Discovery – half way point learnings

This week marked the halfway point of the first stage (Discovery) of the case I’m currently working with. We are three weeks in and have three more weeks to go. So what have I learnt about the process so far?

(Lingo: We refer to experiments as almost any effort we make to test an idea. I.e. customer interviews are experiments, landing pages tests are experiments. Internal activities such as brainstorming sessions are not experiments)

Six weeks is a short time period, two to three experiment might me reasonable

Our Discovery process aims to find evidence of Desirability in the idea. This narrow scope is one of the reasons we believe 6 weeks is long enough. In our process we want to work in iterations, finding a problem and then exploring it with various experiments. Before starting the Discovery process my naive ambition was to run one experiment per week.

The first experiment we wanted to run were exploratory interviews. Since I mainly work with two-sided marketplaces we knew we needed to interview at least 2 groups of people, the “supply side” and the “demand side”. However due to lead times in finding the right customer to interview, contacting them and setting up meetings the first external interviews were held at the end of week 2.

In week 3 the interviews has started for real. Me and my colleague ran ~20 interviews with both business customers as well as consumers. Especially during week 2 and 3 our tempo has been high. However the one experiment per week that was our ambition was not feasible in this case. 

Six weeks are likely a long enough period to run between 3 to 4 experiments. Maybe future experiments will take less time than interviews and that number can increase further but right now 3 to 4 seems like a good ambition.

Start booking interviews right away

One learning then, is to really start booking up interviews during the first day of the case. This is especially true if the target customers are businesses which typically have longer lead times to get in contact with. 

So next time, I won’t worry too much about reaching out to a “wrong segment”, the important thing is just to start booking up interviews.

Exploratory interviews are great

When trying to learn about new markets and finding problems to solve, exploratory interviews are great. Especially when interviewing people from various backgrounds you can easily start to pick up themes that can serve as hypotheses further on.

Pivot to more hypothesis driven research

As great as exploratory interviews are there is a risk of “over exploring” where you don’t really commit to a certain problem. Tomorrow (1st day of week 4) we will shift gears and commit to one of the user problems that has arisen. We will also start exploring a solution to those problems. We are open that the solution might change. The reason we are doing this is to get deeper into the problems and to do this we need to focus more narrowly. 

All in all, this process is really fun, creative and also a bit challenging. I’m looking forward to the coming weeks. Next week I will probably not write a post so see you in two weeks!

Working with new products

I’m fortunate to work in a company with growth ambitions and that also realize growth, in the long run, can not only come from improving the core products. Instead the company also need to push beyond what we do today into adjacent and new areas. We believe this to the degree that our entire organization is based around the concept of Core on the one hand and Adjacent & New on the other. I believe it’s necessary to have separate processes and even a separate organization for creating new businesses. 

New things are inherently more uncertain (they don’t exist yet) and therefore harder to forecast and plan. This means it’s hard to fit creation of new things into normal company processes.  Basic questions such as: “What do you expect your revenues to be this year” are expected to be answered and in Core this is usually not a problem. With new businesses however, these types of questions are at best a waste of time but could also harm the process by forcing teams into plans and expectations that are not the best for the new business. The new business need processes that obsesses over customer feedback and quick experimentation. 

With new businesses we work in three “stages”. The first stage is a very short (6 week) Discovery with a focus on really understanding the customers, their gains, pains and jobs to be done. After Discovery comes Nail-it. Given that we have found some real customer pain-points in Discovery this is where we try to build something that in a very simple way solves (“relieves”) these pain points. The things we build are very much in the MVP- style and we expect to usually throw away most of our code. Finally we have our Scale-it stage. In Scale-it we deploy our learnings from the nail it process and start building a real product that we distribute to more of our users through internal and external channels.

I currently work with a project in week 2 of the Discovery stage (i.e. very early). Next week I will start reporting on our learnings from this process, what’s easy what’s hard. So far I’ve found that frame-works that seem straightforward to use requires a lot of bending to stand the test of reality. More on that next week!

