<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Isaac Yuen on Medium]]></title>
        <description><![CDATA[Stories by Isaac Yuen on Medium]]></description>
        <link>https://medium.com/@isaacy?source=rss-146690383241------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*iyOJGrDaowd_cEPuMrToeg.jpeg</url>
            <title>Stories by Isaac Yuen on Medium</title>
            <link>https://medium.com/@isaacy?source=rss-146690383241------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 17:10:14 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@isaacy/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[PSD2 End-game, 5 days countdown]]></title>
            <link>https://medium.com/@isaacy/psd2-end-game-5-days-countdown-3675b6d98a72?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/3675b6d98a72</guid>
            <category><![CDATA[3d-secure]]></category>
            <category><![CDATA[emvco]]></category>
            <category><![CDATA[psd2]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[ca]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Tue, 10 Sep 2019 00:51:56 GMT</pubDate>
            <atom:updated>2019-09-10T18:18:39.790Z</atom:updated>
            <content:encoded><![CDATA[<h3>PSD2 end-game, 5 days countdown</h3><p><em>This is a continuation of the blogs (see </em><a href="https://medium.com/@isaacy/psd2-6-months-count-down-3a0b0e38480"><em>this</em></a><em> and </em><a href="https://medium.com/@isaacy/psd2-5-months-count-down-7704d73acd11"><em>this</em></a><em>) around </em><a href="https://eba.europa.eu/-/eba-publishes-opinion-on-the-implementation-of-the-rts-on-strong-customer-authentication-and-common-and-secure-communication"><em>Payment Services Directives 2 (PSD2) strong customer authentication to be in effect in EU on Sept 14, 2019</em></a><em>! And of course, all thoughts expressed here belong to me and aren’t related to my employer.</em></p><p>It has been a long haitus since the last update. As our team at Microsoft (and other merchants) were heads down in getting ready for Sept 14th date, it now seems that this isn’t a true <em>deadline,</em> now that <a href="https://eba.europa.eu/-/eba-publishes-an-opinion-on-the-elements-of-strong-customer-authentication-under-psd2">European Banking Authority</a> has allowed the central bank of each country (what EBA referred to as <em>national competent authorities</em>) to decide if SCA would be enforced or delayed.</p><p>Note: For those who are interested in the latest list of delayed countried, Stripe has mostly <a href="https://support.stripe.com/questions/strong-customer-authentication-sca-enforcement-date">up-to-date list</a>, and PaymentsCompliance also has a <a href="https://paymentscompliance.com/table/psd2-strong-customer-authentication-tracker">geo-tracker</a> on the latest country-by-country status.</p><h3>SCA: Who is on? Who is off?</h3><p>At this point, it is pretty clear to most people that Sweden is likely the only major economic power in Europe that will enforce SCA at a country level. From Stripe:</p><blockquote>On 6 September 2019, the Swedish regulator <a href="https://www.fi.se/sv/publicerat/nyheter/2019/nya-regler-om-stark-kundautentisering/">acknowledged</a> the new guidance by the European Banking Authority. However, they expect to enforce SCA on online payments with Swedish cards starting 14 September 2019. Banks that require additional time can reach out to the regulator with a proposed migration plan.</blockquote><p>What we can confirm, is that we have seen 3DS 2.1 challenges from one of the larger Swedish issuers (there are only a few) so Sweden seems to be more readily than other countries. For example, I personally have not seen a 3DS 2.1 challenge from any of the eastern or Southern European countries, and it is scary to think that we will lighting SCA/3DS 2.1 experience across all 31 countries in one day.</p><p>So what makes the most sense at this point for merchants, seems to be a “managed/staged” rollout where each merchant will <em>enable</em> SCA experience on a country-by-country or even at <em>bank-by-bank</em> level. However, without a <em>harmonized </em>timeline<em> </em>(which EBA seems keen to not push), and the lack of transparency at SCA enforcement at per bank level, the costs for merchant to manage this rollout will only increase. And I’m not just thinking abou tthe the technical infrastructure and resources in managing for the rollout: a merchant with significant sizes will bear the costs for experimentations (to evaluate the actual readiness vs. costs), and the communication overheads between merchants, banks, and card networks (who have the most stake in this card payments to make sure that EMV 3DS is <em>good enough</em> as SCA). The continous maintenance over the next two years, will reduce the ability for all stakeholders to move on to new business opportunities and wear everyone out, and ultimately defeat some of the goals of SCA (reducing fraud and increase consumer’s awareness).</p><p>Throughout the journey, one important lesson I’m learning is that a merchant cannot bifurcate <em>authentication</em> and <em>authorization </em>when optimizing your checkout conversion and cost — assuming that both authentication and authorization will cost money (most in terms of customer drop-off, and the usage cost from the payment processors), to reduce costs you will likely want to trigger authentication only when it is absolutely necessary (not if the customer has good credit and when the issuer will not decline an unauthenticated payments). And if there is one advice I will give to merchants who are justing starting on their PSD2 readiness, I will suggest that they choose a MPI who have both <em>authentication</em> and <em>authorization</em> capabilities, and have demonstrated that they will be able to minimize the costs from <em>unnecessary</em> authentication (either via inclusive pricing, or optimized challenge engine that is able to not trigger unneeded 3DS).</p><h3><strong>No. 1 Challenge for merchant: lack of testing (or test cards)</strong></h3><p>However, at this point, by far the biggest challenge for merchants is the lack of testing or test cards. First of all, the issuers in general aren’t able to provide credit cards to a merchant for testing without a rigorous compliance oversight (and rightfully so). The most viable option for merchant is to leverage companies such as Applause to obtain test cards for testing. However, leveraging a third party for testing create another problem: in most cards, these cards are issued to a real identity residing in EU, and by far the most common authentication option from 3DS 2.1 option is SMS, which is usually sent to the tester’s own mobile numbers in Europe.</p><p>So for a non-EU merchant to validate its SCA experience, it typical requires a two-ways or three-ways coordination testing between the merchant, the third-party company, and its tester, usually during early morning hours in the US, to complete one challenge validation of a particular card. Obviously, such coordinated testing is time and resource intensive and simply not scalable beyond 5 or 10 cards. Yet what we discovered is that despite the standardization efforts between the cards networks via EMVCo, the actual implementation between ACS, networks will vary, and the 3DS challenge UX is also highly fragmented between ACSs, making such coordinate testing a lifesaver in some cases; otherwise, the customer would not be able to complete the flow.</p><p>I expect that UX fragmentation in 3DS 2.1 protocol will be a big issue in the post PSD2 world, and the best solution is for issuers to support native/app-based 3DS experience, giving the experience control to back to merchants who knows their customers the best.</p><h3>The future of EMV 3DS</h3><p>The most shocking news from the <a href="https://eba.europa.eu/-/eba-publishes-an-opinion-on-the-elements-of-strong-customer-authentication-under-psd2">EBA opinion</a> is that it invalidates EMV 3DS <strong>as-is </strong>as a PSD2-compliant authentication mechanism. Merchants (and processors, issuers) have been pouring tremendous resources in the past two or three years, into supporting 3DS2 because card networks have been pitching that EMV 3DS is the good enough as the SCA. So to hear the EBA said otherwise just 2 months before the deadline made everyone (to put it mildly…) <em>gravely concerned</em>.</p><p>By now, it is likely too late to explore alternative: the ecosystem across card networks, issuers, and merchatns have already invested too much into this to restart, so the next question is how should the ecosystem evolve EMV 3DS to become PSD2-compliant? More importantly, how to incorporate the <em>inherence element?