Q&A-gency - Thomas Gwin

Hailing from a background in data strategy, Thomas has successfully blended analytics and creativity in various roles. Following a long stint at BBH London as head of insight and effectiveness, Thomas has moved across to Havas, working as Senior Strategy & Effectiveness Partner at Arena media.

Let’s dive into the questions and see what he has to say for himself.

Q You’ve recently moved from the creative agency BBH UK to Arena Media. Thinking about your approach to insight and data, to what extent has your approach changed – if at all?

It’s very much the same world. It’s still marketing, it’s still media, and it’s still advertising. My focus is still creating value for clients through effective advertising. The reality is that doesn’t really change where you work in this industry.

But where the differences lie is in culture. Each agency has its own set of values and ethos – and that cascades down into ways of working, whether that’s focused on insight or how you approach a strategy. At Arena, our focus is on rewriting the rules of marketing – it’s more of a challenger mindset and we set out to be ‘the agents of change for growth’. 

It sounds nuanced, but that  ethos has an impact on everything I do, and it’s also the promise we make to our clients: that we set out to disrupt the category and disrupt communications to make brands more liked, more talked about, more shared.

At a practical level, we work a lot with challenger brands and entertainment-led brands.

Q And has that change in culture and ethos changed the way you approach insight?

Since Arena’s ethos is anchored in a challenger mindset, a lot of what we do is challenger marketing focused.

That aligns us with a lot of brands, dare I say it, that are more culturally resonant.

Now, I know that “culture” is a bit of a slippery word – marketing struggles to define it – and its often spoken about like it’s some sort of mythical beast that needs to be captured.

But considering a lot of our brands are very culturally relevant, whether that’s gaming brands like Blizzard or Bethesda or retail brands like Red Bull or Dr Martens, it has a knock-on effect on the work we do from an insight and data point of view.

A lot of what we’re trying to find is meaningful moments in culture to expand their visibility and impact. That fundamentally changes the role of insight – as well as understanding the audience or the context in very broad terms, we focus on specific marketing tactics that we can deploy.

From a data point of view specifically we benefit from the Havas village model. We’re based in a single building in Kings Cross, but within that same building we have access to so many specialisms across the network across consumer science and analytics, CX and many more. We operate a system called Converged which we describe as the most comprehensive, accurate and human audience dataset in the marketplace.

Q Converged - it sounds like quite a data-heavy approach?

Yes, it is, and it’s a big advantage to have.

As you can probably imagine, it’s powered by AI, but the USP is that this system allows us to fuse all of the relevant datasets about our audience. Meaning we can connect all our intelligence tools to this one central spine, and critically, we can push to activation so there is no loss of fidelity from insight through to activation and measurement.

No loss of data integrity or consistency.

That’s very different to how many other agencies work, where the process is to go to an external agency to gather audience understanding, or use a series of siloed tools and methodologies, and the problem is when you bring it back in-house or try to push it through to implementation, all the richness and data quality they uncovered gets lost in transit.

Q A lot of your career has been based on demonstrating the long-term value of the strategies you develop. But, with performance marketing and short-term metrics dominating so many conversations, is long-term brand insight harder to land with clients than it used to be?

I could probably speak for more than 20 minutes on this question alone.

I would start by saying, looking at all the evidence and academic research in this space, it’s very clear that performance and brand should be seen as working together – it’s called the multiplier effect for a reason.

Having said that, I agree with you that there is obviously an issue in the industry being quite focused on the short term. This then naturally has a knock-on effect in destroying, or should I say, wasting the opportunities to develop long-term value creation.

In my current role, I am very implementation focused from a media strategy perspective, by which I mean channel and budget allocation. This allows me to have tough conversations with clients about brand and performance investment.

It’s important that we’re empathetic to client challenges. Yes, we all know the principles of marketing effectiveness, but clients, they’re reporting to shareholders; they have to nail every quarter in terms of their earnings.

That means their view on media can become quite transactional – if they put enough volume behind it, it will lead to new customer acquisitions. When a client is in a tight spot, it’s really hard to break out of a short-term perspective.