Your traffic is made of users not water

Ok a personal update first, I failed on my last post for my 2019 goal. It was just that Christmas was so good that I decided not to write the last post. I guess there is something poetic of coming close but not succeeding as well? Anyway, I decided to keep writing. I will try to keep a weekly cadence but will skip artificial targets. Ok so on to the actual post. 

I quite commonly hear people refer to traffic as if it is water that can be “directed” and lead as desired. There are at least three examples when it’s common to hear this.

When marketing other services. A site or destination with much traffic often fall into the trap of thinking that they can direct their traffic however they want. True, an influencer with a lot of trust can direct her followers to another product or service but this is a transaction of trust. If the followers don’t understand the product the influencer is directing them to then trust erodes and the CTR rates are likely to be low. 

When optimizing funnels. It’s easy to fall into the trap of optimizing certain steps in a funnel without thinking about what this means for the user. A softer commitment like a “Read more” – Call to action can lead to a higher top-funnel conversion but this will likely be traded for a lower conversion further down in the funnel. I believe the correct way to think in funnels is first to decide how to explain the offering in as good a way as possible first. Then of course optimizing copy can be worthwhile.

When launching new services. When expanding a business to adjacent areas and “jobs to be done”. A common logic is: “if only 5% of our users do the new job our revenues will grow with X million”. These arguments are often powerful, especially if you have large traffic because they generate great looking business cases and also sound quite conservative (“only 5%”). The problem is just that the 5% can easily turn to 0.05% if the users don’t understand what the new offering is or why they should use it. 

In some of the examples above the distinction might feel subtle. But recognizing that the traffic you have on your site is actually made up of users that need to understand and desire the things you put in front of them can make you avoid some costly mistakes.

Products need to prove that jobs are being done

This post refers to “jobs” which is a reference to Clayton Christensen’s work on Jobs to be done (

Your product should help users get their “jobs” done. Apart from helping the users get their jobs done the product also need to show the user that their jobs have, in fact, been done. For some jobs this is the same thing. For example: buying a power drill to make a hole, the user of that product knows when the job has been done (when his new poster is hanging on the wall).

However some jobs are not as visible. Take the example of internet advertising. The customer’s job is to drive traffic to its site (and ultimately to increase sales). However in this case the customer doesn’t know when you have helped him with his job. What you do instead is to provide data giving proof that the job has been done.

This might sound very banal but I believe it’s quite common that products underdeliver on giving proofs on jobs to be done. Customer’s then start distrusting your product and starts suggesting features. I.e. they don’t trust that their jobs are being done so they try to come up with suggestions on changes in the product that could make it solve their jobs better.

This is not the type of discussion you want to have. You want to talk effect and outcomes, not features. And one way to get there is to give the users proof (usually in the form of data) that their jobs are being done.

Getting things done

Blog post number 17. I don’t really have anything to write this week other than that I’ve been reading David Allen’s Getting Things Done and implemented a new personal system for productivity.

I’m using the app “Todoist” which seems OK, however it is not strictly adopted to the Getting things done framework. Makes me wonder if there is an opportunity for creating an “todo-app” (like me previous idea on that is based on the GTD methodology. Like Superhuman it would prod the user do work a certain way with it.

One issue with most todo-app is that they are blank canvases where you can apply almost any methodology. The flip side of this is that you don’t get much help in your organizing work. The app I’m picturing would be very game like in that it would prod you to take certain actions until you “in”-box was empty.

I will try to use Todoist and see if it works well enough. If I feel I want a more helpful app I might try to think up one.

The role of the business case in digital products

Business cases are the backbone of many traditional companies’ strategies. They outline how revenues and cost occur over time and lets management prioritize initiatives after the value they create. As I discussed in post the zero marginal costs of digital products and their malleability leads to a high regard of uncertainty. One practical implication of this uncertainty is that business cases become less valuable. They are not useless, but managers need to recognize this uncertainty and use them slightly differently.