</em></p><p>I expect that in the next two years this will most likely be the place where different players will be pushing products supporting authenticaiton with <em>inherence</em> element. Already one of the networks are proposing a <a href="https://payments.cardinalcommerce.com/what-is-visa-delegated-authentication"><em>delegated authentication</em></a> model where merchant will take on a more active role in authenticating the user instead of through 3DS protocol. This absolutely makes sense since merchant app or web pages is directly interacting with the consumers/buyers and can apply unique authentication techniques, such as biometric fingerprinting, or even behaviorial biometrics. Of course, why EMV 3DS will still be needed in that case though, is for another discussion. I do suspect that processors who can provide such capabilities to their merchants (without incurring new costs) will be the winning solution, so look out for such <a href="https://www.biometricupdate.com/201904/stripe-acquires-biometric-software-provider-touchtech-payments-to-support-gdpr-requirements">processors </a>or vendors!</p><p>Now, all fingers crossed for Sept 14th date….</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3675b6d98a72" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[PSD2: ecosystem readiness, 3DS 1.0 problem, and exemptions considerations]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@isaacy/psd2-ecosystem-readiness-3ds-1-0-problem-and-exemptions-considerations-59927be73227?source=rss-146690383241------2"><img src="https://cdn-images-1.medium.com/max/2600/1*yoi4GPDN94gR4wdijp3toA.jpeg" width="5435"></a></p><p class="medium-feed-snippet">As we are only 4 months away from PSD2, there are more debates around PSD2 now in payments conferences around the world. Our team just&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@isaacy/psd2-ecosystem-readiness-3ds-1-0-problem-and-exemptions-considerations-59927be73227?source=rss-146690383241------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@isaacy/psd2-ecosystem-readiness-3ds-1-0-problem-and-exemptions-considerations-59927be73227?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/59927be73227</guid>
            <category><![CDATA[card-payment]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[3ds2]]></category>
            <category><![CDATA[psd2]]></category>
            <category><![CDATA[authentication]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Mon, 27 May 2019 06:02:21 GMT</pubDate>
            <atom:updated>2019-05-27T06:17:48.468Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[PSD2: 5-months count down]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@isaacy/psd2-5-months-count-down-7704d73acd11?source=rss-146690383241------2"><img src="https://cdn-images-1.medium.com/max/847/1*_B44eWZPPm_S7c3zsb4U8g.png" width="847"></a></p><p class="medium-feed-snippet">It is only 5 months away from Payment Services Directives 2 (PSD2) strong customer authentication to be in effect in EU on Sept 14, 2019&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@isaacy/psd2-5-months-count-down-7704d73acd11?source=rss-146690383241------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@isaacy/psd2-5-months-count-down-7704d73acd11?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/7704d73acd11</guid>
            <category><![CDATA[psd2]]></category>
            <category><![CDATA[authentication]]></category>
            <category><![CDATA[3d-secure]]></category>
            <category><![CDATA[payments]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Fri, 26 Apr 2019 06:08:38 GMT</pubDate>
            <atom:updated>2019-04-26T06:08:38.150Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[PSD2: 6 months count-down]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-snippet">We are less than 6 months away from Payment Services Directives 2 (PSD2) in effect in EU on Sept 14, 2019! As someone who has worked on&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@isaacy/psd2-6-months-count-down-3a0b0e38480?source=rss-146690383241------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@isaacy/psd2-6-months-count-down-3a0b0e38480?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/3a0b0e38480</guid>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[credit-cards]]></category>
            <category><![CDATA[psd2]]></category>
            <category><![CDATA[ca]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Mon, 01 Apr 2019 05:11:40 GMT</pubDate>
            <atom:updated>2019-04-01T05:11:40.446Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[Money 20/20 2017: observations and reflections]]></title>
            <link>https://medium.com/@isaacy/money-20-20-2017-observations-and-reflections-ec00c8a3ec00?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/ec00c8a3ec00</guid>
            <category><![CDATA[fintech]]></category>
            <category><![CDATA[conversational-ui]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[money2020]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Sat, 28 Oct 2017 06:52:45 GMT</pubDate>
            <atom:updated>2017-10-28T06:52:45.272Z</atom:updated>
            <content:encoded><![CDATA[<p>It is my first time going to a conference like Money 2020 and it is an eye opener to see bankers/merchants/start-ups/technology firms all mingled together.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*8cgS0dFkpxmPVLrnr6V1XA.jpeg" /></figure><h3><strong>General Observations</strong></h3><ul><li>Several themes dominated the discussion: AI / blockchain — ICO / lending.</li><li>There is also a realization that customer experience and solving real customers problems is now emphasized over selling more things or optimizing the payment fee/interchange itself. Simpler checkout, deliver additional values to buyers, got the most eyeballs from the audience.</li><li>At the same time, banks and financial institutions are increasing aware of the social implications and untapped potential in the unbanked population. Unfortunately, the mainstream thought are about trying to pull these people back into banking products (pre-paid cards) instead of understanding the root cause.</li></ul><h3><strong>AI and ML</strong></h3><p><strong>AI </strong>is such a buzzwords that the conference dedicated a full day focused on AI. However, the presenters mostly focused on the implication of AI such as job replacement, or doomsday scenarios…</p><p>When you pull back from the futuristic AI conjectures to day-to-day AI application, it turned out that even basic ML can give you much of the knowledge, and Stripe shared how it used a simple technical call word2vec to segment its customers and discover hidden relationship. My colleagues and I were super excited at this simple techniques because there are so much you can do by uncovering these relationships (cross-selling, fraud estimation, financial advices) etc.</p><p>Finally, I should note that AWS is making a huge pitch to the industry by positioning itself as the sole leader in the AI /ML. I talked to multiple developers in the Money20/20, and AWS is their only choice for cloud deployment option.</p><p>The best part of the talk is probably finally hearing from Dr. Michio Kaku. He is one of my favorite physics author and he made a very entertaining talk.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KRXZa8ktgDs37es1xn2NGg.jpeg" /><figcaption>Dr. Kaku on the stage.</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Lb5dsVvMoNoc4189Jy5akg.jpeg" /><figcaption>Dr. Kaku on how to be a billionaire.. very Jeff Bezos</figcaption></figure><h3><strong>W3C Checkout Experiences</strong></h3><h4><strong>FIDO/Payment Requests API</strong></h4><p>Google positioned itself as the leader in both spaces.</p><p>Google leveraged the W3C work on Chrome to lead the discussion in <strong>FIDO</strong>, and emphasize that Payment Request is not only a simpler checkout model but also more secure by incorporating fingerprint / biometrics to increase the trust across web/android phones.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*udeMArnzeQ3RkvztcGfMGQ.jpeg" /><figcaption>From the FIDO workshop from Google</figcaption></figure><p>In the W3C session, it was more akin to an education of how W3C worked on Chrome. Google demoed Payment Request implementation with engineers from Airbnb and MasterPass, on both desktop and mobile. Google also positioned gateway token as the de-factor approach for W3C and ignored basic-card at all. I should note that AirBnB mentioned BrainTree as its processor.