From Full-Service to Full Stack: How Stagwell Marketing Cloud Is Building the Future of Brand Tech

While some advertising holding companies have been finding various ways to encase the creative agency jewels of their crowns onto a firm structure of data and AI-driven martech, Stagwell has been working quietly – and ambitiously – to build a tech stack that’s intrinsically complementary to its agency offerings. At the heart of its vision sits Stagwell Marketing Cloud (SMC), a portfolio of self-service marketing tools designed not to compete with agencies, but to complement them.

Three and a half years into its development, SMC is already changing how brands approach media, creative and communications – from research and influencer marketing to asset QA and performance. But what really stands out is how Stagwell is connecting the dots between insight, execution, and scale.

After watching a demonstration of the tools in the Cloud at an event at Stagwell’s EMEA HQ in London, LBB’s Alex Reeves sat down with Elspeth Rollert, CEO of Stagwell Marketing Cloud, and Lindsay Hong, CEO of SmartAssets – one of SMC’s homegrown success stories – to talk about boring-but-brilliant tech, what agencies get wrong about AI, and how SmartAssets started as a “ball-ache” solution and grew into something far more powerful.

LBB> Not all holding companies have a tech stack so clearly packaged quite like this. So what is the value of Stagwell Marketing Cloud?

Elspeth> About three and a half years ago was the genesis of the Stagwell Marketing Cloud. This is actually my third time working for [Stagwell CEO] Mark Penn – I had been at Uber for seven years and he came to me and said, “I’ve started Stagwell. It’s the modern-day challenger holding company.”

At that time, the merger between Stagwell and MDC had just happened, and they’d just gone public. His position was, “We have some of the best agencies in the world, and I feel really good about our portfolio on that side. But the reality is, modern marketing organisations need both full-service agencies and self-service products. On the full-service side, I feel very good about that. But on the other side – self-service, DIY, in-house – I’ve started to invest in some of that stuff: ARS, voice. The reality is, those products can’t sit within agencies. They’ll die on the vine.

“Ultimately, I want to create a group – Stagwell Marketing Cloud – to bring all our tech products under one house, so we start to build scale, drive efficiency across shared services, and really build our offering in a more cohesive, comprehensive manner.”

His push then, which is still the push now, was: there's some technology we’re going to fund. Lindsay is a great example with SmartAssets, an in-house product that we’ve built, given our deep expertise. You saw in H2 of last year, two acquisitions in EMEA, InfluencerMarketing.Ai (IM.Ai) and UNICEPTA, with the majority of revenue and employee base here in EMEA.

So, the initial pitch to me and proposition is still the same today: we want to be a full-stack shop, from full-service to tech, focused on top-line growth, user engagement, and so on.

From sales, account management, engineers, developers, data scientists, we really coalesced around building out these products. We have three different verticals: media, research and communications. We’re competing in communications with the decisions of the world, and in influencer marketing with the captivators of the world. That’s our competitive set and the work we do.

Then, of course, we’re part of Stagwell. So we lean into the deep expertise we see on the agency side. Aaron [Kwittken, founder and CEO of PRophet] is a good example – he deeply understands the workflow, where the gaps lie, and how we build against that.
We’re in a continuous cycle of listening to agencies; what are they hearing? What are clients saying? How do we build products against that – not to compete with it, but to support a full-service offering? Some clients lean more into agencies, some lean more into products, and we want to make sure we’re meeting them wherever they are.

Also, we’ve spent a lot of time building out the products. Last year, we landed a few different clients. With our ability to then scale those products to work with bigger brands and pockets, we’re starting to see some pretty strong traction. The whole goal of this is we’re not building for one, we’re building for many.

LBB> I loved the Spotify playlist example you gave to illustrate how exciting AI-driven tools can be in terms of what meaning they actually deliver. You hear about “mass personalisation at scale” but it is more fun than that sounds when you see the result?

Elspeth> Even something like email – operational, dull and dry. But the reality is, it's part of your life every day. People can be scared about embracing certain pieces of technology, but it's already ever-present. One of the things we do really well with our product, and even working with agencies or clients, is time to value – you see it right away. SmartAssets is not a 12-month implementation. Lindsay can go in and start proving the value of the product pretty quickly, because we deeply understand those users and how they’re interacting in their day-to-day workflows.

Lindsay> To your point about it being quite simple, life-enhancing things, agencies often go in selling themselves as the cleverest person in the room who's come up with something no one's ever said before, right? That makes marketers feel like, ‘Oh, you can do something I can't’.

With our solutions, quite often we're solving a really simple problem that’s actually just a ball-ache, frankly. And there's a lot of value in that. It's a bit different to the sexy world of creative agencies. They're going in saying “Here's something amazing.” But I'm going in and saying, “You know how painful it is to get assets signed off? How 40% of them are the wrong size, and you don’t know which ones went live? You've got all these emails and can't find the latest version, and everyone’s got an opinion on what’s right and wrong?”
“Oh yeah, God, that really hurts.”
“Would you like us to automate that for you? Take the emotion out of it and put a computer between you and your stakeholders?”

It’s a different sell, but to Elspeth’s point, it makes us a more full-service network. We can come from both the super sexy, insight-led side, and the real-world problem side of delivering marketing at scale. Our solution is definitely one that’s about scale – processing a hell of a lot of content very quickly, but to quality.

