Preview of 2025: The challenges ahead

Babita Earle, international managing director, Zappi

I believe the biggest challenge for our industry moving forward is data quality and representation. While the desire to be mindful of representation is genuine, it has fallen off the radar and been taken over by other more ‘sexy’ topics like AI. There is and should be space for both 

Research is largely misunderstood as an industry. More often than not, it’s one that people fall into. We still, after many years of trying to open our doors to people from all parts of society, struggle to diversify to a meaningful extent and keep them.

What leaders need to realise is that there are very real business consequences of putting representation on the back burner – not only in the data we take in, but in those who are managing and applying empathy to data. AI only increases the need to ensure we reduce biases and make sure what we feed the machines represents society and consumers fairly.

Kelly Beaver, chief executive officer UK and Ireland, Ipsos

As with other industries, market research will be susceptible to the large number of global shocks facing us in 2025, including potential changes with respect to trade, continuing wars in Ukraine and the Middle East, and more elections in countries such as Germany, Canada and Australia.

However, I’m optimistic that these shocks will provide us with new opportunities to provide insights to our clients, as it will only become more important for them to stay close to what consumers/citizens and audiences are thinking and feeling during this time.

Ray Poynter, chief research officer, Potentiate

I think we will see some significant losers, some companies and some roles inside companies.

Graham Idehen, senior director, customer experience – international, Dscout

Balancing speed with depth. In a world demanding instant insights, how do we ensure we don’t sacrifice the quality of understanding for the sake of contracting budgets and reduced delivery timelines?

At Dscout we’ve seen a stark increase in research democratisation and tool consolidation as researchers, designers and product managers deploy a multitude of methodologies, supported by generative AI, to overcome this age-old challenge posed by accessing instant insights.

Danielle Todd, director, The Forge

A perennial challenge, but balancing momentum with focus. It’s always tricky to prioritise long-term strategy investment with keeping an eye on short-term wins and needs. This remains a critical challenge for any brand in our industry.

Amy Cashman, executive managing director of the UK insights division, Kantar

We can’t ignore the economic outlook for clients. Businesses are facing rising costs following the changes announced in the Budget. Inflation has eased but household finances are still under pressure too.

For the consumer sectors we work in, that means the battle to offer value and compete for spend will remain fierce. Clients will have some tough decisions to make and the IPA Bellwether report showed that total marketing budgets were already being put on ice during the third quarter of 2024, with market research spend down by 1.5%.

Over the next 12 months, we need to champion the power of market research to help clients make better, informed choices – from spotting new opportunities for growth, to staying resilient by predicting how consumer behaviour might evolve. We’re an investment to protect, not a cost to be cut.

Jane Frost, chief executive, MRS

In terms of challenges, we’ll need to have an eye on whether consumer and business confidence settles. Many clients have been waiting with bated breath, holding the purse strings tight, while we waited for the chancellor’s Budget. I hope we’ll start to see the nerves ease as businesses get to grips with what the announcements mean for them.

Either way, in tricky economic times like these, research really comes into its own, making sure budgets – whether public or private – are being spent in the right places.

Crawford Hollingworth, global founder, The Behavioural Architects

The market research world is moving super-fast and it’s easy to end up chasing to catch up. But once you fall behind, that becomes harder and harder to do. In our industry, where margins are tighter, projects more competitive and clients are always wanting more, it can feel a bit like being the proverbial hamster on the wheel.

Emma Cooper, chief people officer, System1 Group

The biggest challenge for the industry heading into the new year continues to be data quality. It’s an ongoing issue, made even more complex by the rapid advancements in AI, which both heighten the challenge and offer new tools to address it.

Tatenda Musesengwa, vice-president of audiences, Savanta, and co-founder, Colour of Research (CORe)

There’s a couple to my mind. I’m concerned that the industry is at risk of being in a ‘race to the bottom’ in terms of online sample cost, which is likely to impact quality of the data we’re using. I think we need to find a better balance.

More broadly, we need to continue to make sure we’re getting sufficiently high-up the ‘food chain’. Insights should be setting strategy – not verifying it – and that means proving you deserve to be in the most important conversations.

Neil Bellamy, consumer insights director, GfK

A crisis of understanding research approaches and methodologies. Without it, we risk losing the price premium for providing quality research, setting the industry up to be replaced by AI for instance or cheaper alternatives to the well-managed online panels with rigorous data quality checks that we rely on today.

William Ullstein, UK chief executive, YouGov

While it is important to understand the benefits AI can bring, we also need to be aware of the limitations that come with any new technology. Ultimately, all AI models are judged by the quality of their outputs, which are really defined by the quality of the inputs. The industry as a whole needs to be focused on producing accurate, trustworthy data so that clients know our work can be relied on.

Fiona Blades, chief experience officer, MESH Experience

Declining budgets, re-organisations and inability to move forward quickly.

Mark James, chief executive, Differentology

As AI and automation continue to revolutionise the research world, they bring a surprising twist: a growing environmental footprint. It’s estimated that a 100-word email crafted by AI uses as much energy as 14 LED light bulbs glowing for an hour.

As AI models and data centres ramp up, so does the research industry’s carbon impact, creating tension with MRS sustainability goals. The real challenge for 2025? Unlocking AI’s game-changing potential while staying eco-savvy.

At Differentology, we’re tackling this head-on, building a carbon measurement plan to lead the way in blending cutting-edge tech with sustainability. It’s innovation with a conscience.

Paul Hudson, founder and chief executive, FlexMR

I think it will be similar to the last few years, grappling with tough market economics coming from a variety of underlying factors, including reductions in marketing budgets, unrealistic expectations that AI is solely an opportunity to drive cost-efficiency, continued inflationary cost-pressures, and then wider macro-economic uncertainty related to politics and international trade tensions.

Josh Glendinning, board director and equity partner at Opinium

The difficulties of collecting robust and reliable data have been mounting for many years but this year many of these problems came to a head. These issues are serious challenges for the research sector, but they also underscore the importance of foundational research principles in an era of big data. In 2024, even governments are grappling with these issues.

The Labour Force Survey (LFS) is not only a vital source for many quantitative researchers, it also informs much government and Bank of England policy. Response rates to the LFS have cratered since the pandemic, leading to increasingly unreliable results. Estimates vary, but the LFS may be undercounting the number of workers by up to one million. Some may suggest we can overcome these issues through technological innovations such as increasingly complex weight schemes, unorthodox data collection or even synthetic data.

Can models built on the corpus of online (primarily American) text truly imagine the views of an offline British person in their 70s with a long-term disability, for example? If the source or training data is incorrect, we may end up building our castles on sand. While not all research is as consequential as the Census or LFS, it is important for the industry to confront issues around data collection for the sake of its future success.