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gjIoSwnzpz0yX9P7xfrIjw.jpeg" /><figcaption>From W3C payment request</figcaption></figure><h4><strong>Network Tokenization</strong></h4><p>Nothing new there, except the fact that there is not much progress of the network token (from Visa) and the merchants (N) were impatient with the lack of progress. The merchants also want all the networks to work together instead of creating siloed solutions.</p><p>For the merchants, getting out of PCI scope and reduce security risks / costs trumps the implication in not keeping PAN vaulted in its own storage. As an example, N is already tokenizating its cards with its processors and saw costs reductions. In the short run, I don’t think that there would be a network tokenization solution, PSPs will tap into this opportunity instead. In another words, gateway tokenization resonates with the merchants and is here to stay.</p><h3><strong>Bitcoin / Blockchain / ICO</strong></h3><p>At the surface, main focus wasn’t so much of blockchain itself but on problems ICOs and why regulators justifying themselves why regulation is good, ICOs should be regulated. Behind the scenes, there are at least 3–4 bitcoin startups, and everyone is bypassing the traditional VC model by going directly to ICOs to fund some very crazy ideas including:</p><ul><li>Creating a new ethereum-like public blockchain in China, that focuses of ID and asset managements. (<a href="https://mvs.org/">Metaverse</a>). They are the only Chinese blockchain company at 2020 but they are gaining a lot of momentum.</li><li><a href="https://www.matryx.ai/">Matryx</a> — provide its only cryptocurrency to crowdfund STEM-related R&amp;D challenges.</li><li>The craziest one I heard is a startup that uses ICO to raise money from fans to buy a professional sport-team. They are backed by Isaiah Kacyvenski a former Seahawks player…</li></ul><h3><strong>China</strong></h3><p>I always have a special interests in China and how Chinese firms present themselves overseas.</p><p>These year there are around 6–7 Chinese companies at the Expo. Besides the obvious giants (Ant Financial, UnionPay) a new player called <a href="http://www.lianlianpay.com/international/">LianLianPay</a> has emerged.</p><p>Many PSP in China can been seen as nothing more than a Alipay copycat. Lianlian differentiated themselves by focusing on mobile payment and end-to-end solution for overseas merchants that want to sell in China. And the fact that they are going to hose the 1st Money 20/20 China at HongZhou, (the heart of the Alibaba empire) is certainly going to raise a lot of attention in the future.</p><p>But overall, most of the Chinese firms at 20/20 besides the big two (Alipay and Wechat) are all focusing on transfer money in and out of China, so there is nothing innovating. But focusing exclusively on Chinese / Asian opportunities out of China could be itself a significant market already:</p><ul><li>I chatted with a group that handle restaurant menu and POS called <a href="http://www.menusifu.com/">Menusifu</a> — interestingly even though they are based in NY, all the exec and management at Menusifu are Chinese (which is a great thing)! They initially only focused on Chinese/Asian restaurants in the NY and was able to got more than 3,000 merchants to use their service including little sheep hotpot!</li><li>Alipay did a demo of the an app that focus on travel experience for Chinese tourist overseas. However, the app not only handles pre-trip experience (hotel / flights) but extend to solve mid-trip problems, such as restaurants recommendations, travel maps and route suggestions, and post-trip issue like automatic tax refunds. Things that takes traditiontally 4 or 5 apps at least to do, is now completely managed in this single app, and save the tourist from micromanaging its trip over different app and the cognitive stress. I’d say that this is the most impressive demo I saw at Money 20/20 this year.</li></ul><h3><strong>Conversational/voice commerce</strong></h3><p>I chatted with a few leaders in this space, including CEO from <a href="https://conversation.one/">Conversation.One</a> and the <a href="https://www.chatlogix.com/">Chatlogix</a> team to learn what is the trends in the industry.</p><p>What is clear is that Amazon Alexa and Google Home are the clear choices for voice assistants and none of them are developing Skills over Cortana because the demand isn’t there yet.</p><p>I should note that most of these financial skills focus on replacing an ATM (like check balance, transfer money) but so far these ‘skills’ are confined within the boundary of that institution, and does not support checkout or inter-bank money transfer. In fact, ‘checkout’ over voice assistant isn’t on their radar yet, and we are hitting a limitation in these ‘skill’s unless they are able to inter-communicate with other skills to do more things.</p><h3><strong>Final Thoughts</strong></h3><ul><li>An end-to-end solution like Stripe is preferred over a standalone vertical product. The key is simplify the process so merchant focuses on their business instead of worrying about infrastructure/logistics.</li><li>Tapping the influx of Chinese tourists in selling luxury goods in the US.</li><li>Voice commerce will be big. No one quite know how to monetize it yet.</li><li>ICOs doubt aside, bitcoin/blackchain is to stay and create many new solutions that the financial incumbents aren’t willing to tackle.</li><li>Everyone recognized the ‘unbanked’ is a huge problem. Banks and networks still try to use traditional means to solve that, instead of understanding the ‘why’ and offer a solution. Bitcoin will be one of viable solution for the unbanked.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ec00c8a3ec00" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Payments and e-commerce on Voice Assistants: Alexa, Google Assistant, and Cortana]]></title>
            <link>https://medium.com/chat-bots/payments-and-e-commerce-on-voice-assistant-alexa-google-assistant-and-cortana-b4336c98dcd2?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/b4336c98dcd2</guid>
            <category><![CDATA[google-assistant]]></category>
            <category><![CDATA[alexa]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[cortana]]></category>
            <category><![CDATA[voice-assistant]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Sun, 02 Jul 2017 23:51:56 GMT</pubDate>
            <atom:updated>2017-07-12T05:52:48.532Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/496/0*L3v7TJx9QHC6HPVq.png" /></figure><h3>State of e-commerce on voice assistant</h3><p>2016 and 2017 are the pivotal moments when we start seeing a new tech trend gradually transition the industry from a ‘mobile-first’ to ‘AI-first’ mindset. As mobile app ecosystem reaches maturity, and consumers now experience an app-fatigue, start-ups and innovators are looking for the next disruptor.</p><p>In 2016 we saw at F8 and Microsoft //build, the rise of chatbot as the new app. What we slowly realize though, is that text input is too tedious to get stuff done compared to mobile app, and bots aren’t as discoverable as mobile app. As a result, consumers still used messenging platform mostly to talk with real humans, and the bot ecosystem dwindled.</p><p>In 2017, catching the AI wave, we witnessed “conversation as a platform” resurrected through voice assistants such as Alexa, Google Assistant. Despite the initial lackluster experience, Alexa slowly creeps into our house as a music player, radio, and home automation assistant. Then Google caught up with Google Assistant and launched Google Home. Today Alexa/Echo enjoyed the first-mover advantage with a dominating <a href="http://www.reuters.com/article/us-amazon-com-alexa-idUSKBN1840ZL">70% of market share</a> compared to 23% from Google, and has 10,000+ skills in its directory. In 2017, Skills has become the new App.</p><p>Now, the ultimate goals for Skills developer is probably to make money off it. So how does monetization work in these new voice-assistant platforms?</p><h3>Alexa</h3><p>Unsurprisingly commerce is a built-in capability in Alexa that allows you to buy from… the ‘<a href="https://www.amazon.com/Everything-Store-Jeff-Bezos-Amazon-ebook/dp/B00BWQW73E">Everything Store</a>’. When you make a voice purchase request, Alexa searches through Prime-eligible items from your order history and Amazon’s Choice items, and make use of your 1-click setting to complete the payments.