4 Steps for Founders to Start Thinking Like a Data Scientist

Beginning your startup’s data journey when you don’t know how.

There are steps you can take to embark on your data journey — for free and without a data scientist.

So let’s role play:

You’re a startup founder who is about a year into the adventure. Things are going well, the product is gaining some momentum, and you’re constantly refining your long-term play for growth and possible (fingers crossed!) acquisition someday. You and your co-founder are spread more thinly now than you were a few months ago, but so far that’s ok because the marketing, sales, and even product development tasks are totally in your wheelhouse.

It’s why you embarked on the adventure in the first place!

But…the whole “big data”, analytics, and machine learning things you keep reading about, yeah, they’re not in your wheelhouse.

You know that you need to have a data angle of some kind because you know that prospective clients are asking about it. But you’re not sure what to do, how to do it, or even where to start.

• Option: Hire a data scientist? Maybe, but definitely not yet. (And even when you do, it’s not so easy. Slow clap for Monica Rogati)
• Option: Contract out the analysis? Maybe, but what do you even want them to do
• Option: Bring it in-house and build your own tools, dashboards, etc.? Maybe someday, but that’s a big commitment and you’re probably better served spending your time and resources elsewhere for now.

So what’s left? Below are 4 steps you can take — for free, and for even the least data-minded of you out there — to start thinking like a data scientist and move your company forward on your data journey.

1. Create a question landscape
2. Make your own glossary
3. Bridge the UX gap
4. Build a data roadmap

NOTE: Data scientists do WAY more than these 4 things, and this post is by no means meant to diminish the work of the professionals. It’s simply some encouragement for non-data-minded people to take a first step, or four, without expending a ton of resources.

Creating a Question Landscape:

Creating a question landscape will help you identify what you already know, and anticipate what you will need to know in the future.

If the purpose of using data is to be able to answer your own questions, or those of your clients or investors, then anticipating what the questions will be might help you determine where to go when it comes to fleshing out your data strategy. One way to start is by listing out a bunch of questions you might get and putting them into three categories:

What questions could you answer right now?

What questions could you answer if you did a little digging with your current data?

What questions can’t you answer because you don’t have the data yet?

For example:

• “How many downloads did you have in the past 30 days?” might fall into the first category.
• “What are the age demographics of your most frequent users in the past 30 days?” might fall into the second.
• And “What is the average session length of your top and bottom quartile of users?” might fall into the third.

The purpose of creating this landscape — some of which is charted territory, some of which is uncharted — is to say,

“Ok, we can answer this, this, and this, and we know (or at least are pretty sure) that eventually we’re going to need to answer this, this, and this…so let’s start to strategize how we need to get there.”

“Getting there” could be backing your way into the info using current data, or you might need to bake it into your product now so that you can get that information later. But at least having the conversation and anticipating what questions you will have to answer will put you ahead.

Making a Glossary

Let me ask you a question: what is your definition of a daily active user?

Now, what if I asked everyone in your company — would they have the same exact answer? (No cheating!)

While it might seem obvious, creating a glossary and/or adding documentation to your reports is really important. Everyone in the company might answer the DAU question the same right now, but what happens when your company is 30 people? 50? 100? Do you really trust that they’ll all be on the same page when you have dozens of clarity metrics that drive your product development and business strategy?

The value to this kind of organization is limitless, and while it seems like something that can be put on the back burner for later, the sooner you start documenting and the more disciplined you can be about it, the less likely it is that something will come back to bite you later on.

Bridging the UX Gap

Bridge the UXr and UXi gap by using data and then you can make educated product decisions.

All product people have an ideal UX in their head. It’s the utopia of people interacting with your product in exactly the way you want, and ideally your UX design is robust enough to lead them there — or at least lead them close — most of the time.

But no matter how good your UX design is, there is a UX reality (UXr) that is likely different than your ideal UX (UXi).

How do you nudge your customers from their current UXr to your UXi?

Use data.

Let’s say you don’t have the conversion rates you want — what is it about the churn that can give you insights into how to increase conversions? Where are users dropping out? Is it all users or just a particular cohort?

Maybe you already have data to help answer those questions, or maybe you need to get it (see “Creating a Question Landscape” above), but don’t guess — use data to bridge your UXr and UXi gap and make educated product decisions.

Building a Data Roadmap

A data roadmap is essentially the same as a product roadmap — it’s a tool for communicating direction and progress to internal teams and external stakeholders. I suggest taking a day, or at least half a day, to sit down with your team in front of a whiteboard (preferably off-site) and think really hard about what role data will play in your company. Whether it’s core or peripheral, it’s going to be there and you need to have a guide on how to manage all the associated tasks along the way.

In this post I spoke about how building a data roadmap can answer a bunch of questions that will help your company grow. But if those are too vague for you at this point, start by completing the three tasks listed above: list a bunch of questions you can and want to answer, make a glossary of metric definitions, and identify the gap between the UXr and UXi of your product.

10 questions that your data roadmap can answer

At least then you will have some internal marching orders for what to do next and how to do it. And then, when you’re ready to contract out analysis or even build your own data team, you’ll have a good idea of the goals and what the value will be when they deliver.