it.

Of course I have backup plans, so for me, it was a stressful annoyance, and inconvenience, but it reminded me of just how frustrating it can be to not have full ownership and control over my own money.

Imagine that a young Vietnamese woman has an internship in the US when her mother urgently needs money to pay for a life saving operation. She sends everything she has, and it gets randomly held up by the USA banks, causing it to arrive too late to save her life.

Cryptocurrencies have been around for a long time trying to solve this problem, yet we still don’t have a good solution. It is a shame.

This problem is one that has frustrated me for decades, it is what drew me to fintech and crypto, and the lack of solutions I found while researching for this article is what has driven me to step back from Taureon angel group and back to the startup building side. More on what we are building later in the article.

Where is the decentralized money we were promised?

Bitcoin was launched over 10 years ago, with a simple promise:

However, all variants of Bitcoin and other top cryptocurrencies that we have today, do not fulfil this promise.

Before the various cryptocurrency maximalists start evangelizing that their pet cryptocurrency already does this, let me take a moment to define exactly what peer-to-peer electronic cash system means to me.

“Peer-to-Peer”
To me, this means decentralized, self-sovereign, censorship resistant. This means I own and control my part of it, no one can revoke it or otherwise stop me from having or using it.

“Electronic Cash”
To me, this means I can use it as money in my day to day life. This means it is fast and cheap to transact, and this means that the value of it is relatively stable. My purchasing power does not change much from holding it.

“System”
A system is holistic. It is not only comprised of the backend decentralized ledger, but the entire user experience, which means the app interface, the ecosystem, and everything in between.

Most Cryptocurrencies are “Peer-to-Peer”

Most of the top cryptocurrencies check the first box, “peer to peer”.

Bitcoin and it’s forks, Litecoin, Dash, and various others are decentralized, self-sovereign, censorship resistant, and peer to peer, in varying degrees.

Bitcoin with its superior hashing power is probably the most safe and censorship resistant, while Dash has a strong boots on the ground merchant adoption program and Ethereum’s smart contracts are enabling an ecosystem of open finance products.

One can debate the nuances of the various project’s degree of decentralization for quite some time, but given that they are considerably more decentralized than existing fiat systems, the networks are good enough to use for now.

Almost None are “Electronic Cash”

All of the top cryptocurrencies are highly volatile and therefore useless as a unit of account or store of spending money.

When my friend sends me BTC to buy him a car on Friday, and I can only buy 75% of that car by Monday, that BTC is unusable as a form of cash or money.

Sure it may turn out to be a good store of value in the long term, say 3 years, but that doesn't make it cash any more than oil or soybeans are cash.

Yes, electronic cash must also be useful as daily spending money. Buying a meal needs to be quick and frictionless at all times, which means 10 minute transaction times and $5 transaction fees will also make a currency unusable as cash.

But if I had to rank the barriers to electronic cash, value volatility is the number one barrier to adoption, fees and speed come after.

The “Systems” are Incomplete

Everything currently on the market is an incomplete experience.

We have a lot of backend ledgers of varying completion, but I have yet to see that combined with an easy to use front-end and robust ecosystem.

While we have plenty of wallets that allows users to manage multiple cryptocurrencies, we don’t have a wallet that simply allows you to send cash to your peers, with an easy to use app interface.

So how is it that in 2019, over 10 years since Bitcoin was launched, we still don’t have a complete peer-to-peer electronic cash system?

Let’s Finish Building Bitcoin’s First Use Case

It is clear to me that for decentralized technologies to thrive, we must all come together to solve the first use case of blockchain and cryptocurrency: electronic cash.

Through my research, I’ve discovered that many of pieces needed to make this happen, did not fully exist during the first 10 years of Bitcoin’s life, which is why it’s taken a bit longer than expected.

However, due to the funding wave in 2017, both technical and market developments in the past year finally give us all the pieces we need to fulfil the vision of Bitcoin.

The mission of our new company, Depfi, is to make a peer-to-peer electronic cash system accessible to the world.

How to create a Peer-to-Peer Electronic Cash System

So far I’ve defined some of the criteria of a peer-to-peer electronic cash system:

As an ecosystem, we’ve clearly got 1 out of 5 requirements fulfilled; we have many decentralized ledgers with varying degrees of censorship resistance, of which Bitcoin and Ethereum are leading the pack.

The remaining criteria don’t have any clear leaders yet, but as new technology is developed, the pieces need to seem to be available or nearly available today, they just have not been assembled in the right order.

So how can we go about assembling the other parts of this system?

Before digging into the possible technology we could use, let’s first consider the User Experience so that the system we design is able to get real usage and adoption.

What actually gets adoption? What do users actually want? What problems are people experiencing?

What do users want — The User Experience

I can only draw from my own experiences and observations here, but I suspect I have a somewhat diverse set of experiences and observations to draw from, having lived and traveled among the rich and the poor, been dead broke and well off, in developed countries like US and Germany and in developing countries, particularly those in South East Asia.

In all my travels, I found there are some universal truths about people:

• People want to have a contented and fulfilling life with their family and friends
• This means they need things that keep them alive, safe, and happy, which they buy with money
• This means they need to earn, save and grow their money

And all of this boils down to a simple desire for their money:

People just want to safely send and receive their money between people they know and merchants they engage.

Breaking that down reveals additional desires:

“Just send and receive”

The biggest issue I see with most crypto asset wallets is that they meant for managing a multi-asset portfolio, making them complicated and confusing. For users that want to “send money”, they don’t really care about the various crypto assets, they just want their value to be easily usable to send money to other people or merchants.

Users want to send a receive value, denominated in and pegged to whatever currency they think in, such as USD or Thai Baht, not BTC or ETH.

This also means the money they are sending around does not violently fluctuate in value as that makes it unusable.

Users don’t care how the backend works, as long as they can be reasonably sure that their money will never be lost or taken, and never lose value, ever.

This means the front end needs to simply send the users currency of choice around without exposing them to the complexities of making that happen.