</p><p>Making commerce very efficient to buy from Amazon comes at a cost though. On purpose or by accident, Alexa Skills does not offer any native payment function. As a result, third party Skills such as Domino Pizza needs to handle the checkout process without much help from Amazon…</p><ul><li><em>Me: Alexa, tell Domino’s Pizza to to buy me a Hawaiian pizza</em></li><li><em>Alexa: Sure, what size do you want</em></li><li><em>Me: a Large one</em></li><li><em>Alexa: OK, let me finish your order</em></li><li><em>Alexa: Sorry, we cannot complete your card. Please save your credit card on Domino’s Pizza website…</em></li><li><em>Me: ghhh…</em></li></ul><p>Other skills such as WingShop, Grubthat do checkout often do not have any menu selection, and only allow the customer to re-order his previous order. That user case ignores the fact that customers often want to have something different from what they ordered before, and most people often forget what they ordered.</p><h4><strong>Experience on ordering on Alexa</strong></h4><p>At this stage, ordering on Alexa is pretty much like making order on a phone: it is as tedious as speaking to a real person but then, and it lacks the ‘intelligence’ that understands the complexity of your situation or order, and makes customized decision that address your need. I believe that with the advance in AI, the ordering experience will be much better, but at this point, commerce on Alexa falls short of expectation.</p><p>Unless you’re buying from Amazon, of course :)</p><p>But then, you will already have your dash buttons, Echo Look, Echo Dot….</p><h4>What I like in Alexa</h4><ul><li>Simple order with Amazon can be done completely in voice</li><li>Has the most commerce-related skills in ecosystem</li></ul><h4>What can be better</h4><ul><li>Lack of payment support makes with third party skills hampers any meaningful checkout experience on Alexa</li><li>Alexa needs to be more intelligent in helping a buyer go through the checkout</li></ul><h4>What I expect to happen in Alexa in next year</h4><ul><li>Amazon Echo Show will have checkout built-in.</li><li>Amazon will follow Google / Microsoft by providing its own purchase API for third party skills on Echo Show</li></ul><h3>Google Assistant</h3><p>While most people focused on Google’s <em>forte</em> efforts in artificial intelligence, in a few years, we will look at 2017 I/O as a pivotal moment for Google Payments: Android Pay announced partnership with PayPal, and launched W3C Web Payments on Android and announce partnership with Shopify and Alipay. But beyond Android Pay (mobile) and Web Payments (web), <a href="http://ben-evans.com/benedictevans/2017/2/22/voice-and-the-uncanny-valley-of-ai">the next likely platform shift will likely be voice assistant (not bot)</a>, and Google is positioning itself as the de-factor aggregator by infusing commerce activity into its AI agent through the <a href="https://developers.google.com/actions/transactions/">Transaction API </a>on Google Actions.</p><p>At this point, the Transaction API is still in developer preview, and no merchant has launched a live Google Action yet. So I spent a few days on learning deploy Google Action, walked through the Transaction sample on github, and deployed the action on Firebase.</p><h4>Developer perspective</h4><p>Because of my work, I already had some experiences in Android Pay and Web Payments on Chrome. Transactions API for Google Action extends naturally from Android Pay. If you are already familiar with <a href="https://developers.google.com/web/fundamentals/discovery-and-monetization/payment-request/android-pay">Android Pay on Web Payments</a> or in-app purchase, then likely you will already have signed up via one of the processors that Android Pay supports: Braintree, Stripe, or Vantiv. When you integrate with Web Payments on your website, Android Pay help you to manage the checkout experience by helping your buyer selects his payment and shipping info, and generates a ‘gateway’ payment token and your can pass to your processors, to charge the buyer and complete the checkout.</p><p>Below is a code snippet from Google’s documentation, showing you the structure of the payment request on Chrome:</p><pre>var supportedInstruments = [</pre><pre>{supportedMethods: [ ANDROID_PAY ],</pre><pre>data: {</pre><pre>merchantId: &#39;02510116604241796260&#39;,</pre><pre>environment: &#39;TEST&#39;,</pre><pre>allowedCardNetwork: [ &#39;AMEX&#39;, &#39;MASTERCARD&#39;, &#39;VISA&#39;, &#39;DISCOVER&#39; ],</pre><pre>paymentMethodTokenizationParameters: {</pre><pre>tokenizationType: &#39;GATEWAY_TOKEN&#39;,</pre><pre>parameters: {</pre><pre>&#39;gateway&#39;: &#39;stripe&#39;,</pre><pre>&#39;stripe:publishableKey&#39;: &#39;pk_live_fD7ggZCtrB0vJNApRX5TyJ9T&#39;,</pre><pre>&#39;stripe:version&#39;: &#39;2016-07-06&#39;</pre><pre>}}}}];</pre><pre>//Web Payment on Chrome — payment option request example</pre><p>Google Actions’ Transactions API accepts the same “Gateway Token” model and basically you can copy/paste your Payment Request code to Google Actions and be done in a few minutes! Below is the equivalent payment request sample in Transactions API.</p><pre>app.askForTransactionDecision(order, {</pre><pre>// These will be provided by payment processor, like Stripe,</pre><pre>// Braintree, or Vantiv</pre><pre>tokenizationParameters: {</pre><pre>&quot;gateway&quot;: &quot;stripe&quot;,</pre><pre>&quot;stripe:publishableKey&quot;: (app.isInSandbox() ? &quot;pk_test_key&quot; : &quot;pk_live_key&quot;),</pre><pre>&quot;stripe:version&quot;: &quot;2017-04-06&quot;</pre><pre>},</pre><pre>cardNetworks: [</pre><pre>app.Transactions.CardNetwork.VISA,</pre><pre>app.Transactions.CardNetwork.AMEX</pre><pre>],</pre><pre>prepaidCardDisallowed: false,</pre><pre>deliveryAddressRequired: false</pre><pre>});</pre><pre>//Google Action transactions — payment option request example</pre><h4>Checkout User Experience</h4><p>As a pixel user, this is the first time I really started using Google Assistant. Going through the sample checkout flow, I was quite impressed at how Google Assistant structure a checkout through a conversation-driven experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/628/0*cr_V_dnFlEdszgb8.png" /></figure><p>Typically in an online commerce a buyer needs to select a shipping address. Instead of presenting a checkout page, the way Google Assistant handles this is to explicitly ask me to select the delivery address from my Google account, similar to how a human agent would help me to complete a purchase on a phone.</p><p>After I select the address, then I get to see the entire order and my payment method, and just say “place order”, authenticate with fingerprint or device password, to complete the checkout.</p><p>For the happy path at least, user is kept in context in the entire flow. I am not able to add a payment method within the flow, but Google Assistant provides a link to my Google account to set up any payment or address information. In fact, Google prohibits an Action to redirect the user to another site to complete the purchase. (see Google Actions’ <a href="https://developers.google.com/actions/policies/general-policies#transactions">policy</a>). Since the entire experience is kept within Google app, it makes it easy to be ported to Google Home, as soon as google is able to provide an efficient voice-based order review.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/369/0*hNjuXVCIuGR_vK2U.png" /></figure><h4>What I like in Google Assistant</h4><p><strong>Developer/merchant</strong></p><ul><li>Coherent developer story with Android Pay</li><li>A lightweight onboarding process</li><li>Integration with api.ai confirms Google ambition to infuse NLP and AI into every</li></ul><p><strong>End-user</strong></p><ul><li>Tight integration with your Google account and Google payments means I can</li><li>Google got it right that the user needs to be kept in-context and not be redirected to another place during checkout.</li></ul><h4>What can be better</h4><ul><li>Google Assistant feels more like a chat bot than an assistant</li><li>Lack of voice support in the order confirmation card.</li><li>Changing or adding a payment card is not well supported.</li></ul><h4>What I expect to happen in Google Actions in next year</h4><ul><li>Google to perfect the voice interaction in Google app, and enable Transaction API in Google Home as well</li><li>Google to infuse this Transaction capability into Google ad and checkout experience, on Android devices</li><li>Tighter integration with Android OS to enabling adding payment instruments</li><li>Blurring between the Android Pay brand and Google Payments, with Google Payments slowly taking over the brand recognition.</li></ul><h3>Microsoft Cortana</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/589/0*PLVRAWrSjDLfLhBa.png" /></figure><p>Not many people realize that Cortana also has the capability to invoke a checkout experience. In //build 2017 we launched <a href="https://blog.botframework.com/2017/05/10/Build/">Bot Payments</a> on Microsoft Bot Framework that allows a developer to initiate a checkout. Since Cortana is a channel supported by Bot Framework, it is naturally support payments.