And that’s where agencies have to be more helpful. I think everyone would love to focus on the long term, but ultimately what you’re dealing with is market forces and capitalism. What I think we’re good at is taking a data-led and pragmatic approach – whether that is running our own econ modelling, incrementality or regional holdout tests, we work really, really hard to diagnose and surface positive client outcomes. That creates trust, but to get there you need to work with your client and go on a journey with them – the devil is very much in the detail.

We also try to be creative by looking for efficiencies that can be made through the implementation of the media strategy. And if you can make a saving, it becomes quite a convincing argument to suggest recycling that saving back into brand building.

Q So how do content creators fit in the mix then? I read recently that Unilever has already allocated 50% of its marketing budget alone on just content creators. What does that mean for insight professionals like me and strategic advisors like you?

Content creators are becoming a bigger play within media strategy. Historically the problem here has been measurement, how do you account for results from this channel? We’ve made progress here with studies like the IPA’s own research in the short-term and long-term effects of influencer marketing.

I think if you simplify the whole thing – people are spending more time on social media channels. Therefore, it makes sense that brands need to meet audiences where they live, whether that’s meme accounts or subreddits - it all has a part to play in your communication framework. In this sense, the job of media has evolved to scale what is already working and broaden beyond the community. The world cup is a good example. We’ve all seen the empty seats and the signs of fan fatigue. But look at bit deeper and you 'll notice audiences living inside club, creator and supporter communities every day, not just arriving for major tournament.

Q How has the approach to measuring effectiveness evolved over the last couple of years? What challenges do you face and how have you overcome these?

I do think the industry’s approach and understanding of effectiveness has become more sophisticated in recent years – but implementation is still as tough as ever.

The challenges are less to do with data quality and technical ability; it’s more about ensuring there is effective communication and integration across different stakeholder teams.

In some cases, measurement systems may not be fit for purpose, despite everyone being bullish about them. Similarly, business objectives can lack the clarity needed from the outset. At that point, the whole thing falls apart.

The other big challenge we’ve not yet discussed is when the ‘blueprint’ for measuring effectiveness isn’t in place. That can be for a number of reasons – lack of investment for example. At the other end of the spectrum, you can also aim too high with effectiveness. So it’s about getting the balance right between econometrics and more nimble experiments.

At Arena, we are very focused on learning agendas to make sure testing hypotheses and learnings feed into a living and breathing measurement process.

Q We’ve previously discussed how, regardless of which agency they work for, strategists now often have access to the same insight platforms, data sets, and cultural reports—raising the risk that strategic thinking can start to look and feel the same. Do you still agree with that perspective?

Yes, I do think it’s quite a big problem – and there are three issues which drive it.

The first issue is that everyone has access to the same things. That creates a challenge - how are you going to ‘extract value’ from a data source or tool that all your competitors have access to? That problem is often procurement-led. Linked to this problem, some tools that agencies use lack depth or quality, but are used as a shortcut, because time is so compressed. So everyone is using the same data, and a lot of it isn’t thought-provoking.

The second issue is the lack of diverse thinking, which has become quite topical in recent years. The industry is quite vocal now about avoiding the ‘London Bubble’ effect where a single focus group in Birmingham is equated with regional representation. By not reaching out to the fringes we become lazy. That attitude leads to sameness.

The third issue, which for me is actually the most problematic of all, is that as humans we massively underestimate our own confirmation bias. That’s an interplay between how you read the data, what the objectives are, and what you think and feel about a category, an audience. There is often a circular logic and it’s quite dangerous… I’ve seen this where people make massive leaps of judgment and you’re like ‘hang on a minute, yeah that’s really interesting, but we need some rigour to underpin that thought before we decide that it’s the definitive answer!’

Q So, just picking up on a classic one ‘confirmation bias’, how do you keep that in check?

You can never really stop it completely because it's inherently built into us - we have our own view and our way of making sense of the world.

So obviously the first thing is being aware of it.

But I think people aren't necessarily aware of just how bad it is. I used to run training where I would ask people a series of factual questions and then reveal the answer. And the idea behind it was to show the ‘gap’.

