As I was brainstorming what to share today, I started with the same core principles I do for every article of “What will be most useful to my readers?”. With that lens, I came up three main questions that I thought would be fun to answer that help frame the story of how I built Uber’s Brand Science team:

Andiamo!

I moved to Amsterdam in October of 2019 and became the Global Head in May of 2020. In those 6 months, there were so many pieces that needed to come together, but these types of organizational decisions rarely exist in a vacuum.

There were 4 factors that made the perfect storm:

To put it mildly, Uber’s IPO was a disaster with the stock falling the first day and basically going down almost 50% over 6 months. The knock against Uber had always been “Sure you have hyper growth but can you do it sustainably” .

Under that lens of scrutiny, Brand Marketing spend was a BILLION dollar line item so naturally there were questions around the impact and ROI of it.

Brand Marketing was in the cross hairs.

Rewinding a bit, I first shared that I wanted to move to Amsterdam with my boss in the Spring of 2019 as my wife was preparing to graduate business school and we wanted to try Europe before settling in the States (spoiler alert…6 years later we’re still in Amsterdam and not going back).

What happened next was pure chaos:

April 2019 → I applied for a Senior Marketing Data Science role, interviewed, and got the offer

May 2019 → Then we were hit with a Marketing reorg where all offers were paused.

June 2019 → My offer is officially rescinded. I was devastated!

But sometimes you just need luck on your side.

July 2019 → My skip level manager, as part of the reorg, becomes the Global Head of Marketing Science. We had a really great relationship.

August 2019 → They decided to send me to Amsterdam as the new EMEA Head of Marketing Data Science. I went from going as a Senior IC to a Manager overnight.

In addition, over the next 6 months, both the Asia and LATAM lead decided to leave Uber and things were going well enough that I was asked to take over those teams.

Somethings you can control. Some you can’t.

When I first started as the EMEA Head of Marketing Science, I thought I’d be working with the Performance Marketers because the Marketing Science team was also under Performance Marketing. Naturally they’d be my stakeholders. Duh. But they already had support from Central HQ or other contractors.

Turns out there was a whole other half of Marketers (the Brand Marketers) that had no analytical support. The crazy part was I didn’t even know these Marketers existed. I’d never talked to them or even heard of this part of the organization. Imagine that… you’re in a Marketing org in a company and discover another set of Marketers. It felt like Parent Trap where I discovered my long lost twins!

As I’d learn, these Marketers were spending a lot of $$ and had a lot of impact on the business so there was a huge Analytical desert just waiting for an oasis. I brought this up with my manager and told them that there was this huge opportunity for us to be able to support Brand Marketers with work that would be extremely impactful.

By 2019, Uber was a clear household name and had made bets in Uber Rides, EATS, Freight, and numerous other markets. From a competitive sense, Uber had exited any markets where we weren’t 1-2 which meant we only stayed in competitive markets.

At the same time, we were tapped out on paid marketing in many markets. Many of our core markets we had been in for over 9 years and with a name like Uber that everyone knows, Performance Marketing has strong diminishing returns.

Sad side note: The only place we were getting beat badly was in the US on Delivery. DoorDash was crushing it. 😢

So, the next big bet was on creating and maintaining a unified global brand and we brought in a new VP of Brand leader from Google to build a unified Brand Marketing team at the beginning of 2020. They’d sit complementary to the VP of Perf Marketing and take Uber to the next level. Coming from Google, this new leader was extremely data driven which meant they identified a need for a global brand science team.

How convenient… I saw there was a need. Some one else saw there was a need. And we filled that need.

In most situations when you’re inheriting a team, it’s quite messy. First, Uber had six different and distinct regions:

This resulted in a variety of different methodologies, assumptions, and even trust in how things were being measured. In addition, most markets were severely understaffed with effectively no brand scientist meaning teams would use whatever methodology they wanted to measure their campaigns.

From a resourcing perspective, Brand Science was super barebones. The reorg in 2019 had decimated the number of analysts outside the US. Here’s what that looked like:

So, as you can see and as I mentioned, there were no Brand Scientist 🙂.

Lastly, Uber was a very ops-heavy organization and operations teams have often owned the P&L. This mean that they often looked at Brand Marketing as a waste of money and that they could better deploy that budget elsewhere. 

In summary, a mess of an organization, under pressure to prove impact and a need for global alignment. To be honest, these are the types of challenges that I really enjoy. Data science is a strategic function so being dropped into a strategic clusterfuck is as exciting as it can get (if you like this stuff).

Below I’ll share the carefully orchestrated set of chess moves I used to rise to the top and expand my analytics empire in a bid for world domination.

Just kidding, but yeah here’s what I did as I was still figuring out how to become a Senior leader working on a type of Marketing I had zero experience with. If you’re not uncomfortable, you’re not growing.

After about a month in Amsterdam, I knew that Brand Science in EMEA was the biggest place I could have an impact. I mean, we’re talking about about ✅ hundreds of millions of budget, ✅ world class marketers as stakeholders, and ✅ complex measurement challenges.

So, I pitched to my manager that I should fully support Brand Marketers.

🟢 I was given the green light.

From experience, I knew that Data Science is a very slipper slope. You’re there to prove ROI (which sounds like a good thing), but you’re often the blocker for many things including needing to have more robust planning steps and create holdout groups, blah blah. I needed the Brand Marketers to trust that I was their partner and not an agent of Performance Marketing set out to eliminate their jobs.

