๐Ÿฅ— Plating counts – Food for thought

I’v been thinking a bit about how to explain a certain problem area, which I think many companies are stuck at today (including me). The problem domain I’m talking about is going from being good at collecting data to actually making it into real customer value. I think it is a bit of a paradigm shift, we have been fed with the fact that becoming a data driven company is crucial for survival. Which I truly believe as well. We start collecting data, we start with what we can and what we understand. We become better at it, the data warehouse grows along side data departments. But it doesn’t necessarily means that the data you’re gathering can be acted upon. This is what I call high level of Actability. Actability meaning the percentage of data you actually create customer/business value from. Or having a high Traceability, meaning the percentage of features, function and services you can actually trace back to an insight.

I think this might be because we’re sloppy when it comes to analysis and sloppy about the definition of insights. We have our business data portals, our insights repositories, and other services which serves us data, the thing is that it does not provide us with analysis or insights. We log into them and look for charts and datapoint which has the highest probability to validate your assumptions. I don’t think we do this intentionally, I think we do it because we’re human. And this is the problem domain I have been trying to find a good analogy for.

So let me try to explain it with other words. Imagine the process from growing a seed, into a vegetable, to harvest it, to cooking it, to finally plating it.

๐ŸŒฑ The planting phase:

We plant seeds to get vegetables. We decide what vegetables we want, and what we like. It’s the same for data, we start understanding what data we need to collect. So we start implementing what is needed to feed the data warehouse with some veggies. We can go for quantity, we measure everything, and then decide what data we actually want. It can be fruitful and opens up for experimentation, but it can also be a bit costly. It takes time, effort and cost money to collect and store a lot of data. We can also before try to understand what questions we want to answer by the seeds we’re planting. This makes it easier to have high Actability because you don’t have as much data. I don’t think any of it is wrong, it is just different approaches and I think it is good to understand the pro’s and con’s of it.

๐Ÿง‘๐Ÿปโ€๐ŸŒพ The harvest phase

It is time to pick your vegetables. They look great, the taste ok, it will fill you up but not make you satisfied. You can eat raw carrots, even raw potatoes but it doesn’t really makes it for you. So what can you do?

๐Ÿ”ช The cooking phase

You can refine them, you can combine them and you can make something satisfying with them (if you know how). This is the phase where I think many companies struggles. They don’t know which veggies goes well together, and how to refine them without destroying them. The same goes for data, it is hard to know which datapoints goes well together with other to make sense, to create insights, to make you smarter, full and satisfied. I don’t have an answer to this, expect trial and error.

๐Ÿซต๐Ÿผ but some pointers to have in mind

  1. Define what an insight is at your company
  2. Work on your insights repository, it should be easy to search, easy to track, and not to hard to digest (no raw broccoli)
  3. Get som good pot and pans, but experiment what works for you. Excel files, Confluence, Jira, Google sheet, Notion, Airtable.
  4. It is not molecular dishes for foodies, it is food for farmers. So dare to have it raw, and gradually refine it.
  5. The result does not become better than the crops you grow. So the fact that the quality of the crops makes the dish is not a lie here either. So the quality of the data is of essence.

๐Ÿฅ— Plating it:

Now you have come to the satisfying part, you hav cooked the food, it can be indulged with good result by you, but can it be appreciated by someone else? And if other people does not appreciate it and does not want to eat it, why have you then cooked it? We underestimate how we visualise the data. It lies a great deal in how we visualise the data if it will be acted upon and even understood. So do not underestimate the plating, it might even be a dealbreaker to get over the “we are a data gathering company” to becoming a “we are a data driven company”.

๐Ÿซต๐Ÿผ and some pointers to have in mind

  1. Minimalistic plating has never been better. Make the main crops the focus . Otherwise we get lost in the sallad. Make the key take aways the focus for the visualisation. It is better that the audience take one thing with them and act upon it, rather than trying to remember it all
  2. Try different ways of plating it, UX the shit of it. Ask people how they interpret and understands it.
  3. Set goals for you visualisation. What do you want to achieve with your dish and your plating.

This if of course very broad, because we often have so much data from various sources, quantitive and qualitative. And I guess that is one of the reasons why it fast becomes quite complicated.

Don’t forget to make a plan that stretches from planting to plating. Especially think about how you will maintain it? routines, cadences, ways of working and processes. It does not have to be complicated but it is necessary to be able to have data with good quality and make smart and efficient data driven decisions a habit rather than sporadic activities.

The two Cooking and Plating phase is where I have been experimenting the most these last few years. The analysis and the visualisation part. And it is so much to be done there, really exiting stuff.

How do we create automated User Journeys combined with Qualitative and Quantitative data?

How can we in one view offer the “What” is happening but also the “Why” is happening?

I’m thinking text analysis, sentiment analysis of qualitative corpuses (text datasets), and so on. That could also serve as a insights repository further on, but with much better plating than Excel can offer.

Who is with me ๐Ÿซต๐Ÿผ โœ‹๐Ÿผ