</p><p>A bot developer can integrate with <a href="https://docs.microsoft.com/en-us/bot-framework/dotnet/bot-builder-dotnet-request-payment">Payment Request API</a> in Bot Framework, and offer a checkout experience for his customers that is tightly integrate with Microsoft account. Similar to Google Assistant, a buyer on Cortana can complete the checkout with the address and payments stored in his Microsoft account.</p><p>We are still working on how to improve this experience, and if you have any ideas, please reach out to us at <a href="https://stackoverflow.com/questions/tagged/botframework">our Stack Overflow </a>forum.</p><h3>Where is Siri?</h3><p>Apple in WWDC made an announcement adds <a href="https://www.apple.com/ios/ios-11-preview/?&amp;cid=wwa-us-kwgo-features---slid-&amp;mnid=sWcJG3C2z-dc_mtid_20925e1t39169_pcrid_198753497450_&amp;muid=de63245f-8363-43d1-b55f-c5b9f1753f80&amp;mtid=20925e1t39169&amp;aosid=p238">money transfer via Apple Pay </a>in iOS 11. Noyes Payment Blog has a much <a href="http://blog.starpointllp.com/blog/?p=4670">better analysis of the economic benefits</a> that I refer you to read. What is unclear yet, is how this is related to Siri, and how much Apple believes in Siri as a intelligent agent to help you complete tasks such as checkout.</p><p>But since Siri remains a ‘closed’ system, and its app ecosystem lacks a conversation engine that can understand conversation like Api.ai or LUIS, I am doubtful we will see much more advance from Apple in this area.</p><h3>Summary</h3><ul><li>Merchant ecosystem through voice assistant is in an early “crawl” phase, and the ecosystem has not been created yet.</li><li>Amazon Alexa is likely continue focusing on commerce with the ‘Everything’ store and optimize the experience through 1-click.</li><li>As a result, third party Skills that do commerce needs to handle the entire checkout, and the experience is lackluster, particular in managing address and payments.</li><li>Google Assistant provides the smoothest checkout experience. But it still requires a screen to help user with more complex tasks, such as review order, select address, etc.</li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b4336c98dcd2" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chat-bots/payments-and-e-commerce-on-voice-assistant-alexa-google-assistant-and-cortana-b4336c98dcd2">Payments and e-commerce on Voice Assistants: Alexa, Google Assistant, and Cortana</a> was originally published in <a href="https://medium.com/chat-bots">Chatbots Magazine</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[E-commerce Checkout Design Deep Dive (Part II) — Address Collection]]></title>
            <link>https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-ii-address-collection-3229b80bc709?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/3229b80bc709</guid>
            <category><![CDATA[checkout]]></category>
            <category><![CDATA[payments]]></category>
            <category><![CDATA[address-validation]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[ecommerce]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Tue, 28 Feb 2017 05:59:43 GMT</pubDate>
            <atom:updated>2017-02-28T06:00:38.894Z</atom:updated>
            <content:encoded><![CDATA[<p>Given the importance of the usability of the checkout form, you’d think that most e-commerce sites would spend the efforts to optimize the UI so that the legit buyer can complete his purchase as easy as possible</p><p>What I found surprising is that the sites that process thousands or even million customer have not paid enough attention to the tiny things they can do to help their customers buy things on their own sites. For example, giving the right label for <a href="https://baymard.com/blog/address-line-2">Address Line 2</a>, or <a href="https://baymard.com/blog/credit-card-field-auto-format-spaces">auto-format the spaces in the card number field</a> are obvious changes that have clear usability improvements and involves trivial efforts, yet around 50% of the surveyed sites aren’t applying this trick and so opportunities are lost</p><p>In <a href="https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-i-payment-method-collection-bf9a52ac6d1d#.eueunel8g">Part I</a>, we looked at the common patterns on the credit card collection form. In this Part II we now move on to a burdensome process we all go through when buying things online: giving out your home address to the merchant.</p><p>Similar to how we approach credit card, we first formulate a set of criteria</p><ul><li>Is the user allow to select the country of the address? If so, where is the ‘country’ field placed?</li><li>Does the site collect 1 line or 2 lines for address?</li><li>Does the address form provide a descriptive label for a “line 2” such as apt. no or unit? Researches have show that <a href="https://baymard.com/blog/address-line-2">unclear “address line 2 description” leads to user confusions</a></li><li>For the “state” field, does it use full state name or state code?</li><li>Can it detect that the entered address is valid or not, and inform the user when the address is invalid?</li><li>Does it auto detection the address based on user’s input?</li></ul><p>Again, we categorize the the surveyed sites in three groups: <a href="https://baymard.com/checkout-usability/benchmark/top-100">Baymard-ranked e-commerce sites</a>, the top Internet e-commerce giants (Google, Amazon, eBay) and payment processors (PayPal), and then a set of hospitality service (Airbnb, Expedia) and Internet upstarts.</p><h3>Survey Result</h3><p>In our table below, we also highlight the results with different colors that express a common sense judgment on the feature’s usability:</p><ul><li>Green label denotes the feature that helps improve usability because it reduces the number of fields entered, or has better clarity</li><li>Orange label means that we are not sure if this feature helps or degrade usability</li><li>Red means that this feature hurts usability because it is more ambiguous, or make the flow more cluttered</li></ul><h4>Baymard Ranking ranked E-commerce sites</h4><p>I expected that these sites will set the baseline for common patterns in address form treatment. For the most part, my expectations are met. With the exception of Office Depot and Nike (which is ranked much lower than these other sites anyway), these address forms implement patterns that helps the usability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TYv0xJLyv8NzhtYSDiTX6g.png" /><figcaption>(**) — detect City/State based on zip code, (††) country field is selectable from drop down</figcaption></figure><h4>Address forms from Internet giants and payment processor</h4><p>The surprising observation is that Amazon, perhaps the biggest online merchant on this planet, actually has … let me put it euphemistically, design that are contrary to patterns that we observe from other merchants:</p><ul><li>The buyer needs to select the country at the bottom.</li><li>There is no meaningful test on the ‘Address Line 2’</li><li>It accepts arbritrary input for the ‘State’ field</li><li>Despite the vast number of address entries it possess, it does not auto detect even the city or state of the address for the user.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T5LxgtO-QnyLzY28J6ap2w.png" /><figcaption>Survey results from EBay, Google, Paypal, Amazon</figcaption></figure><h4>Address forms on on several hospitality or service-based merchants</h4><p>For hospitality or travel sites, there isn’t anything to ship, to address form is generally used for billing address collection.</p><p>A pleasant surprise is that both Airbnb.com and Snap Spectable sites have address designs that are on par with the top Baymard-ranked sites. Perhaps it provides a glimpse of why these two business have gone through tremendous growth in their user base —it could be that because the users just find it easier to use these sites compared to other sites.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*eSndSEFNf9Zpcg6NnbODZQ.png" /><figcaption>Survey results from Internet upstarts and hospitality service</figcaption></figure><h3>Observation for common patterns</h3><h4>Treatment for ‘Address Line 2’</h4><p>While almost all (except Groupon) supports an explicit “Address Line 2” field, surprising only <em>50% </em>of the surveyed sites have meaningful text for address line 2, e.g. ‘Apt No. Bldg (optional)’.