So, I went on a bit of a roadshow selling myself to 3 audiences:

The Brand Marketers were Marketing Leaders that I had never come across, worked with, or heard of . The challenge here is that the inverse is also true. They've never came across, worked, or heard of me. So, I set up time with as many of them as I could to simply learn. I used this opportunity to truly understand:

The last one is crucial. If you’re ever entering a new team, understand the preexisting notions they have about you. Some of them had never worked with a data science team while others had worked with world class data scientist at their previous companies. Knowing everyone’s expectations is a MUST if you want to live up to them.

Turns out Brand Marketing is significantly more complex than I thought. There were so many more parties involved including PMMs, Media Buyers, Marketing Ops, and a few other functions I hadn't interacted with before. Also, the range of experience with Marketing Science was all over the place. Some thought Pre / Post was a valid measurement strategy while others knew that you can’t A/B test a national TV campaign. Having such a wide array of expectations was a huge hurdle.

On the Local Operations side, the goal was to understand their concerns as well. I was able to share my previous experiences in Performance Marketing and help alleviate the concerns of wasted spend. In addition, I was able to support in the education and advocacy that Brand Marketing was the only path forward to the increased growth we were looking for (it was surprisingly easy and quick for me to pick this fact up given my experience with diminishing returns of Perf Marketing). Finance had very similar concerns to Ops, so it was the same sort of spiel.

For me, it was about introducing myself as a partner and creating a relationship.

🟢 The goal was to build trust and put a face to the analytics name

The 3rd key decision I made was hiring. I inherited a team of 2 analysts in EMEA, zero from ANZ, and two more in LATAM. To be honest, it made zero sense on how we were resourced. I felt that we were understaffed compared to the impact we were trying to measure.

So, I put together a hiring and resourcing plan that went all the way up to our VP of Performance Marketing to show that we were way overinvested on how many analysts we had for Performance Marketing vs what we had for Brand Marketing. The goal was to show an analyst-per-dollar measured ratio, and the ratio was heavily skewed towards Performance Marketing.

Here was the proposed org structure:

In addition, we already had a team of contractors in India that I was able to partner with to keep the team lean but effective.

Here’s how the hiring played out:

The additional analyst in EMEA was immediately green lit but that’s because I had already had 3 analysts (and one left) so it was just a backfill. The 2 in LATAM I inherited and we weren’t going to add anymore.

The 3 in the US were approved because that’s also where Brand Leadership sits and you need to make sure they’re supported. So, over a period of 6 months, I was able to hire 3-4 analysts and onboard 5 contractors giving me the coverage and skill set I could survive (and maybe even thrive) with.

🟢 Build up a dope squad

One of the things Brand Marketers felt was a lack of continuity and support. On the flip side, one of the struggles I've always had as an analyst was feeling like I'm a part of the business. I killed two birds with one stone.

Each analyst was going to be embedded to a two regions and they were the sole owner of those regions. I was not going to be the type of leader where decisions had to be run by me. You have autonomy and independence and it’s my job to support you when you feel out of your comfort zone.

Example

For example, I had 3 analysts to cover EMEA, APAC, ANZ, and India.

Within those markets, India and Australia were big ones. Within EMEA, you would have the UK and France being big. So, I divvied up the regions amongst the analysts ensuring there was a balance of larger and smaller markets.

Analyst 1 → Northern + Eastern Europe + India

Analyst 2 → South West Europe + ANZ

Analyst 3 → Middle East + Africa + Rest of APAC

In LATAM, there were 2 major markets: Mexico and Brazil.

So, there I split by country ensuring each analyst only had 1 major country + a handful of other countries balancing the workload.

Analyst 1 → Brazil + Colombia

Analyst 2 → Mexico + Chile + Argentina

🟢 Team has autonomy, ownership, and clear resourcing.

The final piece of the puzzle was implementing a new planning and prioritization process. We had a finite amount resources to cover all Uber and Uber EATS campaigns globally.

The planning process at first started rough because the pendulum swung too quickly to the other side. Too many Brand Marketers had now switched to this idea of needing to prove ROI because they were being asked to and wanted to consume all of the analytical resources possible.

The challenge is that not all campaigns were large enough to be measured and not all campaigns should be measured. And so the planning process was my way of streamlining that.

To solve that, every half (and then revisit quarterly) we would sit down and figure out how many campaigns we could resource. Roughly, I knew every analyst could generally do about 1-2 big campaigns and 2-3 smaller campaigns per quarter.

We would then go campaign by campaign and say whether we could measure it. This would first get aligned with the head of the country and then region ensuring that the final decision was made by Brand Marketing leadership and not me. Can’t blame me for something I didn’t do 😉 . Just kidding, but seriously. When you’re in a prioritization exercise, it’s best to have the priorities finally decided by the person who the team reports to.

So, at the start of every quarter and half, every country Marketing lead would know which campaigns were getting analytical support. And the countries that were not supported were obviously frustrated, but they eventually understood it and were able to set clear expectations. It also forced them to think of clever ways to prove ROI (of course with our help and guidance).

The truth was these smaller markets were hard to causally measure. They either didn't have another spend or cities so we couldn’t build synthetic controls. So, these markets had to fend for themselves. Sorry smaller markets ☹

🟢 Planned and priorities were aligned!

And the final piece of the puzzle was the global measurement framework, but that’s for a whole other article. And that’s the story of how I built Uber’s Global Brand Science team!

Stay experiMENTAL,

Sundar

Here it is: The 4 investments that unlock marketing efficiency