👋 Hey, it’s Sundar! Thanks for reading experiMENTAL: my newsletter helping founders and marketers navigate the CRAZY world of consumer tech with secrets from 10+ years in Marketing at Uber & others.

In today’s newsletter, you’ll learn:

• why data driven is no longer enough for a competitive edge

• how Uber was data obsessed and how that unlocked hyper growth

• how you can get your company go from data driven → data obsessed

Why being data driven isn’t enough

The phrase data driven appeared in the early 2010s as companies collected more data and realized they could make more informed decisions from it.

But it’s so mainstream now that it’s in the Oxford Dictionary.

In 2024, being data driven is no longer enough.

With the pace of innovation across generative A.I., cloud services, and other contributing factors, people are able to replicate technologies faster than before.

Here’s a great example:

Klarna, a large enterprise fintech announced it was rolling off Salesforce and Workday, two popular and entrenched SaaS companies. That’s likely $50M+ in savings.

This is the first point in history where Klarna can be bold enough to make this decision whether it goes through with it or not.

This means that companies must go back to the root of every good business:

Understanding their customers.

A company’s advantage can’t be tech alone but an obsession with collecting and analyzing data to identify patterns.

To optimize every touch point and create better margins.

That’s why being data driven needs to be replaced with data obsession.

To understand the difference between driven and obsession, I’ll share how Uber operates at a level that only a handful of companies in the world are at.

How Uber enabled data obsession

Data obsession starts at the top.

And at the top of Uber was a visionary leader named Travis Kalanick.

I’m not going to get into the type of person he was or the controversies around him.

From a professional perspective, this man was a genius.

Yet, so many people look at him and think he had an unstoppable ego.

But to be data obsessed (or even data driven), you have to be ego less.

By definition, you let insights drive your decisions and not your own perspective.

I’ve never seen anyone be better at this than Travis Kalanick and his leadership team.

The formula at Uber was simple:

1. Hire smart people

2. Provide access to data

3. Make smarter decisions

4. Repeat

Here’s how Uber did that.

Data access

Mention Querybuilder to any ex Uber person and they’ll tell you how powerful it was.

Querybuilder was a web based interface that gave every person at Uber access to ALL of Uber’s data.

The genius was that every time you ran a query, it would create a new URL. This meant that you could share that URL and then all of a sudden you created virality for a SQL query.

Virality for a SQL query 🤯 .

It sounds stupid but imagine how quickly other people would tweak that query and use it and all of a sudden you have SQL literacy across the board.

The crazy part was that SQL wasn’t reserved for analysts and data scientists.

Operations, Marketing, Product, Growth, Engineering, and even HR had access to Querybuilder.

And to ensure that everyone knew SQL, Uber gave every single person training to SQL through a company called Vertabelo.

Instead of silos of analysts you have an entire army where everyone is an analyst.

Imagine the compounding impact of that.

Data investments

The chart above shows the ratio of Data Scientists to Data Engineer.

Uber has one of the highest at 8:1.

A bit counterintuitive but that’s a good thing because as companies become mature and data platforms more stable you can build more and more on top.

Effectively, Uber created such a stable foundation that you can scale how many data scientist build.

That creates efficiencies that only the most data obsessed companies in the world can achieve.

I was also part of the Martech and Adtech team at Uber and it’s best in class.

The team focused on automations and optimizations that would allow Uber to make microsecond decisions on $X,XXX,XXX,XXX of Marketing spend.

Those small % add up.

But, it was all built on top of data investments.

And data obsession.

Data literacy

30,000 Uber employees do 1 thing every week:

They look at a dashboard called Summary.

Why? Because people are expected to know their metrics (Read full LinkedIn post).

Every City, Country, and Regional P&L owner knows their business by heart and can recite last week’s numbers.

Conversations and discussions only start AFTER there was data to look at.

People were OBSESSED with Summary.

It was just a dashboard 🤣 .

