<?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 Adam D&#39;arcy on Medium]]></title>
        <description><![CDATA[Stories by Adam D&#39;arcy on Medium]]></description>
        <link>https://medium.com/@adampdarcy?source=rss-46fba4aac79e------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*WoyFpUEvZ0p2qEAV02CqnQ.jpeg</url>
            <title>Stories by Adam D&amp;#39;arcy on Medium</title>
            <link>https://medium.com/@adampdarcy?source=rss-46fba4aac79e------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Mon, 16 Nov 2026 16:02:01 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@adampdarcy/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[JTBD: Earn a new product manager superpower through clustering]]></title>
            <link>https://adampdarcy.medium.com/jtbd-earn-a-new-product-manager-superpower-through-clustering-d2bd45ccbd4c?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/d2bd45ccbd4c</guid>
            <category><![CDATA[jtbd]]></category>
            <category><![CDATA[ux-research]]></category>
            <category><![CDATA[product-marketing]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[clustering]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Tue, 20 Apr 2021 07:19:32 GMT</pubDate>
            <atom:updated>2021-09-20T06:26:21.642Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*zNtuuFr56iG7_svS" /></figure><p>Have you ever commissioned a customer needs survey only to discover everyone looks roughly the same leaving you with nowhere to focus? This is surprisingly common and I wanted to share an easy solution using a well known clustering technique that can give you a much higher degree of confidence you’re building the right thing.</p><p>In Competing with Luck, Clayton Christensen gives two excellent insights:</p><ul><li><strong>Myth of the average customer:</strong> The US Air Force <a href="https://www.thestar.com/news/insight/2016/01/16/when-us-air-force-discovered-the-flaw-of-averages.html">in the 1940s</a> blamed pilot errors after 17 crashes in one day. Turns out the flaw was the 1926 seat designed around the ‘average pilot’ that they subsequently realised didn’t actually exist overturning a 100 year assumption.</li><li><strong>Measure the job, not the customer:</strong> Gives a beautiful analogy of how companies analysing the customer vs the job get measurement errors that are narrated away similar to how for 18 centuries we narrated away the errors when measuring the planets as moving around the Earth vs the sun</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/869/1*5jlf0sgXFI2aY-WJgWmHDw.png" /><figcaption>Many <a href="http://www.pas.rochester.edu/~blackman/ast104/aristotle8.html">complex calculations</a> were invented to preserve the Aristotelian view of the world</figcaption></figure><p>I’ve personally commissioned six large surveys in the last four years embracing a Jobs view of the world, but each time I’ve been disappointed. Common replies are ‘there’s clearly no problem to solve’ or ‘customers tend to answer around the middle when given many Likert scale questions’. I called up a few product friends in different companies and got similar views with one yesterday at a multi-billion dollar US Fintech saying:</p><blockquote>“Quant research is just a tickbox for me now - we have to do it but it doesn’t add value so I go with my gut or use the customer interviews”</blockquote><p><strong>Clustering and the Correlation vs Causation mistake</strong></p><p>After reading JTBD pioneer Tony Ulwick’s <a href="https://jobs-to-be-done.com/jobs-to-be-done-examples-creating-the-bosch-cs20-circular-saw-d70022e08815">blog</a> on this exact problem, he suggests <strong>clustering</strong> as a way to find hidden segments with common needs but doesn’t detail how. I remember learning this a few years back on a data science course but never had the time to practise.</p><p>Clustering is a way of [unsupervised] learning more about lots of data points by finding meaningful groupings that are too hidden to specify beforehand [unlabelled]. This is in contrast to supervised learning where we know the groupings upfront and simply want to train a model to classify future data points e.g. is this a cat or a dog (see a <a href="https://developers.google.com/machine-learning/clustering/overview">great explanation</a> of both from those clever fellows at Google).</p><p>I’ve asked research teams in banks and tech companies I’ve worked in to try clustering. One team didn’t know how, another didn’t have the right resource free (data scientist — precious commodity) and another had segments already defined from previous exercise so mapped the responses to them (definitely more useful, but still felt imprecise). Instead we tend to get correlation based pivots like “Females in mid 40s SEC AB have these needs in common’. The fact someone is female and 40 usually doesn’t give me much to go on in terms of causal understanding of why they would hire our solution — however surprisingly that’s how many people build and market products.</p><p>Then last week whilst on paternity leave, the founders of <a href="https://www.gini.co/"><strong>gini</strong></a> — a fintech I advise — asked for help in applying JTBD to decide what solution to build for a cash flow forecasting product and guess what..</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Tm7BOtkOUgiImijl_mtz2w.png" /><figcaption>I’ve had 4 messages like this in the last 6 months alone</figcaption></figure><p>Finally some time (between diaper changing) to try this out myself and 48 hours later I was able to present three clusters of potential customer each with very unique needs along with associated correlation based insights that were just fascinating and gave their product design team and marketing the rich story they were looking for to build and market their new product with high confidence.</p><h3><strong>4 simple steps to insight</strong></h3><p>Here’s how to do it in four easy steps using a unsupervised machine learning algorithm called <em>k</em>-means in Python (full code <a href="https://github.com/adampdarcy/jtbd_kmeans_clustering"><strong>here</strong></a> and please don’t be put off if you don’t know programming, all instructions are in the <a href="https://github.com/adampdarcy/jtbd_kmeans_clustering/blob/main/README.md">readme</a> file):</p><ol><li>Organise raw survey data</li><li>Find the right number of clusters</li><li>Cluster the respondents</li><li>Analyse and present your results</li></ol><h4><strong>1. Organise raw survey data</strong></h4><p>After many profile questions, we asked respondents to rank job outcome statements by importance to them. A typical question looks like this and note how the mean rank are all too close to each other to be informative:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wBzYyC50cTgvz7HhO8So1A.png" /><figcaption>Respondent ranks each statement with 1 being most important, 5 being least</figcaption></figure><p>We exported the results into CSV format available from most tools (we used <a href="https://www.pollfish.com/">Pollfish</a>). You want each row to be a respondent and each column to be the answer they chose.</p><p>Look at the responses and choose the ones that are needs based. Ignore profile questions like age etc as mentioned they are likely correlations not causation factors.</p><p>We decided to choose only the <strong>most important </strong>outcome to cluster around ignoring the others so we gave a 1 for top rated outcome and a 0 for the others <em>(note: in k-means you need continuous numbers for answers that are comparable like Likert scale i.e. not categorical ones like Dog or Cat.. there are ways to do that but beyond scope of this blog)</em>.</p><p>Below is how the data looked right before doing the clustering:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Q1OpX-bD0Mvj5mQ31JudQQ.png" /><figcaption>Notice each respondent has only one ‘1’ for each question denoting their most important outcome</figcaption></figure><p>Now read this file into a variable called training, then pull out just the JTBD outcome answers (in my CSV it’s column 49 to 96):</p><pre>training = pd.read_csv(&quot;jtbd_survey.csv&quot;)<br>jtbd_needs_only = training.iloc[:,49:96]</pre><h4><strong>2. Find the right number of clusters</strong></h4><p>While we want to find clusters, it’s not useful to have lots. So how many? Luckily <em>k</em>-means tells you. Simply run it several times with different K’s and compare the <a href="https://www.geeksforgeeks.org/elbow-method-for-optimal-value-of-k-in-kmeans/">data distortions</a> (distance each respondent is from each other on average in K clusters). Once the distortion stops significantly decreasing, you can stop. Let’s try 1 to 5 clusters to see which works best for our dataset:</p><pre>distortions = []<br>K_to_try = range(1, 6)<br><br>for i in K_to_try:<br>    model = KMeans(<br>        n_clusters=i,<br>        init=&#39;k-means++&#39;,<br>        n_jobs=-1,<br>        random_state=1).fit(X_pca)<br>    distortions.append(model.inertia_)</pre><p>Once done, you can compare the distortions list in a chart and simply look for the ‘elbow’ in the line. Three feels good as after that there is little loss.