Welcome to this edition of thought leadership, where you can discover the latest eCommerce insights direct from industry experts. In this article, Thomas Pantle, Product Manager at FACT-Finder, shares his eCommerce Kata – a practical and powerful framework for turning small steps into big results.
eCommerce has all the data at hand to create the perfect shopping experience and optimize for pretty much anything, yet sometimes it seems hard to make real progress on your goals. To help people working in eCommerce solve their problems, I have created a framework that is derived from the Product Kata by Melissa Perri and the Toyota Kata by Mike Rother.
What is a kata?
The term ‘kata’ means a structured way of doing things or practicing. The approach, originally used in Japanese martial arts, encourages learners to spend much of their time practicing drills to improve specific moves. By repeating a movement on a consistent basis, the correct technique is gradually mastered. This kata definition applied to mainstream business uses scientific thinking to train the daily skills individuals need for rapid, incremental, continuous improvements.
Why do you need something like the eCommerce Kata to structure your work?
By using the eCommerce Kata, you will make progress towards your goal in a scientific and proven way. This helps you in reaching your goals, as well as your manager in tracking the team’s progress towards the goals and finding ways to support them. Since you’re making sure to always move forward towards your goal, it is a great opportunity to create alignment throughout your team and company.
The eCommerce Kata will help your business tremendously, as it is focused on learning rapidly and inexpensively. Thus, you will just do what is necessary, keep exactly the things that are proven to work and trash everything else. Finally, the eCommerce Kata can also help a larger team with collaboration, not only in terms of alignment but also by creating a knowledge base of learned lessons and by creating visibility within the team about what’s being worked on.
How does the eCommerce Kata work?
1) Company goal, eCommerce KPI, Future State
The first step is to make sure you know where you’re heading. This could be something huge, a goal that is 5-10 years out in the future, but it could also be something smaller. Just make sure that it isn’t too small, otherwise, the eCommerce Kata will lose its value, as the road towards the goal could be clear anyways. This major goal could be the company vision or something that is derived from it as a sub-goal. Depending on your position it could either be defined in collaboration with your manager, or you are defining your goals in collaboration with other managers, to make sure you’re aligned with the company goals. Briefly, this step should be to answer the question ‘where are we heading to?’
2) What are our users doing now?
This second step must be done after the first step for a specific reason. There are likely a lot of things that your users are doing, but most of it is likely not related to your goal from step one. In this step, you want to make sure to accurately capture the current situation, as concise as possible. That usually means investigating some KPIs, maybe just one if your goal is already related to a certain KPI. It could also involve doing qualitative analysis by talking to your users, customer support or someone else.
3) What is the first small goal?
Here you’re starting to plan your actions. The important outcome of this step is to create a reasonable hypothesis of what you need to do to improve the current condition towards the target condition. This is usually the step that requires the most ‘hard-thinking work’, as you need to find a possible solution that solves the customer’s problem, relates to your goal, works within the specific constraints of your company and is measurable. This is also a great step to harness the knowledge and ideas of a team, as you can collaboratively create a good hypothesis. At the end of the step, you should have well-defined steps and expectations of the results.
4) User research, Experimentation
Finally, you get to work on the current hypothesis. The key to this step is to make sure experimenting is done in an inexpensive and quick way. That means if your pre-defined step was to improve your checkout process, you shouldn’t improve your checkout process from the get-go. The right thing to do here is to prove if improving the checkout process does affect the current condition in the right way (towards your goal). So ideally you want to find a way to not involve any coding, but ‘fake’ the frontend while running your experiment. That way, you usually require more manual labor, or your solution isn’t quite scalable, but our aim for this step is just to get information, not to make everything perfect.
Finally, after completing these steps you have to check whether you’ve reached your goal and if you didn’t, just rinse and repeat until you do.
It’s important to notice that we’re not actually doing anything in steps 1-3, and in step 4 we’re still not doing things in production quality. That’s what makes the eCommerce Kata so powerful since you’re making as much progress towards your goals as you can while straining as few resources as possible.
That seems pretty abstract to me, how about an example?
Sure. I’ve created a spreadsheet template in which I track the katas that I’m doing. It begins with the goal that I’m working on, usually something that has been set by leadership and/or in collaboration with my manager. In this fictional example, the goal is to increase the AOV since it has stagnated in the past few years.