“Safe money”

This can mean different things to different people, but in general this means their money

1. does not lose its value due to inflation or otherwise
2. can’t be randomly “lost” or delayed by banks
3. can’t be arbitrarily seized by governments
4. can’t be stolen by common thieves

In developed countries such as the US, most people believe that USD already fulfills these criteria. Europe mostly believes this as well, though confidence in EUR and other European currencies has faltered after losing 30% against USD a few years ago.

In developing countries, people generally believe that their local currency is at risk on all 3 factors above, yet they often have no choice but to use it. As a result you often see people putting savings into USD, gold, and real estate, and keeping the local currency for spending.

People tend to avoid keeping too much in the local banks as the banks are prone to being “hacked”, going insolvent due to embezzlement, or being seized by the government for the crime of doing too well without including the bosses.

Because of this, I believe that developing countries are ready to adopt an easy to use cryptocurrency system, even if the inflation rate of their local currency is reasonable.

“People they know”

Users do not care for 32 digit hexadecimal addresses.

Users conduct transactions between people, not between computers, so any system relying on computer addresses will face a lot of friction for adoption.

Custom handles such as Twitter’s @user system is far, far better. The best experience for a user is to be able to simply select the person they know and send to them.

Venmo’s integration with Facebook and the social graph as well as WeChat Pay’s leverage of the WeChat ecosystem are what made those products successful.

A decentralized system would have to exist outside the boundaries of a social network, while maintaining user privacy, but these networks can be used to bootstrap adoption.

The easiest step in that direction is through a decentralized @ or $ handle system.

“Merchants they engage”

For electronic cash to work, it must be easy to quickly make payments to merchants, online, but particularly in person.

In Asia, paying a merchant is as fast as scanning a QR code printed out and on display at the cash register. In the US, credit cards and NFC payments are quick and prominent.

A complete electronic cash system must have easy support for QRs and NFC with fast transaction time so the merchant can move on to the next customer quickly.

We are assembling a complete Peer-To-Peer Electronic Cash System

Introducing Depfi

After searching for years for this solution, hoping someone would build it, I’ve decided to do something about it! I have joined forces with Rossco Paddison, an entrepreneur and educator who has trained 7500+ people on how to use a Bitcoin wallet and experienced their pains and confusion first hand.

This system is something we think the world desperately needs, and we can no longer sit back and wait for someone to make it happen, so we have formed Depfi.com to complete this mission.

Our organization is focused on user experience and mainstream adoption, and so we are building what we see most lacking in the ecosystem, Mosendo, a decentralized payments app with UX for a mainstream user base.

Based on the wants and needs of users, and the criteria for p2p electronic cash, we are assembling a complete system requiring

• Stable Cryptocurrency — A decentralized, censorship resistant, peer to peer, stable store of value that is low fee and has fast transaction times.
• Simple App — An easy to use client wallet app denominated in only one or two fiat currencies which is accessible globally without ID. (Mosendo)
• Fiat Access — A global network of fiat gateways to facilitate access to local banks

Stable Cryptocurrency

The stable store of value must be decentralized and sovereign grade censorship resistant, otherwise it is not much different from the existing bank infrastructure.

This means all the fiat backed bankcoins on the market such as USDT, USDC, TUSD, GUSD, PAX, and others, are not suitable choices for this system.

Bankcoins are still very susceptible to asset seizure by governments, banks, and even the less regulated organizations issuing the tokens. We can already see this in effect with companies outright stating they can revoke tokens based on violation of terms of service, and with Gemini refusing to honor tens of millions in withdrawal requests by trading companies in good standing.

Rule of thumb — anything backed directly by fiat in a bank is not censorship resistant or decentralized.

Fiat pegged cryptocurrencies are therefore the most promising at the moment. These coins are pegged to the value of a fiat currency but not backed by them.

Ideally the value stabilization did not rely on pegging to fiat value, but no other method of maintaining or evaluating stable value has come on to the market so far. I suspect this will be a chicken and egg problem, where cryptocurrency adoption allows for new methods of determining stable value, and so pegging to fiat is a bootstrapping method.

The most interesting decentralized pegged stable crypto on the market so far are MakerDAO’s $DAI, Synthetix, and Duo.

Others to watch are Reserve and Constant, which are currently fiat backed but have a roadmap to becoming decentralized and fiat pegged.

DAI in particular has hit an important milestone for cryptocurrency adoption — liquidity.

With over $90m in issued stable token, $30m avg daily trading volume, healthy collateralization, and constant steady growth, DAI has finally hit a point that no decentralized stable cryptocurrency has hit before, it possible to run a small but complete economy using DAI.

Scalability and Fees

MakerDAO and DAI are both currently built on the Ethereum network, so we are only considering solutions built on Ethereum. Should Bitcoin network launch a stable cryptocurrency with sufficient liquidity, we would likely support it.

Ethereum should soon be releasing Serenity, a faster, lower fee version of the protocol.

Until then, we have several slightly less decentralized solutions to use such as xDai on Ethereum POA and the more decentralized payment channel network, Connext.

Initially, these give us a way to bootstrap the ecosystem with nearly instant transactions and zero fees, and build towards a multiple layer solution depending on transaction size and use case.

This means in the near future, some transactions will take place via side chains or payment channels, and some transactions will be settled directly on the Ethereum network.

The user will not have to experience ETH or GAS, which can be confusing and somewhat scary concepts for mainstream users.

Simple App — Mosendo

“Money” and “phone” are increasingly becoming synonymous, so for crypto adoption, an easy to use app is required.

The money in the app must be denominated and stored in the currency the user thinks in (likely the local fiat), with the ability to hold a secondary currency (likely USD). We achieve this by using DAI for USD and synthetic fiat currencies from synthetix.

It must be easy to transact with people, not public keys. We accomplish this through the use of decentralized $motags, saved contacts, QR codes, NFC, and privacy preserving social network import (anyone have ideas on doing this without our organization storing user data?).

No ID is needed to use the wallet or hold a balance as it is simply a non-custodial cryptocurrency client, opening doors to global adoption instead of being stuck in a particular jurisdiction.

Private keys are confusing to many users, and cryptocurrencies are currently lost due to lack of understanding. In Mosendo, the user gets the option of hardware wallet, seed phrase, recovery via their 5 closest friends and relatives, and a few other ideas we’re fleshing out.