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zMkkGKLQW1xeVoJeN8HhCQ.png" /></figure><h4><strong>3. Cluster the respondents</strong></h4><p>The way <em>k</em>-means works is it plots every respondent as a dot in a geometric space. So if we had only two Likert scale questions and 500 people, you would have a two axis chart like below:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_dzMi5KGb-eScavhnyi8Gw.png" /></figure><p>As our chosen K = 3, the algorithm picks three random respondents to be Centroids (the blue dots above) and assigns every respondent to the Centroid they’re closest to. Each Centroid calculates the mean of all its new members’ positions and moves itself to that central position. The variance score between the three Centroids is recorded for later (low variance means the clusters are fairly equal in shape). It then shifts the Centroids to new random positions, repeats and again records the new variance. After many iterations it looks at the variance scores and picks the lowest one. That is our chosen clustering where our respondents’ needs are most neatly related to each other.</p><p>Now we had 42 questions so we actually run this in not 2 but 42 dimensions — I wanted to try and draw that but worried I may end up in therapy.</p><pre>model_k = KMeans(<br>    n_clusters=3, #we put our chosen K here<br>    init=&#39;k-means++&#39;, #++ chooses more optimal starting Centroids<br>    n_jobs=-1,<br>    random_state=1)<br>model_k.fit(jtbd_needs_only)<br>result=model_k.predict(jtbd_needs_only)</pre><p>Result now contains a two column list with respondent IDs and their new ‘Cluster’. Let’s add this to the original dataset and export it to CSV:</p><pre>result = pd.concat([training, pd.DataFrame(y_final, columns=[&#39;Cluster&#39;])], axis=1)<em><br></em>result.to_csv(&#39;clustered_survey.csv&#39;)</pre><p>We can now inspect our new <em>clustered_survey.csv </em>to check the cluster column:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/666/1*2kLrkyXWM7mpcOSucuIjZw.png" /><figcaption>See 0–2 cluster assign to each respondent</figcaption></figure><h4><strong>4. Get rich insights from the clusters</strong></h4><p>Now you can inspect each cluster and pull out what is different about them i.e. what is the significant variance between this cluster and the other two?</p><p>We found some fascinating insights using this method and the team are now busy selecting the top needs to tackle on their immediate roadmap for one cluster that stood out in particular.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-mWLRO5yO2oS9oiq5R6nbw.png" /><figcaption>Final output telling a rich story which made complete sense when thinking back to interviews</figcaption></figure><p>Things we focused on for this analysis were:</p><ul><li>What <strong>important needs</strong> did a cluster have significantly more of vs others?</li><li>What did a cluster pick significantly more as their <strong>least important need</strong>?</li><li>What <strong>demographic and profile questions</strong> were significantly different?</li><li>Did they run some jobs <strong>more or less frequently</strong> than the other two?</li></ul><p>Whilst I can’t share the exact insights due to confidentiality, I can share a few high level teasers below to give you a flavour:</p><ul><li>One country showed a HUGE difference in perceived skill level &amp; tooling</li><li>A different cluster were calling out for very specific types of help</li><li>One cluster with certain needs were in a very specific type of company at a very specific stage</li></ul><p>Needless to say this survey has given the team so much more insight about what to build now vs the usual averaging / user segmentation approach.</p><p>Hope you find it useful and feel free to ping as usual on <a href="https://www.linkedin.com/in/adamdarcy/">LinkedIn</a> if you have any questions or need help. I’m now going back to all my old surveys and running this to see what we missed :)</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=d2bd45ccbd4c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[JTBD: The Ultimate Social Job to be Done]]></title>
            <link>https://adampdarcy.medium.com/jtbd-the-ultimate-social-job-to-be-done-1d22b6607e36?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/1d22b6607e36</guid>
            <category><![CDATA[behavioural-psychology]]></category>
            <category><![CDATA[jtbd]]></category>
            <category><![CDATA[social]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[anthropology]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Sun, 29 Mar 2020 07:14:39 GMT</pubDate>
            <atom:updated>2020-06-21T05:07:57.328Z</atom:updated>
            <content:encoded><![CDATA[<p>In the movie Castaway, Tom Hank’s character gets stranded on an uninhabited island. In his lonely and mentally desperate state, he rather strangely hired a volleyball he named Wilson to get him through his ordeal. What was the job he was hiring the ball to do?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*IwBntQ0_mxgykA65.jpg" /></figure><p>As part of my series on Jobs to be Done I’d like to explain what I believe is THE ultimate <strong>social job</strong> and show from experience how powerful it can be if we apply it to the functional jobs we build our products for.</p><h3><strong>The JOB of social</strong></h3><p>In 2012, I witnessed something I found extraordinary; this led me on a journey of self discovery and completely transformed my career as a product manager.</p><p>I was working in Indonesia for Blackberry Messenger — their largest market, which peaked at 90 million MAU before being taken down by WhatsApp. Everyday between 12am and 6am there was almost no one online. Then at 6am (sunrise in the tropics) their largest spike of the day appeared. Tens of millions of Indonesians had woken up and immediately connected to their social network. What was the underlying force that made this behaviour so commonplace?</p><p>After some searching, I quickly came across the huge volume of work by the brilliant evolutionary psychologist Professor <a href="https://www.psy.ox.ac.uk/team/robin-dunbar">Robin Dunbar</a> of Oxford University. What quickly became apparent was that if you want to understand the social behaviour of humans, you first have to understand that of primates. I have summarised this below, before I reveal what I believe to be the ultimate <strong>social job.</strong></p><h4><strong>Why do we need friends?</strong></h4><p>It’s actually a survival strategy we inherited from primates. A group acts as a deterrent for predators, including those in their own species. However, this comes with internal challenges — as we all know it’s stressful living in groups and subject to free riders taking without giving back.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/384/1*peXSOS_k1cYpaWdTLS-l4A.png" /></figure><p>In fact, Harvard researchers were very surprised to find that the number one health and happiness indicator for humans is not their cholesterol levels or wealth, but their social network. The researchers groundbreaking <a href="https://news.harvard.edu/gazette/story/2017/04/over-nearly-80-years-harvard-study-has-been-showing-how-to-live-a-healthy-and-happy-life/">80-year study</a> discovered that the people most satisfied in their relationships at age 50, were the healthiest at 80.</p><h4><strong>What are friendships?</strong></h4><p>Friendships are implicit social contracts that intend to share the costs of survival and reproduction. Sounds weird? Well of course that’s not how we approach it in day to day life. We are simply behaving most of the time.</p><h4><strong>How do we make friends?</strong></h4><p>Primates use reciprocal grooming to create these bonds. A light touch of the skin releases relaxing endorphins — the brain’s natural pain killer. Next time you see any monkeys or apes, just watch for a few minutes and you will likely see it in action, with some species spending up to 25% of their waking day doing just this to maintain their group stability.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/540/1*SpzyGPri1jssIfkUxL0E_g.png" /></figure><p>Humans also create bonds using endorphin hits and spend around 20% of the day doing this, not by stroking each other naturally, but through innovative solutions. More on this later.</p><h4><strong>How many friends can I have?</strong></h4><p>One of Dunbar’s biggest<a href="https://pdfs.semanticscholar.org/a321/9cdea1120d20dda18d5b3d6f640f91e295c9.pdf?_ga=2.235281161.901566436.1584328426-2075167071.1584328426"> discoveries</a> was the linear relationship between brain size and maximum number of friends primates and humans can maintain. Humans have the largest brains because the more time spent out of the trees in dangerous open land, the more friends they need for protection.