First consider a more traditional company Bottle Cap Co that are currently evaluating whether or not they should launch a new type of Bottle Cap product. They have a high fixed cost in a new cap-press, and to cover the fixed costs the company needs sell alot of caps . However at one point the press will be at full capacity and adding additional volume will require the purchase of an additional press. Their profits as a function of volume therefore looks something like this:

At a certain volume the company need to invest in another press. This means that profit optimization is not always as simple as pushing volumes. This is a situation where a business case is a very useful tool. Bottle Cap Co can model their cost structure and find our what the optimal volume is for their current production capacity. Since reality is usually much more complex than the simplified example above this is a very strong benefit of business cases in this type of business.

Contract this to a digital product. The cost component is very straight forward: a product team of 4-5 FTEs and some almost negligible cost for cloud services. On the revenue side, though some digital products have quite complicated business models most of them (advertising, subscriptions etc.) maximizes revenues with higher volumes / adoption. The business case therefore does not provide the same insights for digital products and you usually end up staring at one assumption that determines the entire case: adoption (i.e. how many will use the product).

I don’t believe that creating a business case for digital businesses is time wasted though. Going through the motion forces you to write down any assumptions you have of the product. Maybe there are other assumption in addition to adoption that is very high risk? These assumptions, written down with the levels you “need” to reach, is a good starting point for running experiments / bets.

Managing sustainability from a product perspective

This week I participated in a workshop on sustainability and circular economy. This is a topic I care about and discussing it in a facilitated workshop-setting was both interesting and fun. However this is a setting where the risk of “Good ideas” is very high.

I believe many people recognize the situation. You work around a theme that many are passionate about. The workshop facilitates innovative ideas and at the end of the day you have some 3-5 interesting ideas that everyone is excited about. The issue is what happens next? In a company with weak or no product process, what happens next will likely depend on who was in the workshop. If it was a senior person with some sway, he might come back from the event, inspired and with a clear idea on how to move the needle. This is understandable but not how I believe you actually make in impact in product (and not how it was in my case). 

So how do we get Sustainability into our products. The tricky thing about sustainability is that users usually don’t respond very well to it. They say they want it when asked, but their behaviour usually shows that sustainability is not a strong enough reason to persuade them to take a particular action (e.g. conversion rates to climate compensation products are usually very low). 

What is positive about this is that we’re already used to dealing with similar situations. When the company wants to achieve something (usually financial) we call this a desired business outcome. Users usually couldn’t care less about our desired business outcomes but we still find ways to do it (see me earlier post). I believe sustainability should be treated as any other business outcome. 

The question then becomes, how does sustainability stack up against our other desired business outcomes such as increasing revenues for this specific quarter? I believe this is the real problem that a company needs to solve if they want their product teams to innovate around sustainability. If we want product teams to innovate around sustainability then: define a metric, and set it as a priority. If you have a strong product process this should be enough! What you should not do is to measure teams on financial business outcomes and then throw in good ideas on top, that way usually neither gets done.

Real experiment – De Lounge pt 2

Second post about our experiment with a premium gaming center. 

The experiment

To gauge the interest for our idea we decided to run with a sort of “smoke test”. We set up a landing page that looked like an announcement page for our new venture and described it in a way that sounded like it was already underway and planned to launch. 

We described: 1: our USPs, 2: our philosophy and 3: our membership options. The USPs and philosophy are quite straight forward, however the memberships served to gauge the interest of our subscription business model. We believed that to get profitability we would need another way than just pay per hour to charge for our experience. The subscription was inspired by gym memberships where you pay a monthly fee with several tiers. 

To make our target group find the landing page we bought Facebook ads for a value of ~$50. On the marketing page we then used call to actions “read more”. The buttons lead the user to a google form where we encouraged them to enter their details if they were interested in the service. We also asked them to choose which of the subscription options they would be interested in. 

The USPs: only people over 23, clean and a way to book seats
The philosophy “No mayo”, jokingly summarizing our concept 
Our membership options
Facebook ad

The Outcome

After spending our ~$50 we could see that the campaign had been quite successful and that there was indeed a large interest for this type of experience. Per click we paid an average of 1 SEK indicating that our message was attractive to our user group. We managed to get 60 people “signing-up”, i.e. entering their email in the form. 