In addition, one of the most popular meetings every month was led by the Strategy and Planning team where they shared the latest insights they found.

Imagine that.

A meeting about insights being one of the most popular.

It’s because people were OBSESSED with insights.

Seeing a theme of obsession? 😉 

They wanted any advantage they could find to improve the growth in their cities and create better customer experiences.

This might seem overwhelming and unachievable but I’m writing this article to help everyone realize that being data obsessed is not reserved for the .01%.

There’s no need for the growing data inequality gap.

Go from data driven → data obsessed

As a teaser for this article, I posted on LinkedIn about how Uber was not just data driven but data obsessed.

One of the comments was:

The question highlights a fundamental obstacle for most companies.

They think of data as an investment that requires proof of an ROI.

That’s the wrong way to think about it.

Every data obsessed company (think Meta, Netflix, Google, Uber, Amazon, etc.) starts with a top down POV that data is a necessity.

There’s nothing to justify.

Once you have a top down POV that creates the foundations of culture you can layer on people, strategy, tooling, and eventually better decisions.

Let’s examine a few actionable ways to do this

Interviews

To hire data obsessed people, ask one question in every interview “Tell me a time you had to influence a decision using data”

and follow up:

1. What data did you use?

2. How did you know that was enough data?

3. How did you influence others that your decision was right?

4. How did you communicate your decision?

What you’re looking for is passion for data.

EVERY single person has made a decision using data, but those that are data obsessed get passionate about talking about it.

They make you feel like you were in the room with them.

Bonus tip: Have a case study that involves unpacking data or do a SQL test for everyone like Uber did.

Dashboards

Make 1 dashboard that’s the source of truth.

It should be kept up to date and expected that people know the metrics there.

The Summary dashboard I mentioned above was exactly that at Uber.

It was simple:

• 8 high level metrics

• Preset look back windows

• Drill downs by 1 geo at a time

• Growth in absolute and % terms

That’s it. But it was the source of truth. It was reliable and up to date.

Over time, dashboards become a mechanism for data obsession instead of ending up in the graveyard like 99% of dashboards.

Meetings

Have a monthly insights meeting.

At this meeting, there should be only 1-2 insights presented.

But, they should feel “ground breaking”.

People should come away wanting to experiment or implement the recommendations.

The reason it worked at Uber is because people would then share back feedback from their experiments which made the insights better and created a feedback loop.

This monthly meeting also put an emphasis that insights need to be actionable.

This is not a place just for people to share their latest work.

It changes the relationship with insights.

A great example from Uber was a meeting where the Strategy and planning team shared the relationship between # of times a customer experienced surge and their retention rate.

That analysis changed the entire game between short term surge pricing and long term retention while so many Operations teams were focused on just short term marketplace health.

It fundamentally transformed the way teams operated. That’s powerful.

Documentation

Prioritize documentation (decks or docs) .

There’s a misconception that documentation slows things down.

I’ve found it to be the opposite because it forces:

1. Structure

2. Completeness

3. Record keeping

4. Knowledge sharing

At Uber, many were ex consultants.

They were insanely good at creating decks and it created a culture of deck building (not in the bad way).

I remember feeling very behind when I first got there but over time my narrative skills improved 100x. This led to better influencing and ultimately decision making.

Amazon does something similar with their 1 and 6 page memos:

Documentation is a forcing function to data obsession.

Recap

If you're looking to change your org, it starts with the culture.

Your people and processes have to scream data-obsessed.

Eventually you'll be so data driven you won't have to say you are.

That’s it for this week. I hope you now know a bit more about:

• Why being data driven isn’t enough to compete

• How Uber helped accelerate data obsession

• Actionable ways to become data obsessed

I’m going to on vacation (or holiday as they say in Europe) for the next two weeks and completely unplugged so no articles for the next 2 weeks! Try not to miss me cluttering your inbox too much.

👍️ Loved it? 👎️ Hated it? Let me know!