Fiat Access

Ideally a p2p electronic cash system is a comprehensive experience that does not require ever using a bank account or cash. However, the transition will take time, and users will invariably find themselves needing paper cash or a bank account for various purposes.

Regulations on sending money, including and KYC/AML, create many of the problems we are setting out to solve. The app itself does not require KYC/AML, but instead will benefit from the existing compliance of money transmitter partners to exchange fiat for crypto.

The system can not rely on a single fiat gateway, but must instead allow a network of multiple gateways in every country to allow for people all over the world to gain seamless access to the cryptocurrency system.

Mosendo uses services such as Wyre for this, as well as other payment processors and cash top up merchants we are in talks with (anyone we should be talking to?).

The Road To Adoption

Peer-to-peer electronic cash system will find the strongest product/market fit in emerging economies such as Asia, Africa, and South America where there are acute pains around currency and money.

In these regions it is easy to find examples of hyperinflation, bank theft, government theft, corruption, and other issues that make local fiat currencies and banks less appealing.

These countries face the largest issues when it comes to cross border payments, such as remittances, cross border trade and travel.

Often currency controls prevent capital outflows from these countries, and predatory practices result in high fees for migrant workers sending their hard earned money back home.

Many of these regions already see significant adoption of volatile cryptocurrencies such as Venezuela and Zimbabwe in the face of hyperinflation.

Many African countries already have high adoption of mobile money systems such as mPesa’s centralized SMS based payments.

Many countries already use telecom phone service top up credits as a form of currency, particularly India, Thailand and Indonesia.

Developing SE Asian countries have high smartphone penetration and saw high usage rates for speculating on cryptocurrencies, laying a foundation for cryptocurrency wallet literacy.

Additionally, the relatively small geographic size of of SE Asian countries causes frequent cross border currency issues.

SE Asia in particular will be a region of high growth for a properly implemented cryptocurrency system.

Since the Depfi team is already based in SE Asia, we will be here on the ground, learning from the everyday person how we can make the best experience for them.

From Burmese migrant workers in Thailand to Filipino virtual workers earning from overseas, we empower them to take back ownership of their earnings and avoid predatory money remitters and sketchy banks.

In the future, we dream of a day when not only I as an American can take out my phone to instantly pay an Indonesian street food vendor, but a Venezuelan can easily pay for currywurst in Berlin, and a Nigerian in Canada can easily send money to family back home, all quickly, cheaply, and safely.

Learn more about Depfi and Mosendo

Thanks for taking the time to read this. If you are as passionate about borderless access to money and wealth as we are, join our mission below.

See the latest articles from Depfi’s team:

• Why I’m doing a startup again after years of investing
• 5 New User Challenges, From The Front Lines Of Crypto Adoption!
• The Surprising Complexity of Money Management on a $2-a-day Income

It’s 2019. Where the f*ck is Our Global Peer-to-Peer Electronic Cash System? was originally published in HackerNoon.com on Medium, where people are continuing the conversation by highlighting and responding to this story.

2017 and part of 2018 were good times for myself and many others in the crypto space. The speculative bubble around Bitcoin and ICOs led to a surge of capital, interest, and talented people, freshly converted to the philosophies of self-sovereign money and decentralized technologies.

While the speculative nature of most of the interest caused the market cap to come crashing back down in 2018, hurting many in the process, the massive amount of capital flowing into the space had some positive effects as well.

Many very talented people who had either been working on decentralized tech or were freshly converted, got their crazy ideas well funded, and got to work building remarkable technologies.

Angel Investing

At the time, as founding partner of Taureon angel group, I saw so many smart people building awesome things that I concluded my time was best spent on backing projects with capital, resources, and any useful bits of information I picked up along the way.

We had the privilege of backing some amazing projects that are just now getting their products out there:

• improving privacy technology with REN, Loki, and MobileCoin;
• enabling DEXs with Kyber and 0x;
• fixing IoT infrastructure with IoTeX;
• getting a working version of sharding with Zilliqa;
• tokenizing commodities and trade finance with ABC Platform;
• enabling decentralized file storage with Archon;
• and many others working quietly on building their tech.

Among the 100s of pitch decks we went through, there was one thing that always nagged at me that was missing in the industry: why isn’t there any project or system that lets me use cryptocurrency as a ubiquitous electronic cash? Why hadn’t anything gained traction for the first use case of Bitcoin?

As the downturn of Q3 and Q4 saw most of our angels pulling out of the space completely, and projects in the space going bankrupt left and right, I contemplated what the next moves in the space would be.

Myself and some others from Taureon researched setting up a VC fund and raising external capital to take our activities to the next level, but we kept running into an overarching problem: many of the great technologies we wanted to fund have glaring user adoption problems. How could we ever expect to get ROI without getting users for the technology?

Adoption Research

Personally, I kept coming back to the lack of usefulness of existing cryptocurrencies as spending money. As someone who believes in the philosophy, even I had a pretty hard time justifying using Bitcoin or another cryptocurrency as cash, it was just too difficult, and I didn’t like mixing my payments with my investment assets.

So I set out to figure out how to make cryptocurrencies useful as electronic cash, surely there would be someone working on this problem already that I could support in some way.

As part of my research and observations, I concluded that price volatility was a major blocker to usefulness as cash. If I decide to buy a car on Friday and set aside some money, then by Monday I either have half a car or two cars, the unpredictability of its value over such a short period of time makes it unusable as cash. It seemed unlikely that no one had yet realized this, so I went on a hunt for stable cryptocurrency solutions.

This hunt led me to re-discovering and fully understanding MakerDAO, which at the time of writing has their stable cryptocurrency DAI at a market cap of $92m USD. The exciting thing about MakerDAO was not necessarily that they’ve devised a brand new solution, though they’ve certainly make some massive improvements on existing ideas, but that they’ve finally achieved a market cap and availability of their stable cryptocurrency that has never been done before.

