Uber claims that its app offers advertisers a ‘natural attention advantage’, and last month it published a report, in conjunction with Lumen Research, to prove it.
According to Driving Attention, Delivering Results Uber ads generate 6.6 times more ‘attentive seconds’ than other digital platforms, and they hold that attention for longer than the three seconds that most other media owners can manage.
Clearly, many brands have already got this message. At the end of the first quarter of 2025, Uber announced that its ad revenue was growing 60% year-on-year, and had now hit a $1.5bn annual run-rate.
But it’s still a young business, notes Paul Wright, Uber Advertising’s head of international, with plenty of room to grow and mature.
Over the course of three decades — and roles at Sky, Omnicom, Apple and Amazon — Wright has often been at the forefront of new technologies and changes in the ways in which advertising is bought and sold. Now, he believes the industry is waking up to the power of attention as a planning metric. In this interview, which has been edited for clarity, Wright discusses the critical moments that make ideal marketing opportunities.
Uber has recently partnered with Lumen for research that focuses on attention. Why have you made attention the focus of your advertising proposition?
It’s proving something we thought [we knew] already. We were already tracking stuff. The way we approach advertising is to embed ads into an existing platform, to not interrupt the user experience. If you think about your own Uber use… you’re waiting for a car and looking at the map to see where it is. Or waiting for a delivery, watching to see where the person is [in relation] to your house. We’ve always thought there was something powerful in that, but didn’t know exactly what.
When we first started looking, we noticed high brand recall and that viewable time spent in ads was pretty high. If we’re thinking about attention, that’s powerful, because in most media environments you’re scrolling through quickly. We went to Lumen and said, ‘Look, we think we’ve got something but need to be able to prove it.’
Did you learn anything in the research in terms of differences between Rides and Eats that you could apply?
There are slight differences. On Rides, there’s fairly large engagement time [with your journey]. The ad unit changes through different stages of the ride, which keeps interest up. We have ads where you start with a brand message and end with a wallet message, which encourages you to take action.
That’s always going to be different to Eats, where you’re looking at a post-checkout ad after you’ve made the transaction and are waiting for the delivery to arrive. You’re not going to look at it for minutes [at a time] while the delivery is coming. You’ll check [the app] in a different way.
So there’s a slight difference, which I think is more to do with format, time spent and consumer behaviour. But both platforms have strong attention metrics, which you can see in the study.
Is it fair to say Rides is more top-of-funnel, whereas Eats is closer to commerce?
Yeah. I think a lot of what we see from brands [in Rides] is more activations that would suggest that. We tend to get more transactional brand activity in the post-checkout area of Eats than we would do in Rides, which is more brand based. So yes, that’s backed up by advertiser behaviour.
Who is this message about attention for? Is it mainly agencies, or brands?
Advertising is in more places now than it’s ever been. How do you get consumer attention at the right time, and what does that mean? It’s tough, right? I don’t know how many messages a consumer gets these days. It’s a lot.
[For brands:] how do we capture that moment where a consumer is leaned in and thinking about buying our product? You’ve got to have attention.
I think agencies are beginning to realise attention as a potential planning metric. Moments… critical moments of the day and the week, and critical events in people’s lives; those things become very important.
The challenges with the programmatic world have often meant that we’ve been targeting devices rather than people. That, in itself, is a problem, because are you really paying for below-the-fold on some random website where there’s no attention? Whereas if you’re in an Uber you’ve got good attention. You can pretty much prove it is there. So there’s elements of that, in terms of efficiency for ad dollars.
There’s behaviour and context, too. Attention is all well and good, but in what context is that actually happening for the user?
Yeah, we believe that. If you look at some of the things we do, like the work with Pinterest targeting Taylor Swift concertgoers. Airports are another one. Everyone is going to go through duty-free.
We’ve done work with L’Oreal on targeting with Armani Code purchases. That waypoint data is very accurate, because we know exactly which terminal you’re going to.
We’re great believers in context and attention together, not one or the other.
Yes, to some degree you could put a random message out and get some attention, but for the consumer to be truly engaged it has to be strong. There’s a lot of evidence where, when we get the message and context right, it really just flies.
Has it become harder for Uber to stand out now that (as per Eric Seufert) ‘everything’s an ad network’?
I was on a panel talking about how retail media networks or commerce networks, depending on how you define us, stand out in a world where there’s a lot of noise.
I think we’re relatively unique. We have a well-known brand which is helpful, and because we have this higher brand metrics platform as well, then I think we do stand out for advertisers.
The other thing that stands out is the audience profile and the way we can understand audiences. In my career, I’ve spent more and more time collecting more and more data about audiences, in the vain hope that if we have enough data points we’ll know exactly what that person is going to do, which is generally not true.
What we’re now able to do is only take data points that are critical, like the Taylor Swift example. You know someone’s going to Taylor Swift. That’s pretty good targeting, regardless of anything else they’re in London and they’re going to Taylor Swift, or they’re in Germany seeing her.
The right audience and great data. The key is choosing the right bits of data to use and then adding context, as we said. Then you’re in a good place and can stand out.