</p><p>Given that more than <a href="https://www.census.gov/hhes/www/housing/census/historic/units.html">50% of houses in the US are single-family detached house</a>, we can assume that perhaps half of the site visitor may not bother to fill ‘Address Line 2’. A lot of screen real estates is taken up by a full line for ‘Address Line 2’ though.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZDbTGt_BjFo2iwJ55LXnPA.png" /><figcaption>Home Deport Address Line treatment</figcaption></figure><h4>Address validation and auto-detection</h4><p>The most labor intensive in the whole checkout process is likely filling out your address (and why <a href="https://support.google.com/chrome/answer/142893?co=GENIE.Platform%3DDesktop&amp;hl=en">Chrome’s form fill </a>is so useful). However, it makes complete sense why this is so hard though: the merchant needs to make sure that the address is correct, so that the goods can be delivered into the buyer’s house without problem.</p><p>Accuracy matters when entering your address. A few merchants have a few delightful tricks:</p><p>Take Target.com as an example. When I type in the zip code, it is able to automatically fill in the city and state for me (see below):</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/710/1*tV_jl1q9jxAqU_fMYZxn_w.gif" /><figcaption>Target city/state auto-detection</figcaption></figure><p>Baymard has a great guide about how this can be done: <a href="https://baymard.com/blog/zip-code-auto-detection">https://baymard.com/blog/zip-code-auto-detection</a>. Only 40% of the sites have this feature.</p><p>Baymard suggests a vendor to onboard to an address verification service. However, if you’re a medium/large size merchant, it is likely that you already have thousands of address records at your disposal, and you can easily build an zip/city look-up service on your own.</p><p>Nevertheless, the perfect solution would be that the address is filled out for you automatically. Besides Chrome’s form fill, an increasingly number of merchants have integrated <a href="https://developers.google.com/maps/documentation/javascript/examples/places-autocomplete-addressform">Google Address Auto-complete</a> on their websites. Out of all the sites that provides address auto-complete, I believe Staples.com has the most ‘delightful experience:</p><ul><li>Initially, only the ‘Shipping Address’ is shown to the buyer, making this form much less intimidating.</li><li>When the buyer enters his address, auto-complete suggests the address candidates that he/she can choose from.</li><li>Once the buyer selects an address, the form expands and reveals the specific street/city/state/zip fields. The user can still manually edit individual fields.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/1*IXZYmTGcs7uUf85PZgATNg.gif" /><figcaption>Staple uses Google API to auto-complete the address form</figcaption></figure><p>Interestingly, Google Maps has a virtual monopoly over address auto-complete API. I wonder why the other map API has no similar offering…</p><h4>Finally thoughts</h4><p>Although the mobile commerce is gaining a lot of attention , make no mistake — people are still buying from a desktop browser. People might browse the items or add it to their carts on their mobile phone. The final checkout step, however, is most likely happening on their laptop. In that case, offering a smooth, delightful checkout experience on desktop is the least a e-commerce merchant can do to increase its sale.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/700/0*4e-MDhClSO9glU8y.png" /><figcaption>Reference: <a href="http://www.smartinsights.com/ecommerce/ecommerce-analytics/ecommerce-conversion-rates/">http://www.smartinsights.com/ecommerce/ecommerce-analytics/ecommerce-conversion-rates/</a></figcaption></figure><p>In particular, there is a real opportunity for merchants that sell <a href="https://stratechery.com/2016/surface-studio-nintendo-switch-and-the-potential-of-niche/">niche items or services</a> to retain their buyers who might visit the sites occasionally, or when a specific needs arise. <a href="https://www.psychologytoday.com/blog/am-i-right/201302/the-power-first-impressions">First impression matters</a>. When your customers think about buying shoes, you don’t want them to be reminded of their painful experience about correcting out their credit cards and address info on your websites (and I’m looking at <a href="http://www.zappos.com/">you</a>).</p><p><strong><em>Note: </em></strong><em>It was fun writing </em><a href="https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-i-payment-method-collection-bf9a52ac6d1d#.uujvgcv9q"><em>Part I</em></a><em>, and in particular I was encouraged by message from old and new friends. Indeed, Medium provides a great way for sharing knowledge and ideas, and each of us, no matter how insignificant we think our job or work is, possess knowledge and thoughts that we can offer through our words. And who knows, maybe someone on the Internet, somewhere, some day, might find what you write valuable!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=3229b80bc709" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[E-commerce Checkout Design Deep Dive (Part I) — Credit Card Collection]]></title>
            <link>https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-i-payment-method-collection-bf9a52ac6d1d?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/bf9a52ac6d1d</guid>
            <category><![CDATA[ecommerce]]></category>
            <category><![CDATA[credit-cards]]></category>
            <category><![CDATA[payments]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Mon, 20 Feb 2017 23:11:28 GMT</pubDate>
            <atom:updated>2017-02-28T06:01:27.149Z</atom:updated>
            <content:encoded><![CDATA[<p>(Updated: 2/26/2017, and <a href="https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-ii-address-collection-3229b80bc709#.x9t869of0">Part II</a> is now available)</p><p>In online shopping, the most critical part of the process is likely to be the final checkout steps. It actually take a lot of trust (and typing) when we give our payment and address information to a third party merchant website, and then cross our fingers hoping that we get charged for exactly what we paid for, and the goods will be delivered to our home correctly, not getting shipped to my next door neighbor or a house that has the same door number but is two blocks away.</p><p>Now, step into the shoes an e-commerce site owner for a second, and imagine how you would build a site that steers the buyer to click that “Confirm” button as often as possible?</p><p>If you have read the book <a href="https://www.amazon.com/Thinking-Fast-Slow-Daniel-Kahneman/dp/0374533555">Thinking Fast and Slow</a>, one of the major concept introduced is called <em>Cognitive Ease. Cognitive ease </em>refers to the psychological ease in making decisions, when we see something that we can related to, feel good, or familiar with already. In the e-commerce world, it means that when you build your UI, the experience should be largely the same as the other popular sites (Amazon, Target, Walmart) where he/she might shop at.</p><p>If luring a buyer to give his payment and shipping information is the <em>most important </em>job of an e-commerce site, then the merchant has every interest in inducing <em>cognitive ease </em>in the checkout flow , by making the experience as smooth and familiar to the buyer as possible.</p><p>I have been doing some works on UI layout design recently, and so I thought it would be interesting to survey different e-commerce sites and see how they structure their payment / address page, and share what I have found in this process.</p><p>In this part I, we are going to start with the credit card collection from design. In <a href="https://medium.com/@isaacy/e-commerce-checkout-design-deep-dive-part-ii-address-collection-3229b80bc709#.x9t869of0">part II,</a> we will move on to address collection form design.</p><h3>Research Hypothesis and Questions</h3><p>Assume that <a href="https://www.sciencedaily.com/terms/exposure_effect.htm"><em>familiarity breeds liking</em></a><em>, </em>you’d assume that most of the <em>payment method </em>pages from these top merchants would have similar look and feel or layout. After all, the required information are standard for e-commerce, and the form factors (desktop mostly) are uniform.</p><p>For example, we’d expect:</p><ol><li>A payment form structures its layout more or less inline with how a customer looks at his physical credit card. he form always starts with the card number, then the user enters the expiry date and the name.</li><li>A form collects only what it requires to complete the transaction. For example, since names aren’t always required, it will not collect the cardholder’s name, or the billing address.