The MakerDAO Rabbit Hole

If you’ve not looked into MakerDAO before, the basics are this: ETH holders who want to keep their ETH, but also want to get some capital to use with it, can collateralize their ETH in a smart contract and take out a loan against it. This loan is distributed by creating a token called DAI, which is worth roughly $1 USD.

What’s super interesting about this, is not that one can get a loan on their ETH, but that the entire system is decentralized, even the stable DAI! Unlike the stable bankcoins on the market right now, DAI has the same censorship resistant properties of any other cryptocurrency, no one can revoke access to DAI like they can with bankcoins.

The available liquidity of DAI, coupled with the stability and decentralized nature of it, unlocks a whole new realm of possibilities for programmable money. Now users can participate in the smart contract economy without having to speculate on the value of ETH at the same time. Speculation and utility has been separated!

This enables a new era for open finance, where users can lend or borrow in DAI, invest in synthetic assets in DAI, participate in prediction markets or reputation in DAI, all without having to speculate on the underlying currency itself. Because it is all decentralized, these products are not restricted by jurisdiction or licensing requirements, which means the 4 billion underbanked who are cut out of the high quality financial system are finally able to get access to sound financial products.

But perhaps the most impactful effect of the availability of stable cryptocurrency, is that finally, the last remaining barrier to peer-to-peer electronic currency has been removed.

Building Peer-to-Peer Electronic Cash

With the discovery of the traction DAI was getting and the realization that the primary blocker of peer-to-peer electronic cash is being removed, I deepened my research to solve the other major blockers I identified:

• Transaction speed
• Unpredictable high fees
• Major UX problems
• Adoption roll out strategy

Since DAI is built on Ethereum, I checked on the status of the various scaling solutions that had been in the works. Ethereum core’s solution, now named Serenity, is still on the horizon. Ethereum POA has a sidechain for DAI called xDAI. Neat but perhaps a bit too centralized. Connext is similar to Bitcoin’s Lightning network, but for Ethereum, which recently launched. Meta Transactions allow for the transactions to be aggregated to combine fees and lets users pay in DAI instead of ETH. Ok, several solutions to transaction speed and fees, looking good.

Next, I wanted to find solutions to the UX issues I found. Since DAI is an ERC20 token, its accessible by any ERC20 wallet. The problem with that is one’s money, DAI, is mixed in with a bunch of speculative and useless tokens, which is an incredibly confusing experience when someone is just trying to use DAI for payments.

Other major UX issues include: the need to use ETH as GAS to pay fees to send DAI around, the use of public keys to tell people where to send money, management of private keys, and lack of ways to turn cash into electronic cash.

I downloaded every single app I could find, but nothing indicated a focus on decentralized payments, DAI, or solving these major UX issues.

Uh-oh, how can DAI be used by the mainstream with all these issues?

Fixing Usability

Along the way I had been sharing my research with my good friend Rossco, who I had been advising on the decentralized asset manager he’d been building out of frustration with trying to manage his own assets.

He kept echoing a lot of the same UX problems, which he had run into while he had spent a few months training 7500+ non-technical folks how to use a Bitcoin wallet. The current state of crypto apps was viscerally fearful for them, they were worried they’d do something wrong and lose all their money.

Together we realized that if we ever wanted to realize our dream of having a usable, peer-to-peer electronic cash, we’d have to step up and build it ourselves, otherwise it might be decades before it actually happens.

So Rossco and I joined forces to create Mosendo, and accordingly my responsibilities with Taureon have become more passive.

Our main focuses at Mosendo are on user experience and user acquisition; we rely on the ecosystem to build the underlying tech infrastructure while we build the front end. The first of which is a p2p payments app called Mosendo.

With Mosendo, we’ve already found solutions for many of the UX problems we’ve identified: solving speed and fee issues with layer 2 solutions, solving public/private key confusion with $motag handles and social account recovery, usage of local currencies via synthetic forex, and abstracting away ETH so that users never need to know it exists.

Adoption Strategy

Our main challenges lie ahead on the long road to adoption: turning cash into electronic cash and convincing one micro-economy at a time to switch over to using Mosendo for payments.

For the first challenge, we are building a network of fiat gateway partners so that users in any country will be able to connect via their preferred payment method and top up the app. While the app is a normal crypto wallet and does not require identification, our payment partners will need to maintain compliance in their jurisdictions and may require personal details. We are also investigating the integration of p2p cash transactions.

For the second challenge, one of the most interesting thing to come out of the token ecosystem in the past 2 years is the creation of incentive structures with quick feedback loops and maximal alignment of interests. To this end, we are creating a utility token, DEP, that will be used in our ecosystem. Any time fees are paid or revenues are earned, they will have to be done so via our utility token, which is then burned after using. They are ‘single use’ tokens, unlike the many ‘multi use’ tokens generated over the past 2 years. This means that increased usage of our products results in higher demand and decreasing supply of the token, which in turn incentivizes token holders to evangelize for us.

This is the first of several adoption focused incentive systems we are designing, with a goal of creating a decentralized and distributed army of evangelists that convince people to switch over their spending to the Mosendo app. Until then, its Rossco and I ourselves with our boots on the ground, getting people setup one at a time.

Making A Business

In order to fund these incentives, we have several ways to monetize.

At first, we can earn commissions from our fiat-to-crypto gateway partners. Once we integrate synthetic foreign currencies, there are some arbitrage opportunities available.

After we have significant adoption with Mosendo, we will start rolling out our decentralized personal finance manager, Flipsr, which allows users to borrow or lend and earn interest, diversify their assets, and invest in synthetic assets such as US stocks.

Through these decentralized investment products, we can earn fees which are paid in our utility token, DEP. It’s important to note that the usage of the token is invisible to the average user, the purchase of DEP to pay ecosystem fees happens automatically on the backend.

We’re On A Mission

So that’s the plan more or less! Obviously there are way more details, far too many to get into here, but I hope this has been an informative overview of what I’ve been up to over the past few months.

The reason I’m doing this, is the reason I was drawn to Bitcoin and cryptocurrency in the first place. While I believe that government and regulation are necessary, I also believe that concentration of power breeds corruption and is eventually usurped by the selfish and greedy for their own benefit.