</p><p>It turns out humans can have around 150 friends at any one time:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/948/1*z-4ido0kG97QP0fwanUngA.png" /></figure><p>Seems low? Well you simply don’t have enough capacity to manage more as each new friend requires an exponential increase in brain memory and processing as we remember our relationship with them, plus their relationship with all our other friends — that’s 11,000+ relationships. For comparison, chimpanzees — our closest living relative — reach maximum group sizes of around 50 which is a mere 1,200 relationships.</p><p>Here is a good test to see if someone is in your 150. You see them sitting at the other side of the airport lounge. You have 30 mins before your flight. Would you go say Hi! and have a conversation with that person? You know their name, but wait… what was their child’s name again? Hang on… do they even have children?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/400/1*u38nQmoR4g743l5jdfsvrg.png" /></figure><h4><strong>How do we organise our friendships?</strong></h4><p>We seem to fit our groups neatly into concentric circles with those closest in the centre. Moving from any inner circle to its neighbouring outer circle, the cumulative total increases by 3x (no one knows why). You also get associated exponential reductions in contact and time spent (see chart).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*lpT_GK8P5bl-CuhvqcXAeQ.png" /></figure><p>A chimp brain allows three circles of five, fifteen, fifty cumulative, whereas humans have added a fourth circle taking it to 150.</p><ul><li><strong>Support Clique (5): </strong>The inner circle are your five core friends — two are usually family members. 40% of your social energy goes here. Write down on a piece of paper five people you would call if you got chest pains in the middle of the night or if you went bankrupt.</li><li><strong>Sympathy Group (15): </strong>Next ten friends who we see at least monthly and would be absolutely devastated if one of them passed away. So far these 15 are very likely to know each other.</li><li><strong>Affinity Group (50):</strong> Next 35 may not be that close, but can be useful sources of information. More likely to see friends here than extended family as the latter need less maintenance so in next layer.</li><li><strong>Active Network (150):</strong> Next 100 we only see a few times a year. We remember enough about them and their links to other friends and act as social capital.</li></ul><p>The 500 circle we can relabel from friends to acquaintances (feature request for Facebook perhaps?). The final 1,500 are those where we can put a name to a face, although we remember little about them anymore and perhaps never really did.</p><p>Good sales people and politicians will hack this circle by getting intel before meetings to make the person feel closer than they really are and there are some great products like LinkedIn that can help. However, this is a temporary solution as you would have to remove someone to genuinely connect.</p><h4><strong>Are family and friends the same?</strong></h4><p>Pretty much, although there is a lower cost to maintain kin as we’re genetically hardwired to be more altruistic to kin in proportion to the amount of genes we share. Therefore most friends would appear in the first 50 due to needing more maintenance. This 1950s<a href="https://quoteinvestigator.com/2016/05/05/brothers/"> quote</a> from the great J.B.S.Haldane sums it up pretty nicely:</p><blockquote>I would gladly give up my life for two brothers or eight cousins.</blockquote><p>Having lived in the UK, China, Singapore, Hong Kong and Indonesia I can see clearly that Asian cultures have way more kin than European. This also means however, that they will have less friends as you hit the same time and brain processing constraints mentioned above. There is no human culture that has a name for a relative outside of a 150 size family tree and global<a href="https://www.academia.edu/13081955/Relationships_and_the_social_brain_Integrating_psychological_and_evolutionary_perspectives"> averages </a>seem to give around 50% of slots to family.</p><p>One other quirk is that a romantic partner or spouse takes up 2 spaces in the inner 5 therefore removing 2 friends as they are rarely are already in there. I’m sure we all know someone we quickly lost close contact with the moment they met someone special. Now we can forgive perhaps :)</p><p><strong>What innovations do we use to manage the job of social?</strong></p><p>Friendships are loosely a function of <strong>quality time spent</strong> and <strong>frequency of interaction</strong>. How we spend that time varies by gender with females generally needing much more frequent contact with inner circles — even if via phone or text — vs males who need infrequent contact but high quality offline activities.</p><p>We only use primate style touch for very special relationships. However, as mentioned earlier, humans invented clever solutions to scale the job of social:</p><ul><li><strong>Laughter</strong> was one of the first most ancient methods. For example, tell a joke and three people will laugh including the teller — who amusingly actually laughs on average 50% more. That’s a three times increase in scale vs touch. In a<a href="https://www.psychologytoday.com/us/articles/200011/the-science-laughter"> study</a> by Robert Provine, people watching a stand-up comedy with a friend vs alone, laughed a whopping 30 times more. Again some gender difference here with women laughing 126% more than men and 62% more likely to seek<a href="https://dictionary.cambridge.org/dictionary/english/gsoh"> gsoh</a> in dating profiles vs men who were more likely to offer it. Like yawning, laughter is also contagious so seems to be a very ancient thing that is hardwired into all of us.</li><li><strong>Synchronous Music and Dance </strong>such as drumming to the same beat, singing or dancing to the same song gives twice the increase vs doing it alone.</li><li><strong>Language </strong>gives us a four times the increase as you can hold a conversation with four people before it naturally splits into two groups (watch.. happens every time). Storytelling and gossip are particularly strong forms of endorphin releasing conversation — both for the teller and listener. In fact Dunbar believes the campfire story was where we first began to evoke the spiritual forces that bound us together into larger tribal communities.</li><li><strong>Religion </strong>is believed by Dunbar to be linked directly to evolution of our bigger brains and was the ultimate tying together of all the above to allow us to live harmoniously in our mega cities side by side.</li><li><strong>Printing presses</strong> were a huge product innovation over 500 years ago, as they enabled the reorganisation of the entire World into the nation states we see today. They facilitated the democratisation of important social information like corrupt politicians and dodgy CEOs and created a new kind of stability — not without a few World Wars in between sadly.</li><li>The Internet age saw the invention of <strong>feed-based </strong>social networks like Twitter/Facebook and <strong>private chat</strong> such as WhatsApp/WeChat. This allowed information to be spread globally to more people than ever before, and we are now experiencing another wave if tectonic disruptions and shifting of power lines like we saw with the printing press.</li></ul><p>It is also important to mention that<a href="https://tinyurl.com/dunbar-bandwidth"> <strong>bandwidth</strong></a><strong> </strong>matters for higher quality social engagement. Emojis give higher endorphin hits and more laughter than plain old SMS, high definition photos and video literally transfers megabytes of social information leading to users feeling almost as close as with offline chat; Snapchat have even tried to better it using AI filters inducing more and more laughter although I haven’t seen the data for that yet.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xR7ZzRUfoeZHFR4m3ZbjRw.png" /></figure><h4><strong>What superpower do humans possess to allow social processing?</strong></h4><p>It is important to mention a rather simple brain mechanic that makes all of this social complexity possible for us to comprehend.