We also found interest in our memberships around two thirds indicated interest in our middle subscription option whereas around a third where interested in our most expensive option.

Average of 9 Sek per email address
People generally interested in subscriptions


After looking over the results from our experiment we concluded that there was indeed an interest for this typ of experience. However we chose not to pursue it. This was because of the business case that showed that even if we would be able to drive people to our memberships it would still be a low margin business and one that would likely be hard to scale given how niche it is.

All in all this was a really fun experiment and also taught us a lot about setting up landing pages and driving traffic via simple facebook ads.

Real experiment – deLounge pt 1

My product strategy brain has been a little absent lately so I will use this post to describe a more concrete product experiment I ran together with a friend earlier this fall.

The problem

So both me and my friend are casual PC gamers since childhood but our current living conditions (sharing small apartments with a partner) don’t allow for a large gaming rig standing in the living room. As such, we frequent one of our city’s largest and most popular gaming arenas. This is nice since it allows us to play the latest game titles without having a gaming computer at home. This, however, is about the only nice thing about the gaming arena. 

The keyboards are sticky, the atmosphere noisy, computers often don’t work. It’s not unusual that we spend 30 minutes just finding two computers that work next to each other. On the flip side, the experience is very cheap and you can buy 20 hours of gameplay (lasting us several months) for the price of a burger and a couple of beers . We wanted an experience where we could game in a more grown-up setting and where higher prices would be ok given that the overall experience was better.

Desired learning

We thought we were not the only ones that thought this way. Gaming as an interest is becoming more and more common in older age groups. We believed that there is a cohort of gamers that would be willing to pay more for a “premium experience”.

Our main desired learning for the experiment was mainly to generally gauge the interest. Where we the only ones that thought this way? We also wanted to see if there was a way to monetize the experience in a different way and make it more economically viable than just charging a per hour price. 

In the next post I will go through the experiment we designed and what we learned, both around the experiment but also from a met perspective about running similar experiments.

Product idea! &

I’m taking a break from Product strategy discussion this week and am trying out a new type of post today – “Product Idea” where I describe an idea or a project I’m working on. 

One of my current side projects is learning web development with React. I follow along a Udemy course which so far has been a great way for me to learn but my goal is to create my own projects and web applications. So I have been thinking a bit on fun projects to try.

I’m currently reading David Allen’s productivity classic “Getting things done” and got inspired by the process he lays out for capturing and then clarifying (sorting) “stuff”. One of Allen’s main points is that our brains are not made for holding information. Our brains are good at creating new ideas and concepts. Therefore, using your brain to remember things is very inefficient. You both run a high risk of forgetting things and trying to remember the things takes up a lot of the brain’s capacity. 

So instead of remembering stuff you should capture the stuff in a trusted system. A trusted system can be any system (analog or digital) where you have the appropriate routine to follow up and sorting through. I’ve been trying out different products for this purpose during the last weeks (Google Keep, one note, mailing, iPhone notes) but I find that all of them both want you to take notes and organize notes in the same app. This has the effect that before you can actually start typing a note you have to make a couple of taps / clicks. This becomes a bit cumbersome when you are walking and want to jot something down that just crossed your mind.

Product idea 1: “”

Product vision: The easiest way to get stuff from your head and into digital form

Description: This would be an app (initially web) with only one feature. When you open the app there’s a blank screen. Tap it once and start typing to take notes. The notes are continuously saved (compare how Google Docs works) so even if you just close down your browser the notes are saved.

What it’s not: What’s more interesting is what this product is not. This is not a product for managing notes. In fact you can’t even access your note. When you open this app there is only one thing you can do: getting stuff from your head into digital form. 

Product idea 2: “”

Product vision: The best tool for sorting through your notes

Description: This is the important sister – app to This is where all your notes end up. The main view is intray looking similar to the tinder interface but instead of potential partners presents your notes one a the time and makes it fun to sort them by quickly swiping through them. 

I will report back here with any progress I make on this project!