Hello experiMENTAList, it’s Sundar 👋

I’m a former Head of Marketing Science at Uber where I optimized $1Bn+ in spend across Brand, Performance, and Lifecycle. Now, I share weekly playbooks that help you prove and scale your Marketing ROI.

Now let’s get experiMENTAL!

This guide distills 10 AI strategies from industry leaders that are transforming marketing.

Outpace your competitors by mastering AI.

Get Your Free Guide

The world of Marketing is supposed to be data driven and, yet, a common complaint that I hear from both Marketing leaders and Marketing Data Science teams is that companies just say it but don’t mean it.

It’s the data equivalent of Greenwashing. Let’s call it Datawashing.

Datawashing happens when teams mistake actions for outcomes, progress, and impact. What most people fail to realize is that there are 3 types of analyses, but they often only focus on one. They get stuck in a loop of high effort, low impact work.

Borrowing terms from Cartography (study of maps), here are the 3 different types of analyses:

This type of analysis, often done weekly and sometimes daily, plots exactly what happened and where you've been. This is where teams spend 90% of their time, but that’s a big mistake for 3 reasons:

First, by nature, it creates a reactionary environment. Think about the last time you were in a Monday meeting and leadership casually throws out, "Okay, but why did that happen?" It's a fair question, but think about the energy in the room that that creates.

The meeting often goes silent. Someone usually pipes up and says, "We'll circle back or get back on it." They send out a request. The data science/analytics team goes into a beehive of activity. Two days later, you have an answer. One day later, everyone forgets it and moves on.

Second, the answer is often not actionable. When you figure out why something moved, it's because you took an action that caused that result, or it was out of your control.

In both cases, you can't really do much about it. When things break, the teams that broke it should be on it. When things are going right, you already took the action to make it go up. Continuously digging every week on what happened doesn't lead to action.

Again, how many times have you been a WBR where you look at numbers but it doesn’t feel like any actions are being taken?

Third, simple questions are not simple answers. As the person that's been on the other side of requests like this, it was in my nature to be as exhaustive as possible.

Which means I’d dig and see if the changes were caused by:

That’s a lot of work in a little time. If you think about the normal cadence of activity:

Hmm, I wonder why Data Science teams aren’t pulling more insights 🫢

Topography is the study of forms and features of land surfaces. It’s where you begin to plot the details and the relative size of things to each other. But, don't go chasing waterfalls. Please stick to the rivers and the lakes that you're used to (if you don’t get it, these are lyrics from the song Waterfalls by TLC).

This second type of analysis is where the fun starts and it’s what most teams imagine their data science teams should be doing. This is where you find your nuggets of impact. At a high level, these activities include:

Here are some examples of analyses:

Unfortunately, teams at best get 40% of their time to do this (Thursday/Friday example above) and realistically maybe 20% because there’s some other fire drill.

This type of analysis helps answer the question of “Where should we go?” Navigational analyses should be run 1x a quarter during planning cycles to shape the strategy of the business and Marketing.

Example

You want to determine which Market to start ramping up Marketing in. This has profound impact on budgets and the business, but data science teams rarely have the opportunity to take a step back and run this type of analysis.

Without it, 4 months later, the team digs into it and realizes the country they picked has poor LTV and they could have gone with another country.

Navigational analyses set up big bets and identify opportunities for scale.

Teams do this about 0% of the time because Data Science teams are not invited to planning or there’s no space for this strategic type of work.

So, what should you do to fix this?

The key to stop Datawashing is not to do something new. It’s to do the 3 analyses in different frequencies.

Normally:

Instead:

By shifting your resources to higher impact activities (Topgraphical and Navigational), you surprise surprise create higher impact. It also has another positive loop in that your team will be happier, attract better talent, and create more impact which leads to them being happier and so on.

Easier said than done, but here are some ways you can create an environment that allows for this type of strategic work:

That’s it for this article! If you’re a Marketing or Data Science leader struggling to bring out the best in your DS team, let’s see if I can help.

That’s it for this week!

Stay experiMENTAL,

Sundar