So behavioural and cultural insight informing Uber’s strategy?
Culture is fascinating, isn’t it? One thing we haven’t talked about much is that we do a lot of work with West End theatres, targeting tourists in London. We can drive people to purchase tickets, or whatever else happens to be on when they’re on the road and going through London.
Again, context is perfect there.
There’s a bit of a schism among marketing academics about attention. Are you aware of this? One side (Byron Sharp’s) believes that only a threshold amount of attention is required for an ad to do its job, and anything above that is superfluous. The other side (Karen Nelson-Field’s) believes attention is more nuanced, and that duration and quality of attention matter (and that sometimes, fleeting attention is beneficial). What side of the argument do you fall and why?
The idea that we have to stick with one particular thing or another right now, I think that’s true. We’ve changed measurement models as many times as we’ve gone through the evolution of advertising.
To your point, if everything’s an ad network then we’re gonna have to think differently about these things. Everything’s contextually relevant for the moment in time, isn’t it? Our view is if you’ve got consumer attention at cultural and key moments, then you’ve got an opportunity and we should be measuring it. The next stage is what does that mean for the business? Are we going to fixate on a particular model over another? No, I think we just have to keep proving what we can do within the overall mix.
Why focus on attention and not, for example, the commercial impact that your ads provide?
The great thing about the uniqueness of the Uber platform is we can do everything direct-to-sale when it’s on the marketplace. For example, if you’re a restaurant partner or an FMCG advertiser on our marketplace, you can see sales. That’s pretty immediate.
So you have a performance platform and brand platform, which we’ve got, with brand and attention measures. We’re trying to cover all the measurement options we should have.
To what extent are you able to track the effectiveness of your advertising?
Outside of brand lift and favourability, the next stage is to say, OK, well if you’re getting attention, what does that mean? With the first survey we did with Lumen Research there were elements of longer attention, and greater brand awareness and brand lift. That’s stage one.
Then the question is, do we take it further into other areas to really understand broader outcomes? I’m sure we’ll work with advertisers over the next few months and work through that. We need to continue this journey and understand this area in more detail.
You’ve spoken about AI driving efficiency in media buying and how it doesn’t necessarily mean better marketing. Can you expand on that?
AI is a buzzword. There are some types of media buying where I think it would drive efficiency and other areas where it would not, because it might require a little more sophistication. What I hope is that the industry doesn’t throw everything at AI simply because it’s cheaper and hope that that makes for great media planning, because, generally, it does not.
This goes back to my earlier point about devices versus people. If you’re targeting people you need to have a good understanding of what people do. If you want to [use AI to] buy programmatic across websites, that’s fine. But if you want to do a big cultural moment, like a promotion around the Las Vegas Grand Prix, or something around Coachella or Glastonbury, then you might need a little bit of humanity in there to really understand consumers.
As ever with these things, it’s never one-size-fits-all, and it should never be a quick panacea that solves everything. I think smart companies will be providing AI layers, which will be sophisticated to a particular need. I can see AI being a benefit, but it’s not for everything.
How is AI changing Uber’s ad business? What other ad tech have you found especially useful for building the business?
We use AI very much in our business. That’s what improves little things, like how long is your journey going to take before you arrive or when is the delivery coming. A lot of that is driven by AI. We’ll continue to work on these areas and work out what else we can do with it.
Our business is based around getting people to places and things to people. That’s our starting point, rather than, ‘Here’s some AI stuff we can do’. Consumers are always our starting point; with ad formats and our business. But if it benefits the consumer, then we will use AI.
What’s next for Uber’s advertising business?
I think we’re still quite young in the market. We’ll continue to grow markets and opportunities. We recently announced that we’re in Saudi Arabia. Growth in markets is one area. We’ll continue to focus on measurement and opportunities around formats. We have our JourneyTV product in the US, on the back of seats in cars, which is going well.
Are there certain nuances or challenges in terms of going into particular markets that you’ve learned? You’ve mentioned Saudi Arabia and the US, for example.
The nuances are more on the Eats and food delivery and grocery side, in terms of the ways [different] markets operate within that space, because of maturity levels in terms of consumer acceptance of food delivery.
There are variations in scale and usage, but if you look, we’re big in airports whichever markets we’re in, to be honest. There’s always something fascinating in the data.
We have Uber Pet in some markets which we don’t have in others, which is always interesting because Uber Pet is pretty popular in the UK. It’s the taxi ride where you take your pets with you.
We have a lot of advertisers interested in that group, [which speaks] to my point about simple targeting. All you need to know is that someone is in an Uber Pet and you can target them accordingly.
[In other areas] we’re focusing a lot on our green and electric fleets. They are growing a lot and we’ve done a lot of activations around that. It’s proving very popular with consumers and it’s what they like to see, particularly in some markets.
Finally, congratulations on being shortlisted for ad tech personality of the year last year. Did you win? And if not, who beat you?… Because we’d like to interview them.
Zuzanna [Gierlinska] from MiQ won. It was good. Look, it’s nice to be recognised, and there were a lot of great candidates. It was a fun night, I seem to remember!
Featured image: Robert Anasch / Unsplash