</p><p>Homo Sapiens are unique in that they are the only species to have developed 5th order intentionality (see<a href="https://en.wikipedia.org/wiki/Theory_of_mind"> Theory of Mind</a>). This allows powerful shared belief systems to be created like religions and the brands we know and love today. I’ll use religion to explain these 5 orders as it is believed to be the ultimate evolutionary solution to get us to the 21st century, but you can replace<em> God/act with compassion</em> with <em>Apple/express my creativity </em>to get powerful <a href="https://www.researchgate.net/publication/48927340_Children&#39;s_Brand_Symbolism_Understanding_Links_to_Theory_of_Mind_and_Executive_Functioning">brands</a> too:</p><ol><li><strong>I</strong> believe I’m Adam (I look in mirror and know it’s me looking back)</li><li><strong>I</strong> believe <strong>God</strong> wants me to act with compassion (Belief in two minds)</li><li><strong>I</strong> believe <strong>God</strong> wants<strong> us</strong> to act with compassion (personal religion)</li><li><strong>I</strong> believe <strong>you </strong>believe <strong>God </strong>wants <strong>us </strong>to act with compassion (social religion, but although you believe, I may not believe so can still be deceitful and disrupt the group)</li><li><strong>I</strong> believe <strong>you </strong>believe <strong>we</strong> believe <strong>God </strong>wants <strong>us </strong>to act with compassion (communal religion where it follows <strong>we</strong> both believe)</li></ol><p>Only Homo Sapiens developed that last level. Our 500k year old cousins the Neanderthal got to 4 and they didn’t make it. Our closest living relative the Chimpanzee only has two levels. It is that fifth level that gives us the stick i.e. if we both know each other believes in the power of the almighty, then we will both turn up for rituals, not steal from each other and ultimately create stable communities.</p><h3><strong>So what is the Ultimate Social Job?</strong></h3><p>Tom Hank’s character in the movie was in fact hiring the volleyball product to be his friend so as to maintain a much needed social connection in order to survive and keep his sanity intact.</p><p>My hypothesis is that this — along with the above human innovations — were simply solutions to the same social job, that of <strong>keeping our social groups stable</strong> as we grew our species via self organised family and friendship groups.</p><blockquote>Ultimate Social Job: Keep my social groups stable</blockquote><p>Now if, like me, you carry this social job in your back pocket when you solve your core functional jobs e.g. ‘send money from a to b’, you will discover more and more that humans rarely do things in isolation. You will then begin to make your product much more powerful and in some cases even go viral : this is something I’ve fortunately managed to do three times in my career so far and it really is the most satisfying of experiences.</p><h3>How to use in your products</h3><p>I’ve been accumulating these insights for the last eight years and as I’ve used them to create a few hit social products, I thought it would be helpful to share a few highlights:</p><h4><a href="http://indokasih.com/">Indokasih.com</a> — Charity Crowdfunding Startup</h4><p>In 2013, marathons became popular in Indonesia and people were posting selfies with finisher medals on social media. If selfies are about projecting status to our networks, I questioned whether we could use this energy to get them to run for charity, raising money from friends for shared causes? Turns out we could; we converted their energy into cash that local &amp; international charities then converted into schools for disadvantaged kids, national disaster aid and environmental protection efforts.</p><p>One interesting social experiment raised an issue in that when someone pledged to a friend, only 20% actually followed up with payment even though we emailed them every three days for 21 days. We had the idea to cc the friend who was running and it immediately increased to 80%. This was a clear example of how a <strong>social job </strong>needed stability and be perceived as following up on obligations to friends, otherwise you could lose those friendships. It also taught me that most of the time we all wear ‘masks’ on the Internet, so building products that make us visible encourage our best behaviours.</p><h4>PayMe from HSBC</h4><p>This was the biggest success so far directly applying the job of ensuring social group stability to the functional job of paying and requesting money from friends. It even surprised me when we saw 75%+ of all Hong Kong millennials habitually use this within the first two years of launch owning over 70% of the p2p payments market. This is whilst competing with the formidable tech giants Ant Financial’s AliPay and Tencent’s WeChat.</p><p>There are many social jobs at play when it comes to money as it really has the power to create huge group instability — religions even made the effort to codify expected behaviour.</p><p>One interesting point to share was related to emojis. When someone asked for money back, 90% of the time they would use only emojis in the message field instead of text. Why? Because emojis are less difficult to be misconstrued as you’re operating at a deeper level than language. To test this I once successfully conversed with a non-English speaking Japanese man via LINE messenger for ten minutes using only emojis. A simple Sushi emoji said so much more than “Give me the money you owe me”. It says “remember that awesome Sushi we ate together last night?” evoking memories of the quality time we spent together eating good food and not of the unsettled debt (<a href="https://www.researchgate.net/publication/314665691_Breaking_Bread_the_Functions_of_Social_Eating/fulltext/58c42f05aca272e36dd6f210/Breaking-Bread-the-Functions-of-Social-Eating.pdf">eating socially</a> is also another huge solution to social stability).</p><p>As discussed above, laughter, music, storytelling all play huge parts in endorphin release between friends; our products should try to evoke these into memory as much as we can. Make people feel good and they will come back for more.</p><h4>Gojek (Mapan)</h4><p>My team and I are currently applying these insights at South East Asian super app Gojek. They own a subsidiary called Mapan that focuses on fighting poverty. There are hundreds of millions of people in Asia who still don’t have access to a bank account, smartphone or the Internet. Even low-income families with access often pay more for products and services than those in the city, often for an inferior service. An injury, such as a broken leg, can bankrupt a family who have no insurance cover.</p><p>To counter this we are building the largest social network of tech savvy female entrepreneurs on the planet to help reach the bottom of the pyramid with services on par with those in the cities. Each of our part time entrepreneurs/influencers reaches on average eight rural/inner city families, providing us with the scale and reach to ensure we can utilise technology to build a sustainable business.</p><p>Here, the social learnings around how rural communities organise themselves and how much trust is placed between people vs formal institutions is just fascinating. Sadly though inequality is so very visible and we are extending our team of smart technical people to enable us to try to build an alternative model based on equal access for all.</p><p>I can’t share too much more on this one, however if anyone wants to come and help us and is up for a big career move to Jakarta then please ping me on LinkedIn right away.</p><h3>Summary</h3><p>It was a struggle to pack eight years of research and experience into a single blog post, however in sharing this I really hope I can help my fellow product researchers, designers, managers, engineers and marketers to create products that are more human and less transactional — especially in the financial services industry, which, in my opinion, is just plain broken.</p><p>Feel free to comment and please reach out if you have more questions.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1d22b6607e36" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[JTBD: Outcome based Product Roadmapping]]></title>
            <link>https://adampdarcy.medium.com/jtbd-outcome-based-product-roadmapping-b6e3438b7f0c?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/b6e3438b7f0c</guid>
            <category><![CDATA[product-strategy]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[product-marketing]]></category>
            <category><![CDATA[product-roadmap]]></category>
            <category><![CDATA[outcome-driven-innovation]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Wed, 12 Feb 2020 14:11:15 GMT</pubDate>
            <atom:updated>2020-02-24T11:13:35.193Z</atom:updated>