Through the use of decentralized currencies and technologies, we can create better systems that are less concentrated and less corruptible, and where better to start than the most effective system of control used by the most corrupt governments today: money.

My own frustrations while young with not having ownership of my digital money has led me down a path of discovering how money is used and misused all over the world and driven me to have a strong need to make sure this exists in the world.

Mosendo for me is really a culmination of my desire for this to exist since I was young, before cryptocurrencies existed. Watching fintech products such as Dwolla, Simple, Venmo, Revolut, and many more, try to fix the existing banking problems, hoping they will succeed, but watching them fall short due to regulation every time.

After observing these pioneers for decades, I think we’ve finally hit onto something that will completely reshape the way money is used around the world, and so it is now my turn to take a swing at making it happen.

As Roman general Hannibal Barca says, “we will find a way, or make one.”

Regards,
Tom Howard
Co-founder and Chief Strategy Officer
Mosendo

Thanks for taking the time to read this! If what we are doing resonates with you, there are a few ways to continue this conversation and bring this vision to life.

More articles about the Mosendo mission

• It’s 2019. Where the f*ck is Our Global Peer-to-Peer Electronic Cash System?
• 5 New User Challenges, From The Front Lines Of Crypto Adoption!
• The Surprising Complexity of Money Management on a $2-a-day Income

Why I’m doing a startup again after years of investing was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

What is DAI?

DAI is a stable cryptocurrency on the Ethereum blockchain.

1 DAI is worth roughly 1 USD.

Since DAI is a token on the Ethereum blockchain, it allows smart contract creators to accept DAI instead of volatile ETH.

DAI is also a fully decentralized cryptocurrency, which means it has many of the same benefits of Bitcoin and Ether: you can be your own bank.

Other stablecoins such as USDT are generated by a central organization holding USD in a bank account. This is not decentralized as the bank account can be shut down or the company can decide to take back the token.

The DAI cryptocurrency is peer to peer electronic cash.

How does it work?

Imagine Ethereum investors who are holding large amounts of ETH. They believe in the long term value of ETH and want to continue holding it, but they also need some spending money.

Using the MakerDAO smart contract, they can deposit and lock up the ETH, and in exchange, generate 67% of it’s value in DAI tokens.

This way, they can get some spending money from their ETH, but if the ETH price goes up a lot they will still own all their ETH and can still make money off it!

Of course, if the ETH price goes down too much, they will have to pay back their loan or risk losing their ETH.

Through this process, DAI is generated and spent, making it available for anyone to use as a decentralized and stable cryptocurrency.

tl;Dr

1 DAI = $1 USD

DAI is decentralized cryptocurrency

DAI is backed by scarce crypto assets

DAI is P2P electronic cash

Use DAI with Mosendo.com

What is DAI and how does it work? was originally published in Mosendo on Medium, where people are continuing the conversation by highlighting and responding to this story.

Originally written January 2017.

“This app was supposed to be built in 2 months. Now you are telling me, 10 months later, that it will take 2 months to finish it. How the hell did this happen?”

I did a lot of software product consulting in 2016, and I was surprised that a good number of the gigs were to fix projects which had gone completely off the rails. At the same time, I had the privilege of being a fly on the wall to a few software venture failures, while advising friends in the midst of the trough of disillusionment.

It seems that tech venture failures still happen surprisingly often, despite having experienced tech builders being on the team, and despite knowing the common ways these things fail.

If it’s not inexperience or lack of knowledge that is causing these failures, what is causing them?

All the projects I’ve looked at all had a wide range of symptoms and were in very different situations, but I can still tie them all back to a single point of failure, which led to cascading effects in different directions.

That single point of failure results in a lack of awareness around a few key principles, and a misunderstanding of what a software product is. Some of the points I will try to make are subtle, and therefore easily overlooked, but they make a world of difference.

First, let's review the symptoms of 3 different projects I’ve observed in 2016.

Failure Symptoms

The Scrappy Startup

In May 2016, I was brought in to rescue an app project for a new marketplace startup. They already had their backend in place, and needed the app for their service providers to participate in the marketplace.

They had started the project in July 2015 with a 2–3-month timeline, but the project was about 10 months behind schedule, and only 50% complete!

For some reason, the more work that got done on the app, the longer the timeline to completion seemed to get! The CEO was understandably quite upset that it was taking 6x longer than expected.

These delays have negatively impacted their efforts to raise money from investors and shaken the confidence of their service provider partners.

The Corporate Venture

Throughout 2015, I was a sideline observer to a new product being built by a capital-rich company. They had already validated and gotten user traction with an MVP and good marketing, and were ready to build the production version of the product.

They brought on an experienced development team, with a six-figure budget, and a 3-month timeline to launch the next version.

12 months later the budget had gone up 5x and they still did not have a working product.

After all that time, they ended up missing the window of opportunity and the idea they were building was no longer relevant to their customer base.

The Funded Startup

In September 2016, Rand Fishkin published a blog post about Moz’s attempts to build several software products, many of which they ended up shutting down.

Moz had raised $28 million between 2012–2016 to build out and market these software products. The software ended up taking about 2 years longer than expected to launch, and once launched, the products fell short of expectations, causing layoffs, and the shutdown of a few products.

Remember, all the people involved in all of these projects were smart people, and some of the people on each of these teams have even worked in successful tech startups for awhile! If the tech people are still failing at building software, how can a non-technical have a chance at success?

Why Did They Fail?

At first glance, one might place the blame squarely on the engineering teams on these projects. Maybe if it took them 2–6x longer to build the software, they were just not good at software development?

But that is like blaming a broken bone for being too weak to withstand jumping off a 5 story building. Sure you can try to strengthen your bones, but it's probably a better strategy to take the stairs.

Similarly, each of these companies lost sight of reality by retreating into a sandboxed room of “software development” and then tried to leap off the roof of product-market fit.

The single point of failure that ties them all together? They all failed to continuously reorient their product, people, and process with reality.

They all got off to a strong start, got a bit of traction of some kind by exploring reality, talking to users, testing the market.

But at some point, they all said, “ok, we now know enough, let's retreat into a room for a while and built out this grand vision we have formed.”