</li><li>It uses common terms that can apply to different card types. For example, Visa/MasterCard/Amex/Discover all use different terms to refer to the 3-digits printed at the back of their cards (for Amex it is the 4-digits at the front). <em>Security code</em> seems to be a more generic and well-recognized term than other choices.</li></ol><p>With this assumptions I formulated a set of survey questions:</p><ul><li>Does the form collect cardholder name?</li><li>Does it auto detect the card type (Visa/MasterCard/Amex/Discover) based on the number that the buyer entered, and show the appropriate help for entering<em> security code</em>?</li><li>What is the first field to enter? Name or Card Number?</li><li>If the name is needed, how does the form position the cardholder name field?</li><li>Does it refer to the 3-digits at the back of a card as <em>security code, </em>or CVV / CID?</li><li>Does it collect the billing address of the card?</li></ul><h3>Source of researches</h3><p>To address my questions, I surveyed 27 different e-commerce sites across the web. Since the majority of the e-commerce still happens over the desktop, I focused on desktop browser only.</p><p>Now, let’s look at the patterns from that individual sites.</p><h3>Survey Results</h3><p>In our table below, we highlight the results with different colors that express a common sense judgment on the feature’s usability:</p><ul><li>Green label denotes the feature that helps improve usability because it reduces the number of fields entered, or has better clarity</li><li>Orange label means that we are not sure if this feature helps or degrade usability</li><li>Red means that this feature hurts usability because it is more ambiguous, or make the flow more cluttered</li></ul><p>Here are the results:</p><h4>Baymard Ranking ranked E-commerce sites</h4><p>The first set of sites comes from <a href="https://baymard.com/checkout-usability/benchmark/top-100">Baymard Institute’s Checkout Usability Rankings</a>. Out of the 50 top ranked sites, I picked out 7 sites from well-known merchants that sell physical goods, and are highly ranked: <strong>B&amp;H Photo</strong> (Rank #1), <strong>Home Depot </strong>(Rank #2), <strong>Staple</strong> (Ranked #3), <strong>Office Depot </strong>(Rank #5), <strong>Walmart</strong> (Rank #5), <strong>Target</strong> (Rank #8), <strong>Best Buy</strong> (Rank #9) and <strong>Nike</strong> (Rank #20). I expect that most of these sites would set a benchmark for good layout design patterns, with Nike being sort of like a benchmark for an average merchant.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2qMCwBuMs1-bTO2EL4KXOg.png" /><figcaption>(*) — Requires explicit card type selection</figcaption></figure><h4>Credit card forms on several Internet giants and payment processor</h4><p>The second set of sets come from several internet giants (Google/Apple) and payment processor (PayPal, Stripe). Note: I deliberately skipped Amazon in this round.</p><p>The surprising thing I found though:</p><ul><li>Ebay has two completely different layout for managing authenticated and guest users. However, the good practices in one flow aren’t carried to other flow and vice versa.</li><li>PayPal’s payment form is one of the more elegantly designed form. But the split between Ebay and PayPal means that these best practices aren’t adopted in Ebay. It would be interesting to see if the payment method activation rate has changed, before and after the split.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uv95oO1ZK0DMW3GMI1dmnA.png" /></figure><h4>Credit card forms on several hospitality or service-based merchants</h4><p>The last set of sites are from 10 other established e-commerce merchants that sell services or digital good. The reason for including these sites in the survey is to use them as ‘control’, and reduce the possible bias for certain industry (merchants that sell physical goods).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qnkTZVN4cnRDLGsgW2BwAw.png" /></figure><h3>Observation on common practices</h3><h4>Treatment in card number and cardholder name</h4><ul><li>Almost all the <em>sites collect cardholder number </em>as the first field on the form</li><li>However, · &gt;50% of these sites don’t collect cardholder name at all!</li><li>&gt;50% of these sites allow or auto-format the space between card numbers (e.g. “4242 4242 4242 4242”). Researches show <a href="https://baymard.com/blog/credit-card-field-auto-format-spaces">23% of buyers type their card numbers in distinct 4-digits blocks</a> .</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/314/1*6pz6xkK48dUTfMTUnN4q5g.png" /></figure><ul><li>If the buyer needs to enter the cardholder name, the field is usually placed below expiry date and security code fields</li></ul><h4>Card type detection</h4><ul><li>Most sites support <strong>auto-detection of the card type</strong> as the user enters the card, and shows the appropriate logo</li><li>However, some sites have neat treatments around helping buyer entering the security code by auto-detecting card type and shows an appropriate image hint. For example, if an Amex card number is detected, the form shows the front of an Amex card, and highlight the CID position (see below)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/256/1*JgHWY6q14dCXsd5wB4yqMQ.png" /><figcaption>B&amp;H photo — credit card payment form</figcaption></figure><h4>Security code/cvc/cvv?</h4><ul><li>Most sites refers to CVC/CVV2 on the card as “security code”.</li></ul><h4>Billing address requirement</h4><ul><li>Most sites still require a billing address for the card.</li><li>However, there are great way to help user enters the correct address. We will talk more about that in Part II.</li></ul><h4>General layout design</h4><ul><li>Sites mostly use explicit labels, although we are observing a new trend that inline. the label in the field. Google sites (across Play Store or Payments) takes this design approach.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/379/1*F0GiLxsDvI2gzHSoRNQ_jQ.png" /></figure><ul><li>The use of animation is essential in creating a delightful experience. Out of all the sites I looked at, the Target has a nice experience where the fields are first inline, and upon typing the card number, the form then show the rest of the fields for credit card</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1023/1*gLhkl90yNy7rig3D-m3W8w.gif" /><figcaption>Target.com credit card form</figcaption></figure><p>In the next article, we will look into the address form design and find the common patterns!</p><h3>Reference</h3><ul><li>Top 50 E-Commerce Checkouts — <a href="https://baymard.com/checkout-usability/benchmark/top-100">https://baymard.com/checkout-usability/benchmark/top-100</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=bf9a52ac6d1d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Thoughts on the future of C#, VB.net, and VBA]]></title>
            <link>https://medium.com/@isaacy/thoughts-on-the-future-of-c-vb-net-and-vba-12c525d48fbb?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/12c525d48fbb</guid>
            <category><![CDATA[programming]]></category>
            <category><![CDATA[microsoft]]></category>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Sat, 04 Feb 2017 23:24:47 GMT</pubDate>
            <atom:updated>2017-02-04T23:24:47.067Z</atom:updated>