One of these questions was, in the UK, what percentage of the UK is built on (as in physical tarmac or roads or buildings) hard surfaces? Do you know the answer?

20%?

No it’s around 5% - and no one believes it! That’s because we all live in cities. It’s silly, but processes like this are quite powerful to demonstrate how biased we really are.

Q Going back to your point about ‘extracting value from data’, what do you mean by that?

It’s about being sure that you’ve explored every avenue you can, then taking those different data sets (whatever they may be), and finding a credible way to connect them so it’s all joined up - that’s where the value comes from.

Q. For a long time, we’ve joked between each other that we’re viewed by creatives as the ‘anchors to creativity’. Has that tension changed as more of the briefs are being shaped by data?

I think that view is more linked to data rather than insight. I do think most strategists, planners and creatives see audience insight as a helpful springboard for creativity. You know, you would have to worry if someone disagreed with that in our industry, right?

Data can feel more complex, more scary, more transactional, I think it because of that it can be seen as something that sort of kills off creativity. Because of my own experience, I have always sat in between those two worlds. And I’m a bit like a broker in that respect – I can see both sides of the argument.

My focus has always been: what is helpful for the creative process? In some cases, data can be super helpful, whilst in other situations it’s less pivotal, or not helpful at all. And that’s fine. So you have to lean into it and make it helpful when you can. If you already see it as the enemy, it will probably stay the enemy.

From my point of view it’s about leaning in and leading. Creative testing is famously resented by agencies, but the reality is if clients are investing huge sums of money - they are going to want to test the work. So you have to influence the process positively, own what’s actually being measured, make sure that the measurement marries up to the strategy and creative idea.

Q The best insights can come from unexpected places, spaces and conversations. Does the industry still create space for that kind of discovery?

Ha, its really hard to think of an answer that’s not really boring.

So to be controversial, I do think that people have always been quite empowered to bring their maverick thinking to strategy. That’s what they do right?

There is also quite a lot of encouragement for people to do that too. We typically celebrate maverick thinking; in the environments I’ve worked in at least.

But to be a maverick, you need to understand the category, the brand equity, the end customer. You can’t shortcut the process and magically land on a spikey insight or find an edge without doing the foundational work. 

That’s where disruption becomes most powerful. The strongest disruptive ideas aren’t born from being disruptive for disruption’s sake. Once you have that foundation, you can push boundaries with confidence and create work that genuinely challenges convention, rather than simply reacting against it.

Q Your point about being category experts, that has implications for how MR companies operate. In order to get to these spikey insights does that mean agencies have to focus on being “experts”?

I think it depends on how you define ‘experts’.

We worked together in the past on automotive work, which was an example of needing to understand the category inside out. That’s hugely beneficial. All categories require deep specialist knowledge – but some even more than others.

And then there are other instances where you are looking to work with an agency that is a specialist in a specific methodology or an application of research such as semiotics or behavioural science.

Q. This is a question I ask everyone but I would love to get your take on it. AI has really gained momentum this year with the proliferation of tools for researchers like me, but also agencies like yourself to use to streamline certain tasks and also execute final content. Focusing on its role within insight, do you think AI will make insight better, faster, or both?

The reality is the landscape is changing rapidly and it’s hard to gauge what the long-term impact of AI is going to be on our industry.

I think anyone answering this question is under a lot of pressure to say something really positive or something clever – but when you really unpick most answers, the key benefit outlined tends to be one defined by efficiency gains.

Thinking about how we use AI, and specifically talking about our Converged platform, the key benefits are about connectedness and speed. At Arena, we are constantly building agents, tinkering, experimenting, the potential is so vast. Being part of thought-leadership on this topic is essential.

With insight we have to be a little bit careful I think. I’ve seen use cases of synthetic audiences where the data used was clearly not collected for the end purpose, which is obviously not a great way of doing things.

At the same time there are so many more strings you can pull on now for insight – for instance LLMs have become an interesting source of data.

So I guess I’m giving you the boring answer and hedging my bets by saying, you have to 100% embrace it, lean into the innovation but balance that with a careful and well-articulated perspective. It’s great for automating completing timesheets, for anyone not doing that yet!