And the thing that often happens when you do that, when you stop checking in with reality, is those unanticipated aspects of the product experience pop up, and the team just takes their best guess at fixing it. And then more things pop up on top of that, and more best guesses are built on top of it, and you end up with a product full of compounding risk and complexity.

Here is how it broke down for each of the projects:

The Scrappy Startup did not have a strong understanding of the needs of their service providers and their process. They started the project with a spec that was not much more than, “build an app where they can manage leads and customers in the field.” In fact, they discovered over the course of the year that a very complex business process flow needed to be built out. Often these discoveries happen while the developer was building a feature and asked, “wait, what happens when they do X?”

The Corporate Venture had a similar issue, they wanted to rebuild their MVP to be less buggy and more user friendly, but in the process of trying to think through the coding of each part, the engineering team discovered that there were conceptual flaws in the MVP that needed to be completely rethought, which resulted in continuous feature creep instead of continuous learning.

The Funded Startup seemingly also lacked the feedback loop between their product ideas, and what they built. I’m sure there were unforeseen technical complications due to the ambition of the projects, but the fact that the usage of the products in some cases completely missed the mark suggests that they were not regularly re-orienting their products with the needs of their users.

—

The common thread between all these failings is a lack of orientation with reality.

It is this lack of regular re-orientation with reality that causes confusion on development teams, which leads to massive delays in completing the software. In fact, usually no one on the team really knows what “complete” even means!

When there is a lack of clarity around a product, the teams get confused, and they spend a lot of time figuring out how something “should” work. But in fact, the only way it “should” work is the way your users want to use it. By not actually discovering how users would use it, you have vagary & confusion, not clarity & execution.

Key Principles for Preventing New Product Failure

A startup’s purpose is to explore reality and discover new possibilities. The entire startup system should be organized to consume reality and learn from it.

Regularly re-orient with reality. Do this every week at a minimum. This means checking that your ideas and products work in the environment they are meant to work in. Often this will mean some sort of user testing, to see how users actually react to your product (not the idea of it). Setup systems that get the people on the team, the product decisions, and the process, regularly adjusting to learnings from reality.

Short Iterations. Tackle and finalize fewer things at one time. This applies to both the innovation/discovery process and the execution process. Quickly test new ideas once you have them so you don’t waste time planning around ideas that won’t work. Build and finalize one feature so you can test it in production before moving on to the next one.

Get specific. Broad concepts do not give results. Digging into the precise details of what will happen and how is the only way to create an experience that works. Software, in particular, is the collection of 1000s of precise definitions, and can only work with specifics. “Amateurs talk about tactics, but professionals study logistics.”- Gen. Robert H. Barrow, USMC

Product is User Experience. The product is NOT the tech that makes it. How your product is made is irrelevant to the user, they only care that they get the experience they want. Whether my Uber is a veteran taxi driver, a student, or a robot doesn't matter to me as long as I get from A to B, quickly and safely. The extremely complicated real-time matching & routing software they have built to make it work is of no importance to the rider.

What Can You Do Right Now Increase Your Success?

Applying these principles into the whole organization will take some time and thought, but there are steps you can take right now to slow down the compounding accumulation of product risk.

1. Take whatever product ideas you have right now and go show them to your users. It doesn’t matter what state the idea is in, as long as its somewhat tangible. This could mean a semi-functioning software, a polished design in InVision, wireframes, or even sketches on paper. If some parts are non-functional, you can manually make them appear functional, like wizard of oz.
2. When talking to your users, don’t pitch them, don’t explain your idea, just see what happens when they try to use your product.
3. Ask them to say their thoughts out loud while they go through the experience.
4. Observe all the problems that arise from the experience that have nothing to do with the non-functional state of it.

That is the reality of your product. Never retreat from it.

Related Stories from DDI:

• Which is More Promising: Data Science or Software Engineering? | Data Driven Investor
• 5 Industry Transforming Blockchain Applications | Data Driven Investor

Why were tech products still failing in 2016? was originally published in DataDrivenInvestor on Medium, where people are continuing the conversation by highlighting and responding to this story.

The Guide To Designing A Magical Chatbot Experience — Part 2

The framework for conversational design

Originally written June 2016.

Part 1 discusses ELIZA’s 8 Laws For Designing Magical Chatbots and chatbot design choices. Read it here.

Magical Bot Experience Design Framework

When building with the ELIZA’s 8 Laws in mind, you should not be making a one size fits all experience to shoehorn all your users into. Instead, the interaction with the bot will vary heavily from user to user, based on their persona and usage history.

To achieve this, you will need to think about the different paths your users might take, and how you can personalize for them on each path.

There are an infinite number of paths a user can take, so we will need to push users towards a finite set of paths we design, and not allow them to follow or have a need for paths we have not designed.

There are an infinite number of paths a user can take, so we will need to reduce the number of paths they do take, and stop them when they leave our area of expertise.

First, we define our Domain, so we know when the user has wandered too far.

Then we break down the paths we want to support, by brainstorming User Personas, Intentions, and Phases.

Finally, we get ready for implementation by defining Contexts and Stories.

Domain

You can’t compete against Facebook, Apple, etc when it comes to making a general AI that can handle anything a user asks. Instead your bot needs to be really good at a specific domain, and focus only on that domain.

The more specific and limited the domain is, the easier it will be to provide a magical experience.

When a user starts going outside the domain of the bot, it should gently let the user know they are going outside the bot’s area of expertise, and gently remind the user about what it is good at.

What is the domain your bot is specializing in?

User Personas

Once you have your bot’s domain defined, you should be able to identify the different types of users who would get value from your bot. You want to be designing your bot experiences around those users.

Think less about things like age bracket, and more psychological state of mind.

What different types of people would use your bot, but get different types of value from it? Is the use case of one user persona different from the use case of another?

A well designed bot will determine the user persona early, and provide the right experience for that user.

For instance, a travel bot might have the following personas:

Each of these personas travel in distinctly different ways.

The business traveller wants comfort, convenience, and speed; cost is probably not an issue. They also will be quick to make decisions about their travel.