            <content:encoded><![CDATA[<p><em>I wrote this after reading the .</em><a href="https://blogs.msdn.microsoft.com/dotnet/2017/02/01/the-net-language-strategy/"><em>NET strategy</em></a><em> and </em><a href="https://arstechnica.com/information-technology/2017/02/microsofts-developer-strategy-c-for-fancy-features-visual-basic-for-beginners/"><em>Peter Bright’s analysis</em></a><em> on Ars Technica. Although I still write code in C# occasionally, I’m a program manager at work, and these opinions do not represent those of my employer.</em></p><p>I remember that when I started my internship 12 years ago, one of the first projects was to create a project tracking system using <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.NET%2F&amp;h=ATPRahJSO6AFChnSb0gexn-bSxbqyt-5xY14aG-H8UsOPNxz_QspH1L-QtS7n0HE3_9EpWl_o9gwwcbBnVVEYtQFO7Ldhm-EYwQAchCKF6U75gcjPxgwzPLFeKlEC2C2PtH56HzESDimVQ&amp;enc=AZPS9ufCO8_49tdLC3M2bNFIw2bLkSsxivBoYfG2FxAJL78nkhRvhJ--O8-7GoZW_jJYG1ijRfJyoQpxY3dVF1Th1MrkG4iLI_q2ATUa0zPVLdpEx6DayuXmNkRumMys5TP0FpmeyYq73RFTGhQaugEZhfQ71TIfwbPdn68UUY-EAcTgTZRqOUXcnO41wMhtcTQ&amp;s=1">VB.NET.</a></p><p>However, the team also had some legacy applications on Excel that were of course, on VBA, and I needed to fix bugs as well. At that time I thought, “Why would Microsoft be so annoying as to create two languages that were so similar but different and entirely incompatible!” At that time, VBA dev really got away with declare Dim as Variant type… coming from the strongly-typed world I had no idea why VBA would allow such monstrosity and but not in <a href="http://vb.net/">VB.NET</a>! Of course, we thought that <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.NET%2F&amp;h=ATPjIVVDL_LUKtADo_jCS4wG2bsTJ14TBoFv6gb-RY5HAw8AuB7r0gc0CdPhg6556Lb0i-83m8zhbdCHUCtCE5veWNtHhSb_U21zOZUyywUXqf6OepqBkV5P3Yd-CL8wWLroWYmwPbhnvg&amp;enc=AZPyigIzmMahSwcL-kVTpsfH5fsvfp2f2xwvYg3sWhNSbvCNCTvVDTKj3jHaxSKPsa8P1Vw7QJqJyMDX_jYXR_CAcWpMP5JbhaNc7aGiUq1a2022fuRGqzEuJVAkS6CPEeJMM265hjYW3VnHenHbQRM3lXHmnccDkWJ382kbhXaZVi7vMvNNVF3QRAsQPAL4Zt4&amp;s=1">VB.NET</a> was at least syntactically a better language than VBA (it is object-oriented!), and thought that one day all of the VBA programs in Excel workbooks would be auto converted to <a href="http://vb.net/">VB.NET</a> running .NET core underneath.</p><p>This never happened. <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.NET%2F&amp;h=ATPt91hav01FJU0Aq1C8VMQeXv0G9qDfbBieqI6yKC6OIGBvmf3UmDohfYFGLdj7TtfXggXMEqSIsDLgRbK8p8NM0i987VyvF7Y5B4A3MiMJCK_9foeRwWpl2vt9hrEdUVeJKdsf53vN9w&amp;enc=AZN7ANG19xYpQQqLAAxIlMMlnGD4e8LXwm3Ca9MwKohhNgGalfl82o9WKd2wQeUwy98TmoOlFv16vvqQNepvQNek3VrYsYxKQx8cMJQpIwEZvJSYML_nduwVmZo83soStlP5cpELig61jHvQAebNlTU-7-lGrAl2VOT1ExG3CvnCZAdD1xOc5tm1xriP6dPgc2Q&amp;s=1">VB.NET</a> lived on. Eventually when I started working on VSTS office add-in,one of the “hairiest” problem was to make sure that the Add-in code didn’t interfere with the VBA code that the user wrote. That was a nightmare. Still VBA lived on. It is weird that the add-in code was written in C# and not in VBA. I doubt anyone at MS at that time used <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATMglaJkP-zlorYbafPwrF5N-38FSHaG4mayWxW7f5XzGJboC3nF-RVNBvTnVbq-rHiSLIL34IoN_SD2g3dl3Jv54VWSLssAUu6DWFPQSYyI5LTAN3SNRUWQ6jpNkiy48e6768-uQ9Fd5Q&amp;enc=AZOBmp9lb8h-Ca2c5nF64eFbnE86jVT8QrMqn5fZHc0Q0e8GQWeM4VFH_MuWCKmYJQ0TxEzivwNKIPiaPBUtbtT6VhglwsMKEVv-kPt4tTJckzzlDGUc7kv3ki2bOOFl0d0sIfcwz7w-zTTzr-m0oxScPJnEKX0fZ_BxEag54KpP9exLXl1myHd0f6rVHCX-9IU&amp;s=1">VB.net</a> for really anything other than the <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATPnEgmsWWUrFpNNQiPHLrcJgH0ZD4Qz7z03_G2LJBiCt9XkF92_mp8OqgAGMZa8UO2wVknHlyWAqqK4k5JEqKcbqAlST_AVxB7ykpzjMUXzi5WDIy2BDcx1o7wvdSwnOCUGXFnnCo_J4g&amp;enc=AZMLHJ2h2Q_swUnAFH1W04fuEY5Pr8hflGk6xwdixIVqc53TPdcJAcozTlFK9AXzW3gvy3lIE4ZirrA1T2wqOGiXyxmc6nxBLxN_c8IdG5tCL1n3oLHHDmGAPxly4BdPCx0ng5fGVDj9Vg2A00iTMNa8crvKbZnfg1CN3xtYxIc7M3N1vdAqkdQ3Asbq23_g22E&amp;s=1">VB.net</a> team. Unfortunately, it was also where the efforts were invested. VBA as a language didn’t have update and was as hard to use and debug as ever, despite the fact that it was still popular.</p><p>Fast-forward almost 10 years later. I’ve not encountered anyone still using VBA at any form. I guess if the team include “small basic” into the VB family then perhaps there is some traction in education, but the language is basically incompatible with any tooling for education and certainly not in any major cirrculum. And, to be fair, learning <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATOhsUQsbDJFRvZyrHwacgQ44H5Z1FQzhoqR7jwe3DY3Ng2F-96_zldToZNljG_8U81ndMZwMnBugGR7X8K4nKPG7jcwHs4kU0b2kxaSZc_97_wVLfNd0mQCxWCyRpclipvTN455BTsXeg&amp;enc=AZN2jnrTgzaPXHkSD8LXUSi9s-29CHnrEVs51YOSzuW9orUWChnVnHZMOX3U5gduuZUBHz9rmtcn9_19QXtUMachaEvABf5XCpQWxcsdCbunUh9lqeyyOl8eDIchmwin1rFPsnYU6bI0P_ps6UEmlaQ8o_JY68msnKrBcXBVU7030EYtzZJy_A7TJXYLPOmf8IY&amp;s=1">VB.net</a> is just bad for mind, and you will incur more bad programming habits out of it. Investing in it is… allow me to put it strongly… waste of human intellect.</p><p>Why is VBA so hard to kill? I think that innately has to do with human’s attitude toward learning — we aren’t likely to learn something unless our lives depend on it! For thousand of business professional, knowing VBA = security of their jobs and there is an incentive to keep it and not replace it with anything else (or they would be out of jobs). And unlike consumer software, enterprise software takes years to be updated and №1 priority it that all the files/data needs to be backward-compatible! That ensures that VBA will stick around until, either 1. the generation of VBA users fades out from the industry, and 2. enterprise aren’t using Office software to manage their work. (2) is not good for Microsoft.</p><p>My thoughts:</p><p>C# — there has been WAYS too many features since C# 5. Really, async is like the last thing that was actually useful. All the new stuff are just add-on but actually adds burdens to the developers because some newbies would use this C# 7 stuff and no one would have any ideas what this is about in the CR. Churns in C# should stop at one point.</p><p><a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATN4I0A5amRjE7GY6JZ1-JkNxbWB-eFCS0e1XlOR1g7P4lclD0tQ1eQYYUw7-cliVhDZbPTVJtePNcfDN0oS4MTJmze5T60opdcYDa_Gi_r4cEF7UgGgU8jX837kwUvdMXgvz68MSqZ-aQ&amp;enc=AZMGQ0HTNY_aQjD5RWpgUMDAPzntB18NxlVP6S-HxKp8wUhVVqqMNa1rJWCIrdNulKjygY4LHPwzdZyTS9BzoTCLKxnzdeN1-z8OkdT2sLS9Gejes3S5wXIW8Mn5UpISkl6giNw8UM_TCfraLQS0JHt7Q2w4ZrvW3e337GwcmSevJxKFKTc0cHngZvYfBGM-SJE&amp;s=1">VB.net</a> — really, if you can make UWP/Xamarin games in <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATPumpnjx1hIAr4xDJeQIdV_F59mFdC7Ckx_8_aVkFhX2G7rHDqmEQqzEisJ8R61Ckv4Yy8m2hgwH3Tow6JppHGILQhxz9qiQTqj3JmDTGSrcJJyzzudSr1T50E-aIOpEmnOQE_bqVLwdA&amp;enc=AZOK64pD4krYDkjGK3wRWsJG7pPsiBTLCkIPaMexVHCmfSnS03QMRHQ_4rkY9pDTkFnMHwOgLqu9d7cHO0zqRK0yVCCUsW7zJfD6OxD9MR3j3_U_5VzMXWgyRbLZA0bXbWuqZaLpyxta4rUygx2eIxO2HQ8sgVcMfsR5_QPS-8EYG7o9qtmsdqdcQd6Lem6rgps&amp;s=1">VB.net</a> and there is really no point for anyone to learn <a href="https://l.facebook.com/l.php?u=http%3A%2F%2FVB.net%2F&amp;h=ATM43edXYMKtpkiw6EX6EJRTOHsdZAEA525G4wQXo8f9ZQMtFZpcPO6Qxp21NypFiy08ZpxzVG0-wMDHQBeO9_TPXiRXUvyDwgEDJtN5V6kh6hOwDkJOIg_NpNLJ9AxB_MzZyNufO2KbTw&amp;enc=AZN2qjhpAGVy-Kpum1V5woPTP5mbKjHQv8NhQGCm3vWYLNjAe-fjhRkfV-G__tRgVPvo3U7SterNVFisy1YhMjKayPZnoL1giSdfdenaxSmY38geb9jyKHiiVyh-V8uVkRxWAkv5PWuc69fsqgmdBRXuyBvOrf9eh1hVBeUMEDBhAW7Hg_7DwpDSrQTO9Cs4th0&amp;s=1">VB.net.</a> Dev wants to learn the language that is as applicable in many (popular) platform as possible. pure .NET isn’t popular out of the server world, and VB isn’t available on UWP or Xamarin. Why keep investing any energy on this at all!</p><p>VBA — it is actually the area with biggest threat and opportunity. I see a trend that as we are moving everything to the cloud, JS is slowly displacing all other languages as the defacto languages for web-based plugin. That’s all goodness except that it basically puts all these business professional at a level playing field with any new hired. 20+ years of expertise will be displaced by this cloud migration. But that’s bad news for Microsoft too since the enterprise rethinks their investments on the non-pro business software development, Microsoft is at level playing field with other competitors.</p><p>The right move, I think is to enable VBA in sharepoint and cloud-based office suite. Anyone can upload their existing workbooks to the sharepoint and Excel Online will be able to handle the VBA stored in the workbook, and allow the user to edit it and recompile it. I know that VBA is out-of-fashion in the cloud work, but technical challenge aside, many will helped by this. And it will ensure Microsoft’s grip on office software for years to come.</p><p>Note. these thoughts are solely mine and not related to my employer, nor do I have any proprietary information about the development of these languages. I’m now a PM and really out of the dev world for a long time:)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=12c525d48fbb" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Guessing Game Bot Privacy Statement]]></title>
            <link>https://medium.com/@isaacy/guessing-game-bot-privacy-statement-46a33f9116f7?source=rss-146690383241------2</link>
            <guid isPermaLink="false">https://medium.com/p/46a33f9116f7</guid>
            <dc:creator><![CDATA[Isaac Yuen]]></dc:creator>
            <pubDate>Tue, 22 Nov 2016 05:55:31 GMT</pubDate>
            <atom:updated>2016-11-22T05:55:31.495Z</atom:updated>
            <content:encoded><![CDATA[<p>This Guessing Game Bot collects no personal information from the user and all the data collected are used for analytics only.</p><p>Copyright 2016 Isaac Yuen</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=46a33f9116f7" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>