            <content:encoded><![CDATA[<p>Have you ever experienced this question before:</p><blockquote>Hey Product Manager.. got a tasty new assignment for you. Oh and I need a suggested MVP and 1 year roadmap by Monday!</blockquote><p>Ok, so the question is usually a little more polite than this, but it’s genuinely something you’ll face many times in your product career as other teams such as tech or marketing need to plan their resources and activities well ahead.</p><p>Let me share <a href="https://docs.google.com/spreadsheets/d/1YVHcm8dIP_3NTglYWsk2UR_QBS7aREVBv_-YQvP9VZk/edit">this hack</a> I recently used when dealing with this exact request at the South East Asian super app <a href="https://www.gojek.com/">Gojek</a> that stitches together a few well-known frameworks. It solves your immediate problems, forms a basis for a more robust validation later and becomes the single place for all big cross-team alignment.</p><p><strong>Problem</strong></p><p>I recently took over a large digital product and after a week got invited on the Friday to a Monday off site whereby we needed to present next year’s roadmap to 40 senior stakeholders who all had very strong opinions about what needed to be built.</p><p>I’d received many Slack messages beforehand with bold feature ideas. But I didn’t want to promise anything I just wasn’t sure about. So how to get them all aligned so quickly without freaking out our engineering teams?</p><p>After reading all our research slides and procrastinating for half a day, full on panic was setting in.</p><p><strong>Ah hah moment</strong></p><p>I decided to take a short walk that Sunday afternoon and an idea hit me. Why not take all the team’s ideas over the last 12 months and put them all on the table. But then express the pain points from the research using JTBD style outcome statements giving each one an <a href="https://medium.com/@adamdarcy/jtbd-how-to-find-the-opportunity-for-innovation-49144ea4a1aa">opportunity score</a> to enable feature ranking/excluding. Then the entire conversation can be filtered to customer pain, market size and engineering effort.</p><p>In just a few hours of intense focus, I had a defendable outcome based product roadmap showing:</p><ol><li>What <a href="https://medium.com/@adamdarcy/jtbd-what-the-job-c394dba1d164">functional jobs</a> we were solving for e.g. <em>listen to music privately</em> — turns out we had 5 jobs (♥️ by all as grounded our thinking going forward)</li><li>What were the pain points expressed as <a href="https://medium.com/@adamdarcy/jtbd-getting-from-in-depth-interview-to-outcome-statements-505699ee01b2">customer outcomes</a> e.g. <em>minimise the time it takes to start listening </em>with associated <a href="https://medium.com/@adamdarcy/jtbd-how-to-find-the-opportunity-for-innovation-49144ea4a1aa">opportunities scores</a> so we could rank them (♥️ by research, design and product)</li><li>All feature ideas linked to above problem statements (♥️ by engineering who just hate waste)</li><li>An MVP cut of features followed by subsequent packages of suggested releases over the year, each with a clearly defined needs based proposition and associated market size (♥️ by engineering and marketing)</li></ol><p>It worked.. the next day in a short 20 min presentation followed by Q&amp;A, we managed to get everyone to the same place whilst shutting down many of the bigger ideas that either weren’t solving high priority problems or giving us any any scale — and at Gojek we only solve for scale!</p><p>We were now ready to begin design and setup tech resourcing whilst our research team hungrily validated my qualitative based assumptions with a quantitative survey whilst marketing got busy with GTM plans and engineering began t-shirt sizing.</p><h3><strong>Step by step guide</strong></h3><p>As per my other blogs on JTBD, I’ll use the following job to help provide a clear example of the framework:</p><blockquote>Listen to music privately</blockquote><p><a href="https://docs.google.com/spreadsheets/d/1YVHcm8dIP_3NTglYWsk2UR_QBS7aREVBv_-YQvP9VZk/edit?usp=sharing">Here</a> is the final output in Google Sheets if you want to skip ahead.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*n-WA9DzGM7EijHl84QSDug.jpeg" /><figcaption>Listening to music privately is our core example job</figcaption></figure><p><strong>Step 1: Define the job</strong></p><p>Define the functional job your product is trying to solve for. Could be multiple and can use <a href="https://medium.com/@adamdarcy/jtbd-what-the-job-c394dba1d164">this blog</a> to help you. Put this/these into column A. Every pain point and feature will be anchored to a core job from now on.</p><p><strong>Step 2: List outcome statements and rank for importance &amp; satisfaction</strong></p><p>In column B, list all outcome statements customer is trying to achieve as they execute the job in column A. These are the core needs you might address. Use <a href="https://medium.com/@adamdarcy/jtbd-getting-from-in-depth-interview-to-outcome-statements-505699ee01b2">this blog</a> to help pull these from any qual research you may have done.</p><p>In column D, put which step of the job the customer is at when they need this outcome to give you more context. I use 4 simples steps here to bucket them:</p><ol><li><strong>Setup: </strong>Before listening e.g. finding listening device, connecting wires</li><li><strong>Execute:</strong> The moment you play e.g. initial volume, wrong device</li><li><strong>Monitor:</strong> While listening, how to monitor and adjust e.g. volume, track</li><li><strong>Conclude:</strong> When you want to stop listening e.g. stop music, put away</li></ol><p>Finally (fun part), for each outcome statement, add a score out of 10 for how important your think this is for your average user and how satisfied they currently are with their solution. I’ve written a post on this part <a href="https://medium.com/@adamdarcy/jtbd-how-to-find-the-opportunity-for-innovation-49144ea4a1aa">here</a> for you. Call out that these numbers need to be validated as a next step, but that doesn’t stop you filling them in right now to get you to a draft roadmap.</p><p>The opportunity scores will appear with a higher number being something you should probably try and solve for with your features and a lower one maybe not important. If importance and satisfaction are both above 6, the tool marks it as hygiene i.e. you just got to do this to even play the game.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1016/1*g6r8JgIlsTz8eedzAmFpgg.png" /><figcaption>Give each outcome an Importance and Satisfaction score to get your opportunity score</figcaption></figure><p><strong>Step 3: List all features and do MoSCoW assessment per outcome</strong></p><p>In the top row, list all the big features you’ve heard from customers, founders, team members (big believer that ideas should come from everywhere in innovation). This makes everyone feel heard.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/648/1*dB9iRiMeyeHURxgq1JrzHQ.jpeg" /><figcaption>Everyone gets a say in JTBD, but be sure to keep it high level for this exercise</figcaption></figure><p>Don’t go too low level here. You want everything on a single screen otherwise you will lose your audience and they won’t use the tool going forward. Remember this is a stakeholder communication and alignment tool.</p><p>Next, for each outcome statement, mark the feature using MoSCoW method of prioritization as follows:</p><ul><li><strong>Must: </strong>If we don’t do this feature we CAN NOT solve the pain point</li><li><strong>Should: </strong>Unless we are strapped for time, we really should also do this</li><li><strong>Could:</strong> Would be nice to have but really not essential to solve pain</li><li><strong>Won’t:</strong> What won’t we do to solve this pain point (important to call out)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/761/1*ATH7-6wR3L4lZV8GOaGhXQ.jpeg" /><figcaption>Easily see what features we should focus on for each pain point</figcaption></figure><p><strong>Step 4: Select your MVP and write market value proposition</strong></p><p>Ok, so we can now clearly see what features are solving what pain points. Now for a bit of validation, ask yourself 2 questions:</p><ol><li>Do we have features that cover all the hygiene pain points for this job?</li><li>If a pain point doesn’t have any Must features, why is that? Have we forgotten a feature? Or maybe it just isn’t important enough? That’s where marketing sizing can help which we will discuss next.</li></ol><p>The next step is to package releases by selecting what must have features we want to build to solve what pain points in the market. I had Timo our marketing guru on a Slack between his Sunday family time to help me understand the market size for those pain points.</p><p>We probably don’t want to select features that solve pain points that fit only 20% of total addressable market cares about e.g. noise cancellation. We probably do want to package features that together hit 60% of the market’s pain points and really differentiate us e.g. just let me start listening quickly. This is an art more than a science so be ready for bit of back and forth here.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*6Ft9ls4pk0HKzSKD2t41FA.png" /><figcaption>Package the features into releases that solve for large pain points not forgetting hygiene factors</figcaption></figure><p>The value proposition is so so easy once you have identified the biggest market pains you will solve. For example, above I crossed checked my suggested release 1 with <a href="https://www.apple.com/airpods-2nd-generation/">Apple’s actual market</a> value proposition and was pretty spot on:</p><blockquote>Simply take them out and they’re ready to use with all your devices. Put them in your ears and they connect immediately, immersing you in rich, high-quality sound.