Conversely, the 9 to 5 vacationer only gets to travel once per year. They want to maximize the value they can get for their money, and will spend a long time deliberating on the best option for them.

Intention

What is the user trying to accomplish by talking to your bot? Typically they are looking for either:

• Information
• To complete an Action

Think about the different types of information a user might be looking for, or the actions they may want to take. This can get incredibly detailed so it may be more practical to think about intentions at a higher level right now. For example, ‘furnishing my office’ instead of ‘want to buy this desk’.

What high level types of activities do your users hope to accomplish with your bot?

Travel bot example:

Phase

Bots should be aware of contextual and conversation history, so they need to be able to keep track of users in different phases of their interaction.

Phases are major shifts in the mindset of the user, usually over longer periods of time. This is not referring to the different steps of a shorter conversation.

A traveller who has booked a flight is now ready to book a hotel. Someone looking for a vacation destination is definitely not ready to look at hotels.

Of the different personas and their intentions, what phase are they in with their intention? Are they pre-action, mid-action, post-action?

Travel bot example phases:

Each category consists of a very different user mindset, and the user will be looking to accomplish very different things.

Consider following up at the end of their natural lifecycle. You already know what hotel they booked and when, why not check in to see how their stay was? Provide better recommendations in the future based on that feedback.

Notice that in this example, the phases apply to all the user personas, however the information and actions they seek for each phase may be different. For your bot, the phases may only apply to certain users.

Contexts

Contexts are the information you need to help place the user in the correct user story. Using context, you can infer their persona, intention, and phase.

Sometimes the context is available from their profile or channel and the fact that they are talking to your bot is a context. More often, you need to gather context from their queries, or from their conversation history. Worst case, you will need to actually ask them to provide some context for you.

Guidelines:

• Infer context as much as possible.
• Favor asking for information at the point that it is needed, not before.
• Do not overwhelm the user with a bunch of questions during on boarding.
• Be creative. Instead of asking them a bunch of questions outright, try to think about what you can infer from fewer questions.

Based on your answers from the Persona, Intention and Phase section, think of the different contexts required to put the user in those buckets. You will want to keep the number of contexts minimal at first, but you should expand and give them more depth as your bot grows.

Travel bot example contexts:

If you have a user’s Nationality already, getting their current location can help you figure out a lot about their Persona & Phase. If they are in their home country, they are probably not a digital nomad or expat. If they are already located in the country they wish to travel to, they are in the arrived Phase.

Notice that Persona, Intention, and Phase are contexts, but we often use other contexts to infer those contexts.

Stories

Now you tie everything together and write out all the user stories.

It might be useful to use a spreadsheet, or mind mapping software such as MindMeister for this part.

Core Stores

You will need to think about all the plausible combinations of Persona, Intention, Phase, and Context, and create stories for each.

Start with the cases where you have perfect information, and plan the ideal path. These cases are the core value of the bot, and should demonstrate the best of the domain expertise.

Example

Spring’s shopping bot nails the core story of discovering and buying clothing items. The user immediately knows what they can purchase there as they are presented with store categories. Then they are quickly shown product images which are relevant to their needs.

Getting Context

Think about what should happen when some information is missing. What can the bot do without that information? How should it get the needed information?

The bot might need to prompt the user for some information, or maybe it can find the context in the conversation history.

Try to only prompt the user for more information as it is needed, with as little back and forth as possible. If 3 bits of info are needed, can the user provide them all in one response?

Counter Example

The CNN bot is a great example of what NOT to do. They don’t seem to make any attempt to connect me to my browsing history. They also did not make use of the context I provided in the chat session.

When asked for stories about space, CNN bot returned irrelevant articles that included the word ‘space’, and never brings that interest up again. CNN is missing out on a huge opportunity for personalization and user re-engagement by completely ignoring context.

Onboarding

Once you have some core stories written, write the introduction and onboarding sequence.

In the introduction, the bot will need to explain the scope of its domain expertise, and give some hints about paths the user can take. Then the bot may need to initiate a short onboarding sequence, where it gathers the contexts needed to put the user on the right path.

Do not try to collect every possible context it might need at the beginning. Save context prompts for the beginning of the Story that the prompt is needed.

Example

GYANT, a medical diagnosis bot, has a good example of a quick, friendly onboarding. It gathers the bare minimum information it needs from user, and moves on to the core use case as soon as possible.

They could improve by asking a few diagnostic questions first, to get the user invested in the outcome, and then asking for some of these details as they become relevant to the diagnosis.

Going Off The Rails

Make sure to include stories for when the user goes off the rails. If they leave the domain of the bot, gently remind them of the available domain expertise.

If they are mid story, think about why they might go off the rails. Did they change their mind? Did they mess up a previous input? Give them a way to start the current story over, or leave the current story altogether.

When you are writing your stories, be sure to clearly define the contexts needed for a story, as this will be necessary for the technical team to implement the story.

Example

The 1–800-Flowers bot handles going off the rails really well. It first prompts the user to fix the local context, and if the user still has issues, it gives them an escape hatch, allowing the user to keep going, start over from the beginning, or talk to a person.

Iteration

It does not end here!

AI chatbot making is a never ending job. Users constantly surface new use cases. AI needs ongoing training.

This framework is meant to be a living document. It will need constant improvement and refinement based on user feedback.

In fact, I advocate getting it in front of users as soon possible and learning from the interactions, because it’s nearly impossible to predict all user behaviors ahead of time.

You don’t need to have a functional bot to learn from user interactions, you just need users who think they are talking to a bot. You could create a mechanical turk, where the user thinks they are talking to a bot, but a human is actually responding as if they were a bot.

Feedback

We are in the very early days of bot design, and this framework is very much a work in progress.

I’d love to hear from you on how this framework worked for you, and where it broke down. Just leave a message in the comments!

Originally written June, 2016.

This response is to the following question posed by an entrepreneur in a private forum

I have zero programming skills, but like everyone else, enjoy the idea of creating a software that other people would find useful.

I’d love to know more about how apps work, what goes on in the background, associated costs, things like that. The problem as with anything else is simply not knowing where to start.