As generative AI means there will be many more assets produced, it’s all very well being able to create a lot, but you also need to assure quality, as well as make sure they’re brand-safe and media-ready. That needs automating too. It might not be as sexy, but there’s lots of value in there, and it’s easily understood.

LBB> You hinted in the presentation at things like insights from media that can be really useful for creative. And creative is more and more looking for those insights. I'd love you to delve into more of the kind of inspiring side of things that come out of the insights Smart Assets delivers.

Lindsay> We have, I think, over a hundred questions now in the LLM, and we're adding to them all the time through our partnerships with brands. They’ll say, “Can you tell me whether having sound on matters?” or “Can you tell me, for our influencers, how often they’re showing children compared to our competitors?” They keep asking us questions, so we build out this massive database, and the model gives us the answer.

It looks at every asset you’ve produced, tags it up with what’s in there, and then cross-references it with performance. So it might say, “Every time this has been optimised with sound off, the click-through rate goes up 20%,” or something like that. Really, there’s no limit – if you can think of the question, we can execute it automatically at scale.

And that creative dataset – what’s actually in your assets – we’ve never really had that before. Going back to the boring stuff, try getting creatives to tag up assets in a consistent way. If they’d wanted to fill in spreadsheets, they’d have become accountants, you know? We do that part as well. So taking away the boring admin opens up a world of interesting insights – things you can ask, because you finally have the dataset.

LBB> Let’s imagine a Stagwell brand client wants to buy the full stack. They want to take everything that SMC offers. They've got a problem they need to solve for their brand. What would all the different parts of Stagwell Marketing Cloud do for them?

Elspeth> I mean, I come from Uber, so I’ll go back to that.
You start on the research side. Who’s my customer? Who am I reaching, and who’s not listening to us? So, on the media intelligence side, you use UNICEPTA. What’s the sentiment toward my brand and my competitors? Start with that base.

Then, I might see something in the data I want to double-click into. I go into Quest DIY to field a specific survey, to get under the hood and understand that better. Or I use Quest IC to tap into a business traveller audience – a very high-value audience for Uber.You’re pulling different levers, depending on whether you want to do research on the spot or ongoing brand measurement. You’re looking at your brand and those of your competitors, at your target audience, and there are a number of levers you can pull.

All right, now I understand my audience. How do I reach them? How do I land a story with them? How do I influence them? If I want to target key decision-makers in Washington, London, wherever, I lean into UNICEPTA or IM.Ai to land my story with the audiences that matter. As I get more of that message out there, I use influencers to make sure it’s landing in the most effective way. We do that today with a global FMCG partner.

Then, the question is, “how do I tag up those assets in influencer, as I put them into market?”

I wouldn’t say we’ve fully built out every single capability across the group yet – we have a lot more to do on the media side – but from understanding an audience to landing a message, to getting smarter about creative and learning from that, that’s where you start to see it all come together. We didn’t get into it too much, but in some of the slides I showed, we’re building out more of the foundational SMC platform. So, as a user – say I’m a CMO at Uber – I come in, and I have single sign-on access to all these applications.

You’re going to continue to see us start to create stronger linkages across those products. You’re not just getting single sources of truth in this product or that product – they start to speak to each other. You get a more comprehensive suite of solutions, instead of just a set of standalone tools.

LBB> Lindsay, is that one of the benefits to developing a software tool in an agency network environment for SmartAssets?

Lindsay> A good example of that is the nappy example you saw today [which demonstrated that whether a baby appeared in a nappy ad made no impact on effectiveness]. It’s really interesting finding out what kind of content works with a brand-specific audience. This becomes a measure for selecting your influencers. IM.Ai has a whole suite of measures for influencer selection – hashtags, demographics, does their creative style speak to the preferences of your audience? We can become that tool, and we’re actively talking about API integration now, so that people can use SmartAssets as part of that. That’s a huge opportunity.

I think the key thing about developing a software startup in a network environment is, number one, the agencies want to use our tech. It’s good for Stagwell for agencies to use our tech. So, there’s a long-term intention to build a very good product. That’s different to VC-backed AI startups, where they’re just trying to achieve certain leading indicators to sell those assets off.

That’s a really interesting piece – we want to use our own technology, so we’re invested in making it good. We work together. I mean, if you look at the layout of the Blue Fin building, we’re on one floor. All the agencies sit together. I know the leaders of all the agencies. We chat to each other all the time. They’re able to influence how I develop my product, so it really speaks to the use cases. You don’t get that if you’re not inside the network.

Likewise, if you’re a big tech company, you’ve got a huge development workforce, so you can spin things up quickly. And you might think that advertising and marketing is a nice cash cow, but you don’t have people in there who’ve lived the reality of trying to get 17 assets for Poland ready by tomorrow in 13 different formats. They haven’t done it. They don’t know it.

That puts us in a very Goldilocks spot. And all those VC-backed AI startups – many will fold, many will merge. We’re in a nice sweet spot to develop technology that actually works, that will sustain for the long term, and that is based on real use cases. We’re working together to knit those products together. And the truth is, we don’t know yet exactly what those future workflows will be. But we’re in a very good position to discover them before anyone else and put them together.