</blockquote><p>Their wireless H1 chip for seamlessly connecting between all your Apple devices clearly solved the pain point around time spent to start listening.</p><p>I might wait for release 2 to solve more niche pain points like noise cancellation as the engineering team may need more time to figure this out and pending quant I’ll assume less people care about that so it’s more incremental. The fiddly job of ensuring a smooth fit in anyones ear when they do exercise sounds like a smaller market too so will also hold on that.</p><p><strong>Next steps</strong></p><p>Now everyone is nicely aligned, go do your quant survey to validate the assumptions on market size for these pain points. With a good research team this takes 2–4 weeks if you are good. Once back, you will likely adjust your opportunity scores, which could change your product roadmap. Same applies to engineering estimates.</p><p>The think I love the most about the survey is that if you go big, you get to play around with the results to group people based how much pain they feel for different outcomes. These become needs based segments that you may want to give slightly different messaging to. More details in <a href="https://medium.com/@adamdarcy/jtbd-how-to-find-the-opportunity-for-innovation-49144ea4a1aa">this</a> post, but as an example, you may discover 2 groups:</p><ul><li>People who get frustrated at setup time and need the wireless feature also happen to love exercising</li><li>People who get really frustrated as they switch devices a lot are office workers who constantly move between conference calls and music and may mistakenly have loud music streaming from the player they aren’t actually connected to.</li></ul><p>Can imagine an instagram ad having very different messaging for these groups.</p><p><strong>Finally</strong></p><p>I really hope this framework helps anyone who finds themselves in the same position as I often have.</p><p>Would love to hear any feedback so can continuously improve as have used on 3 products so far and works great for me and my teams, but doesn’t solve everything i.e. can we quickly factor in effort into the equation somehow earlier.</p><p><em>Note: Original JTBD framework and ideas are a mix of </em><a href="https://www.youtube.com/watch?v=Q63PZR7mG70"><em>Clayton Christensen</em></a><em> and Tony Ulwich and highly recommend reading </em><a href="https://www.amazon.com/Innovators-Solution-Creating-Sustaining-Successful-ebook/dp/B00E257S7C"><em>Innovator&#39;s Solution</em></a><em> and checking out </em><a href="https://strategyn.com/jobs-to-be-done/"><em>Strategyn</em></a><em> for more detailed info from these 2 brilliant people.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b6e3438b7f0c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[JTBD: What the job?]]></title>
            <link>https://adampdarcy.medium.com/jtbd-what-the-job-c394dba1d164?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/c394dba1d164</guid>
            <category><![CDATA[product-strategy]]></category>
            <category><![CDATA[outcome-driven-innovation]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[jtbd]]></category>
            <category><![CDATA[ux]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Mon, 27 Jan 2020 18:40:32 GMT</pubDate>
            <atom:updated>2020-01-28T09:44:05.196Z</atom:updated>
            <content:encoded><![CDATA[<h3>JTBD: What is the job?</h3><p><strong>The most fundamental and often trickiest part of JTBD is defining exactly what is the job you are help customers get done.</strong></p><p>I’ll use a social payments product called <a href="https://payme.hsbc.com.hk/personal">PayMe from HSBC</a> I created back in 2015 to try and help explain.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*4SzNaRKGLjxezJDP0khHkg.jpeg" /></figure><p><strong>First, get an idea of the job you want to either help with or do..</strong></p><p>I’ll assume you have an idea already to get stuck in, although will call out the two very different starting points:</p><ul><li><strong>Startup</strong> — If you are a startup founder, you would have already seen or felt a problem and can simply begin to define the job from here (the roots of the PayMe product were actually from London in 2012 when I found it was such a pain to pay back small amounts of money if you didn’t happen to have the exact cash and coins).</li><li><strong>Established Company</strong> - If you are an existing company, you probably want to start from your current position — unless you are trying to reinvent yourself. When I worked in HSBC Asia’s innovation team, we defined exploration themes that were in and around the banking industry e.g. payments, blockchain, digital identity, where we could leverage our assets e.g. customer base, distribution, brand trust to try to enter and win adjacent markets. It was in this team we analysed and sized the pain points around paying friends back in our largest market — Hong Kong.</li></ul><p><strong>Unpack the problems..</strong></p><p>So we found a high level problem in that it was very difficult to pay people in Hong Kong due to cross-bank fees and the just awful user experience of local banks (Hong Kong was/still is surprisingly far behind on digital UX). We would witness people writing checks over the dinner table or walking from one bank’s ATM with cash and putting it into another bank’s with the wind literally blowing it away sometimes if they weren’t grasping it firmly enough (true story).</p><p>The majority of the time, it turns out, these people were paying their friend’s circles. Friendships (thanks <a href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3385686/">Robin Dunbar</a>) are a function of quality time spent and frequency of interaction and due to matchbox apartments packed with generations of family, Hong Kongers simply love to go out after work to eat shareable local foods like dim sum or hot pot and drink beers. In all these shared moments, money was inevitably exchanged — albeit with a lot of friction. If you think you can improve the experience by 10x, the market really is your for the taking (if you haven’t read, I highly recommend <a href="https://www.goodreads.com/book/show/61329.Crossing_the_Chasm">Crossing the Chasm</a>).</p><p><strong>Allow the core functional job to appear..</strong></p><p>Slowly but surely, an abstract core functional job will begin to emerge from the dust left behind from your problems statements.</p><p>A well defined job statement should clearly express the function the customer is trying to execute, whilst removing any emotional or need statements (that comes <a href="https://link.medium.com/6yUqi4U0A3">later</a>, don’t worry). It should also reveal the object and if required a contextual clarifier. Also avoid using adjectives or adverbs so you can really get to the core.</p><p>After some deliberation.. the job we were helping Hong Kongers with was:</p><blockquote>Pay money to friends</blockquote><p>I instinctively wanted to put the adjectives like ‘small amounts of money’ here to refine the job, however that would have been a mistake. Firstly, define small? It’s ambiguous to the reader. An affluent person may think $1,000 is small whereas a young university student may pass out at the thought of owing a classmate that much cash. Secondly, it is important not to add these type of constraints at this point. When you solve the job for your intended market, you may be surprised at the other uses you didn’t think of, which may make you miss the opportunity to create a much larger impact later.</p><p>Contextual clarifiers like ‘to friends’ help hone in on the core group of customers you want to solve for, but again we could’ve left this out due to it creating market constraints. On reflection I believe it was the right thing to do as it really helped us focus from the beginning on the inner friendships circles that made us successful.</p><p>So we found our job and that now becomes the point of analysis for future work. Just to remind readers, JTBD removes us from customer journey maps so we really focus on the core mechanics of the thing you just want to get done and improve it.</p><p><strong>Emotional and Social jobs</strong></p><p>Whilst analysing for solving the core functional job ‘pay friends money’, we will see many side jobs the customer wants to execute.</p><ul><li><strong>Emotional jobs: </strong>How do they want to feel when paying friends back?</li><li><strong>Social Jobs:</strong> How do they want people to perceive them whilst they pay their friends back?</li></ul><p>You see the difference? One is about being in control of their own emotional state e.g.