Is there a way to learn all of this information without getting nitty-gritty into the details? Like, the “The Entrepreneurs Guide to Software Development, Non-Technical Edition”?

For example, I watched a video recently of Ray Dalio explaining how the economy works. I didn’t have to learn a lot of economics jargon, read a bunch of books on how to become an investor or trader, or anything like that…. it’s just a very straightforward explanation of how a seemingly sophisticated system can be easily explained.

The Beginning of Software Development

Imagine building a skyscraper, 1 atom at a time.

In the old days, that is what programming was like, so people mostly just built smaller things, more on scale of a hammer.

Then a layer of abstraction was invented, a programming language, which was a bit like having access to atoms that were already pre-arranged into molecules.

People could now build a toolbox of tools.

Then another layer of abstraction was invented, which was like having access to pre-made bricks, made up of particular molecules.

People could now build houses.

Modern Websites

Layers of abstractions have been added over the decades, where the community works hard for a long time on incredibly complex pieces of code that solve hard technological problems.

At some point, they get it right, it is accepted by the community, and it gets abstracted away. No one has to worry about the nitty gritty details anymore. At least most of the time.

Now where we are today, is like having pre-fabricated and mass produced sections for skyscrapers.

Anyone can come along and say “Hey construction worker, I want these basic skyscraper features which Acme Prefab has readily available.”

Then the construction worker looks at the simplified installation instructions provided by Acme Prefab, drives some rebar down, places the prefab sections, pours some concrete, and a few months later you have a Skyscraper!

That skyscraper is something like a Wordpress site. Most of the use cases for WordPress have been established, and basically everything you need to do has already been done.

Making Web-apps

Now what if you want to create a system that is VERY specific to your business?

Like an operations tool for a large and complex organization, or a highly specialized social network, or perhaps your business solves technology problems which have not yet been solved by the community.

Well odds are, Wordpress won't cut it, all the premade plugins and snippets out there aren't made for your use case.

So now your Skyscraper has all of these features that are very specific to your business, and the prefab sections just won't cut it.

You need to bring in engineers that will re-design and re-engineer every part of your building so it fits to the specifications you need.

Perhaps your Skyscraper design will work with prefab windows, or maybe your windows need a unique shape, or need a new type of glass.

The engineers will have to peel back those layers of abstraction, whenever they don't fulfill your specific needs.

But once it is built, you now have Facebook.

Understanding Costs

These days it's very rare to need to peel back the layers of abstraction very far, which is why basic web apps can be created for $5000 instead of $5 million.

In fact, Ruby on Rails, one of the more recent abstractions, has made it a lot easier for people to learn programming, since you don’t have to deal with a lot of the complicated and tedious things like the inner workings of the HTTP protocol.

Ruby on Rails even uses the a “scaffolding” metaphor as one of their abstractions!

—

The thing is, every layer of abstraction inherently creates inefficiencies in your programming.

In order for each “black box” to produce the correct outputs every single time it is given inputs, it needs to handle every feasible use case, even if the majority of those use cases could not happen in your specific program.

You can dig into every single layer of abstraction, and build a custom version of it that is optimized for your specific business.

In this way, all technology projects can be nearly infinitely optimized.

I think this is where most non-tech people run into issues with cost overruns.

The non-tech person asks for a way to manage all their content and have a good looking blog, and the engineer can install Wordpress and a good looking theme in under an hour.

Then the non-tech person says, “well hey, since that was so fast and cheap, let's also build out a social network for all my users.”

But the problem here is there are so many uses cases for a “social network”, and there is no pre-made software out there that efficiently handles all of these use cases in an elegant way.

So now the developer spends lots of time looking for a good solution, and finds that in order for one to meet their needs, they gotta do some custom work. This may take 100 hours.

Okay, so that is 100x more than the CMS site took. What a huge difference!

And if they built their own custom social network, it could easily take 10,000 hours!

This is can be difficult for non-tech people to fully grasp. How can “Social Network” be 100x or 10,000x harder than “CMS”?

Good vs. Bad Developers

A good developer will know when it makes sense to optimize, and when it is good enough for your business as-is. They can be difficult to find though.

I once took on a project where an aspiring entrepreneur had spent a year and over $100k to build out his web app idea.

The project was still not finished, and he was running out of money.

We threw out all the code and rebuilt it for him, bug-free and better than before, in 2 weeks, for $5000.

Why had the other developer taken so long to provide an inferior product?

He had a chip on his shoulder about using pre-made components. He considered using “someone else’s code” to be “cheating”, and so he tried to build the entire app from scratch.

This was completely unnecessary for this guy’s business idea, there were plenty of libraries and frameworks existing that did what needed to be done.

Making Business Decisions

I think a good heuristic for a non-tech person to keep in mind, is that as you get more detailed, costs increase exponentially rather than linearly.

This is especially true of complexity, where the effort of having 3 modules (say Accounting, Payroll, and Time-tracking) interconnected is at least 1 order of magnitude greater than the sum of all 3 modules in isolation.

—

Does this help you understand?

If you are interested in learning specifics of web technology, I HIGHLY recommend this Web Development course on Udacity, it’s with a co-founder of reddit, Steve Huffman. I thought the web was incredibly well explained.

If that is too advanced, one of their beginner courses would probably be good for you.

The Difference Between an Engineer and a Programmer

The Software Engineer is akin to the physical engineer responsible for figuring out the complicated bits of the Skyscraper, whether that is designing the perfect prefab sections that will fit most situations, or engineering an entire skyscraper from scratch.

The Software Engineer counterpart will be the one you want actually creating the layers of abstraction, or replacing them in your project.

The Programmer is akin to the construction manager/worker, who has an understanding of engineering principles well enough to assemble the engineer parts, and make non-structural changes to the plans as needed.

Programmers are great at connecting the pieces, but not so good at understanding complex systems and second or third order consequences of a particular implementation.

Most of us probably only need a Programmer in our business, someone to assemble the existing tools out there and make light customizations to our projects.

Another common mistake made by non-tech people is hiring a programmer to do an engineer’s job. This results in massively ugly projects with tons of bugs and huge cost overruns.

Beware the programmer who thinks they are an engineer.