After that, my manager and I have defined a reasonable in-between goal that affects the overall goal, in this case, more than 60% of the customers have an order value that is above €40, so they’ll receive free shipping. This includes a somewhat predefined strategy from my manager as we’ll try to use the free shipping to leverage our AOV. It could also be less specific, e.g., increase AOV by 20%.
From there, I start to fill in the columns from left to right. The first one always captures the current condition regarding our (intermediate) goal, which is in this case the percentage of customers having an AOV over €40.
Next up is our current obstacle. This is the most difficult step in practice, as you need to be honest to yourself and critically think about your problems. It’s especially hard if you have been working on the same topic before, maybe because you’ve designed the specific page that you’re about to identify as the current problem. So, it is really important to practice detachment and think about your shop and user experience objectively. Usually, in the first few iterations of the eCommerce Kata, the obstacle will be a learning topic, which basically means there are things that you don’t know, that may or may not relate to the target. In the first few iterations it is your main job to figure these things out, and especially figure out what isn’t really affecting the goal. In our example, the first obstacle is the fact that we don’t know why customers with <€40 order value aren’t buying enough to get free shipping. We included the hypothesis that €40 could be too high for them to make free shipping worthwhile.
After that my favorite step follows: Defining the immediate next step that we can take to tackle the current obstacle. This gives you the opportunity to apply creativity and collaboration since you can fully include your team or consult some colleagues if you can’t come up with a solution yourself. At this point in the eCommerce Kata is also the only chance for innovation, since you’re proposing a solution here (Note that proposing a solution isn’t proposing the right solution, if a proposed solution doesn’t work, it has to be discarded). If we previously included some kind of hypothesis in our obstacle step we need to find a way that proves or disproves this hypothesis. If we didn’t do this, we also have to define the hypothesis in this step. Both are possible since these two steps are usually done in quick succession. Coming back to our example, a possible step to tackle our obstacle could be to survey a small base of our previous customers about the topic. Two important things here: First, this is a qualitative step but depending on the situation it could also be a quantitative step (e.g., looking into analytics data). Second, make sure whatever you’re doing is cheap and fast. Cheap means, it should involve minimal effort, so if you don’t have a surveying tool of any kind in action, your step shouldn’t involve surveying. Fast means the time until you learned enough to reiterate the kata and define the next obstacle/step should be roughly a week.
Now that you have a clearly defined obstacle and step, it’s time to extract the essence of your hypothesis. In the ‘expected’ column we keep note of what would happen if our hypothesis was 100% correct. That’s it. Keep it for later, it will help remind you about the lessons learned. In our example, the expected outcome is to find out, how many customers thought €40 is too much for free shipping.
Finally, after you waited for the data to trickle in, you need to keep a record of what really happened. For that, we have the last column, ‘learned’. These last 2 columns give us a nice overview of what happened in the iterations we’ve taken so far and can thus inform future decisions. In our example, only 10% of the customers thought that €40 was too much for free shipping, which isn’t giving us enough leverage to increase the number of customers ordering over €40 from 40% to 60%. However, through the free-text field, another 10% of customers feedbacked us, that they didn’t even know about any free shipping in our shop, potentially crucial information that we can utilize in our next hypothesis.
From there we keep iterating until we think we’ve learned enough and understand the underlying problems. Then we switch our iterations to actions that we expect to change the condition and keep iterating them until we meet our goal. You can see in the example how it took us five iterations to get to our goal, but I have to be honest with you, that could still be on the faster side. No need to get frustrated though, since we set the maximum length of an iteration to about one week, you’ll still achieve something in just a few weeks, and many of the iterations can be done quicker.
That’s all the basics you need to know to get started. From here, it is mostly about you and your company practicing this way of working and learning a lot of things about your customer (that you likely don’t yet know about). If you have any questions on this topic, please feel free to contact me at firstname.lastname@example.org or via my LinkedIn profile.
About the author
Thomas Pantle is a Product Manager at FACT-Finder. He was formally trained as a management assistant for eCommerce, graduating with excellent results while simultaneously learning the ropes of product management. His most desired outcome is to facilitate others and enable them to deliver great results. Thomas is optimistic enough to always dream of perfect outcomes but realistic enough to chase after ‘good enough’ results. If he were to describe himself with three hashtags, they would be: #Strategist #CuriousLearner #RuthlessOutcomeFocus
For more eCommerce insights, connect with Thomas on LinkedIn.