“make me feel like I am a good friend by always paying money back” or “make me feel like I’m not being cheap”. These really depend on the values of the individual, although are usually shaped by the culture they are operating in. The other is about how others perceive them socially e.g. “make others feel like I always pay my debts” or “make others see I’m not being cheap”. These are more strongly shaped by the culture so can change a lot from country to country e.g. taking 10 minutes to split a group bill line by line may be socially acceptable in a culture that perhaps values thriftiness, but may make you quickly lose friends in another.</p><p>For each functional job, you will find many associated emotional and social jobs. Keep these in mind as they can create huge differentiation between your’s and your competitor&#39;s product even though you are both solving the same job as we found with PayMe.</p><p><strong>Is there just one core functional job?</strong></p><p>Yes and no. Yes you should set out to solve the core job and use that for your analysis, but at the same time you may find you solve other jobs too with little extra effort. With PayMe for example, within a few weeks we changed a few lines of code to enable another functional job:</p><blockquote>Request money from friends</blockquote><p><strong>Summary</strong></p><p>So there you have it. A functional job statement that is abstract enough to give you lots of room for maneuvering around problem statements and markets to finally creating a winning product. Now go launch something spectacular! 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c394dba1d164" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[JTBD: How to find the opportunity for innovation]]></title>
            <link>https://adampdarcy.medium.com/jtbd-how-to-find-the-opportunity-for-innovation-49144ea4a1aa?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/49144ea4a1aa</guid>
            <category><![CDATA[product-innovation]]></category>
            <category><![CDATA[outcome-driven-innovation]]></category>
            <category><![CDATA[ux]]></category>
            <category><![CDATA[jtbd]]></category>
            <category><![CDATA[product-strategy]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Wed, 15 Jan 2020 17:13:00 GMT</pubDate>
            <atom:updated>2020-01-15T17:17:23.977Z</atom:updated>
            <content:encoded><![CDATA[<p>Once you have derived your outcome statements from your in-depth customer interviews (I explained outcome statements in this <a href="https://link.medium.com/Xrv1QGQUg3">blog</a>), you will want to create a survey to get quantitative feedback to see which pain points are felt most by which segments.</p><p>It always take a bit of time to get a survey prepared and sometimes longer to get the results back. However once you get the results, some really juicy insights can jump out to take you closer to creating a winning product that solves a genuine problem.</p><p>I’ll continue with the example from the previous blog for the job to be done:</p><blockquote>Listen to music privately</blockquote><p><strong>Profile and competitor questions</strong></p><p>Begin the survey with some profiling questions in like age, location, income etc as these are very useful to segment results later.</p><p>Ask how they get the job done today e.g. “How did you listen to music privately today?” I would answer “Apple Earphones” (It is better to use a multi-choice question that consist of top 4 to 5 products in the market that solve this job, so you can do competitor analysis later). This also helps the respondent visually recall the last time they executed the job, which gets their mind prepared for the outcome questions that follow.</p><p><strong>Outcome Statement Questions</strong></p><p>Layout the outcome statements as questions and allow participants to score each of the following criteria out of 5:</p><ul><li><strong>Importance</strong>: How important is it for the respondent to achieve this outcome when they are performing this job.</li><li><strong>Satisfaction</strong>: How satisfied is the respondent with the outcome when they last performed this job.</li></ul><p>Let me continue explaining using the following outcome statement for the above job “Listen to music privately”:</p><blockquote><em>Minimise the time spent to start listening to music</em></blockquote><p>As a recap, this was around the pain point discovered in the interviews whereby customers were having to untangle their earphones each time they wanted to begin listen to music.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/329/0*GjqG_8oECKgIXFDd.png" /></figure><p><strong>Importance</strong></p><blockquote>[When listening to music privately], how important is it for you to minimise the time spent to begin listening?</blockquote><p>Try it yourself. Rate between ‘1 — extremely unimportant’ and ‘5 — extremely important’. For me this is a ‘5’ as I like to listen to a song or two whenever I get a free moment, therefore I want to maximise listening time. It could be a ‘1’ for people who usually listen to music for a long period of time each time, as their set up time is at a relatively small percentage of their total listening time.</p><p><strong>Satisfaction</strong></p><blockquote>[The last time you listened to music privately], how satisfied were you with the amount of time you had to spend to begin listening?</blockquote><p>Notice I changed to ‘the last time” at the beginning of the question. I do this because the way humans recall memories is just plain flawed (for more information, I highly recommend the books <a href="https://www.goodreads.com/book/show/34466952-the-power-of-moments">Power of Moments</a> or <a href="https://www.goodreads.com/book/show/31138556-homo-deus">Homo Deus</a>). It’d be easier for them to more accurately recall the last time they executed this job even though they may vary how they perform it over time.</p><p>Again, try to rate it. For the 5 years ago me I would give it a ‘1 — Extremely unsatisfied’, purely because when I used the Apple earphones above, whilst the rubbery chord makes it more durable against breakages, it is harder to untangle each time I wanted to listen to music compared to cheaper plastic materials.</p><p><strong>How many questions?</strong></p><p>Theory states you should have 75+ outcomes for a job, however in my experience this creates a mind numbingly boring exercise that sends people to sleep, and you will probably get very inaccurate results in the second half of the survey as the respondent’s concentration has wane. In my opinion, 20–30 is the maximum number of questions people can handle in a single session. My team recently did a 50+ questions survey for low income mothers in Indonesia and even though it was a guided survey outside of convenience stores, we found we had to split it in half and join results later.</p><p><strong>How many people?</strong></p><p>It depends. Textbooks tell me 150–5000 respondents depending on what kind of market and how much you want to segment your data later. The larger sample size the better, however it comes at a cost and you may just not be able to find that number of people for very specialist jobs e.g. performing heart bypass surgery.</p><p><strong>The Results</strong></p><p>Aah, the great feeling of anticipation when the results comes back and you are (im)patiently waiting for the summary to be reported. But what do you do with the results? Well the good news is they can be used for many very important things.</p><ul><li><strong>Product Strategy</strong> — As a product manager, the thing I love most about JTBD and why I really can’t wait to get my hands on the results is it literally outputs my product strategy.</li><li><strong>Value Proposition Creation</strong> — Discover needs-based segments of potential customers and market the product to them in ways they will respond. I’ll write more on this in another blog.</li><li><strong>Competitor Analysis </strong>— If you included ‘how did you get the job done today’ in your survey, you can easily see where that product is failing.</li><li><strong>Market Strategy</strong> — Allows selection of needs based segments so you can formulate how you will attack each one. e.g. people tend to type the job or desired outcome as the keyword for google e.g. ‘quick way to untangle earphones’. We can create a bidding strategy around these keywords.</li><li><strong>Business Strategy</strong> — The survey lets you know what areas are under-served or over-served so you know which strategy to employ in the market e.g. People want better performance when listening to music privately and are underserved. If I can solve for this in my product, I can charge more.</li></ul><p><strong>Example of using output for Product Strategy</strong></p><p>A quick example of how I might use the survey results for my product strategy is: a single outcome that score favourably might literally map to a feature that I need to include as a high priority in my first release.</p><p>According to my (fictitious) survey, it shows that the entire market also agree with my personal frustration around the set up time to begin listening to music. We can create a needs-based segment and differentiate on solving that pain point. I was very happy when Apple finally did solve this with the AirPod launch. As we weren’t limiting ourselves by thinking of current solutions, we can now take a look at how good enough Bluetooth has become to play a role in solving this problem. This is real product innovation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*9srF22OKt3HqeIcenMRN8w.png" /><figcaption><em>Outcome statement: Minimise the time spent to start listening to music</em></figcaption></figure><p>I’d still only give them a ‘4 — Satisfaction’ instead of ‘5 — Highly satisfied’ as I still struggle to switch between my MacBook and iPhone.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=49144ea4a1aa" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[JTBD — Getting from in-depth interview to outcome statements]]></title>
            <link>https://adampdarcy.medium.com/jtbd-getting-from-in-depth-interview-to-outcome-statements-505699ee01b2?source=rss-46fba4aac79e------2</link>
            <guid isPermaLink="false">https://medium.com/p/505699ee01b2</guid>
            <category><![CDATA[innovation]]></category>
            <category><![CDATA[jtbd]]></category>
            <category><![CDATA[product-management]]></category>
            <category><![CDATA[outcome-driven-innovation]]></category>
            <category><![CDATA[jobs-to-be-done]]></category>
            <dc:creator><![CDATA[Adam D'arcy]]></dc:creator>
            <pubDate>Wed, 15 Jan 2020 14:52:47 GMT</pubDate>
            <atom:updated>2020-01-16T01:15:58.832Z</atom:updated>
            <content:encoded><![CDATA[<h3>JTBD — Getting from in-depth interview to outcome statements</h3><p>One of the common questions I hear in the JTBD process is how to get to customer outcome statements from your in-depth interviews.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/289/0*CB42JlVX4oNsFs-7" /></figure><p>For me this really is the most critical stage of the whole framework and requires real discipline to remove any bias on current products or solutions they might use.</p><p>Here are some tips I’ve tried myself after reading various books and blogs. I hope you find them useful. It take a few goes, but I promise it does become easy if you stick with it and definitely worth the effort in the long term.</p><p>To help explain using a real example, I’ll use the following functional job to be done throughout:</p><blockquote>Listen to music privately</blockquote><p>This is a job most of us like to execute when we need some escape from the noise of daily life. The job could of course be abstracted to “Help me escape from daily life”, opening the range of potential solutions, however I won’t go there here and I’ll write another blog on defining the core job soon.</p><p><strong>Focus on Outcomes, not Solutions</strong></p><p>The great thing about outcomes, is that later on they lead directly to solutions. But you must avoid the temptation now. It takes some practise, but innovation only truly works when you learn to detach your mind from how things are done today.</p><p><strong>Tune your ear to hear their pain points</strong></p><p>Focus on the underlying job process and listen carefully to hear the difficulties they face along the way. e.g. when I decided I want to listen to music privately, I might be frustrated at how long it takes to untangle my earphones. Or I might find the left earphone isn’t working because the cable is severed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/329/1*96MfdkSwUprYyEg_8wqytg.png" /><figcaption>We’ve all been here right?</figcaption></figure><p>Zoom in on these pain points. Ask how they would like the step ideally performed if there were zero limitations.</p><p>You are getting closer to writing a couple of outcome statements. But first let’s try to define the structure to express it better.</p><p><strong>There are generally 3 types of job outcomes</strong></p><p>I have found all outcomes can nicely fit inside 3 categories. When executing a job, people are always trying to:</p><ol><li>Eliminate <strong>time</strong></li><li>Eliminate<strong> inefficiencies, </strong>and</li><li>Avoid <strong>problems.</strong></li></ol><p>If I listen to music 5 times a day and have to untangle my earphones every time, that pain point may even increase on each iteration. So what I really want is to reduce the <strong>time</strong> it takes to setup and also eliminate the <strong>inefficiency</strong> of the cables keep getting tangled again and again. If I plug my earphones in I want to avoid <strong>problem</strong> of only one ear working.</p><p><strong>Some basic grammar rules</strong></p><p>Below are some rules to help express your outcome statements in a uniform way so you can compare them fairly when weighing up what to solve for later.</p><p>The syntax is:</p><blockquote><strong>Minimise</strong> the <strong>[“time”|“likelihood of” |”likelihood that”] [object of control] [contextual clarifier]</strong></blockquote><p>I’ll explain each part below:</p><ul><li><em>The type of improvement required</em> — use the word “<strong>minimise”</strong> as customers always trying to reduce process, not increase.</li><li><em>The unit of measurement : </em><strong>[“time”|“likelihood of” |”likelihood that”] </strong>(3 categories above)</li><li><em>An object of control </em>— this explains precisely what the customer is trying get done faster (time), more predictably (avoid problems), or with higher throughput / output (eliminate inefficiencies).</li><li><em>A contextual clarifier </em>— this identifies the conditions or circumstances a customer may be encountering. While it can often be left out, it helps to ensure the outcome statement is not misinterpreted.</li></ul><p><strong>Some examples of outcome statements</strong></p><p>Now we have some rules, let’s try to apply them to the pain points we discovered above. For the frustration of untangling earphones, we may want to write an outcome statement related to <strong>time</strong> like:</p><blockquote>Minimise the time spent to start listening to music</blockquote><p>Firstly, notice there is no mention of earphones. That is a solution and constrains our thinking. The category is ‘time’, the object of control is ‘listening to music’, context ‘start’ indicates that it is during the setup stage of the job execution and we begin the sentence with ‘minimise’.</p><p>We might also write another to capture the <strong>inefficiency</strong> of the tangled cable pain point as:</p><blockquote>Minimise the likelihood of having to do manual tasks before I can listen to music</blockquote><p>Another few examples related to the <strong>problem</strong> of severed cable pain point resulting in the left earphone not working:</p><blockquote>Minimise the likelihood that there is uneven sound in ears when listening to music</blockquote><blockquote>Minimise the likelihood that I have poor sound quality during playback</blockquote><p><strong>Beware: What you need to remove from your statements</strong></p><p>Remove both of these from outcome statements:</p><ul><li>Remove <strong>Ambiguous words</strong> like ‘easy’ and ‘reliable’ as they mean different things to different people and can be removed by asking what they think are easy or reliable to drive to an objective outcome. E.g. If they person says ”I want it to be easier to start listening to music” you could ask “what do you want to be made easier?” “It takes too long to untangle the chord” until you can get to a more objective statement like “I want to reduce the time it takes to begin listening to music.”</li><li>Remove <strong>Solutions</strong> words such as those that describe specifications or constraints e.g. if they say “I want stronger chords on my earphones” (stronger chord is a solution.. but why do we need chords at all? And earphones? What if technology has now evolved to just insert plants in my ear?). Get them back to the process and how they’d like it to be ideally and you’ll find a more objective outcome statement like “<strong>Minimise</strong> the <strong>likelihood</strong> that I can’t hear sound when trying to listen to music”.</li></ul><p><strong>Finally.. confirmation</strong></p><p>Confirm each outcome statement with the interviewee to eliminate guesswork e.g. “<strong>Do you mean that</strong> you want to minimise the likelihood that you can’t hear sound when trying to listen to music?”</p><p><strong>Then what?</strong></p><p>What do do with all these outcome statements? These go into a survey so you can see which pain points are shared across the market segments. This ultimately leads directly to your market and product strategy which I&#39;ll explain in this <a href="https://link.medium.com/LzSWQl0Ah3">next blog</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=505699ee01b2" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>