Generative AI is reshaping the market research and data analytics industries, with organisations across the board integrating it to drive new efficiencies by replacing or streamlining processes.

For those early in their careers, the industry’s embrace of AI also has the potential to redefine their role and what their future career may look like.

In the past two years, generative AI has evolved from experimentation to efficiency driver, with many agencies changing their processes and business models accordingly. At the same time, synthetic data is a potentially transformative but less widely understood area, with clients and practitioners alike still determining what its impact will be.

For those who are early in their careers or starting out in the sector, these technological developments add a further layer of complexity that they must get to grips with, while simultaneously learning the fundamentals of research and other required skills.

Christina Tarbotton, research director at Boxclever and co-author of a forthcoming MRS Delphi Group report on the ‘generation Z’ workforce, says: “The landscape has shifted so dramatically in such a short period of time, and the young researchers of today are going into an environment that is just totally different.”

Keeping up

Jayant Srivastava, insights manager at The7stars, says: “It’s getting quite difficult to keep up the pace with it, and a lot of the things that are now being spoken about are more along the lines of synthetic focus groups, or creating chatbots based on a customer segmentation, for example. It’s moved, in my eyes, from being that time-saver to going down the avenue of much more complicated tech.”

While Srivastava can see benefits, he says this environment brings greater challenges for those who are starting out or early in their careers. “It isn't necessarily a specialism for a lot of people who initially got into research – and obviously, there are lots of more complicated ethical and regulatory questions that come with all of that. It’s adding a lot more complexity, whereas at the start I felt like it was very much more of a simple time-saving tool.”

Like many other agencies, Tarbotton’s employer has set up an AI team dedicated to keeping on top of developments in the space – so while she feels confident that the work she is doing is secure, she has some concerns over industry implementation of AI if there isn’t full understanding of research.

“It’s a little bit of the Wild West at the moment, so it’s [about] making sure that for example, if a new company pops up that promises whatever kind of special advanced analytics, that we're making sure that the data is being stored safely, we understand how the processes work, if they're using things like synthetic data,” she says.

“I've been quite lucky in that I feel quite close to it, so I feel quite confident that everything that I'm doing in my day-to-day work is very controlled and being done in a very ethical way. But I do worry for someone who maybe hasn’t had the proper grounding in it first.”

The forthcoming MRS Delphi report, set to be published in November, includes the results of an industry survey with both early-career professionals and researchers further on in their careers.

The research found that regardless of experience level, participants shared the same amount of concern, to an extent, about the industry’s approach to AI – 42% of junior researchers agree the industry is doing enough to adapt to AI, in line with the senior researchers ( 36%). Tarbotton, who led on the research, says: “I wouldn’t be surprised if they are thinking about it in different ways.”

She hypothesises that researchers who have worked in the industry for a long time are thinking about the significant impact AI will have on their working life, whereas those with fewer years of experience are more focused on the industry’s need to adapt.

“From a younger researcher perspective, from what I read in the open ends, there was a lot of positivity surrounding AI. A lot of these people are going into it with very minimal background of working and research, so AI, to them, seems incredibly efficient and great.”

Tarbotton says a lot of junior researchers voiced concerns about companies being reluctant to try new methodologies and adapt. She says: “They’re a little bit more ‘gung ho’ in terms of being open to trying new things, while more senior researchers have the tried and tested things that they can be a bit reluctant to stray from.”

Supporting development

The introduction of generative AI, as with technological developments that have come before, has offered early-career researchers to spot opportunities to redefine how research is conducted and even coach others within their organisation, or clients, on best practices. However, in a fast-moving space, and an uncertain economic environment, are they themselves getting access to the right support and resources to really understand AI?

Srivastava says: “There are definitely some resources in place, but I think a lot of people are still doing a lot of that learning themselves and just actively exploring what’s out there. It’s quite difficult to get an industry view on some of these things because they do develop so quickly and it feels like from an industry perspective quite often are a little bit on the back foot to make sure you can analyse these things properly in due course, which is obviously the correct approach, but because it is so fast moving that kind of leaves a little bit of a vacuum in the interim.”

Josephine Hansom, formerly youth practice lead at Savanta and now running a next generation-focused consultancy, says: “I’m getting a sense that because gen Z know about AI, investing in their training in other things is being put on pause – because what do we train them about? What will that thing still be?”

Hansom adds: “Instead, it feels like a lot of seniors are wanting juniors to tell them about AI and seniors are not necessarily investing in juniors as much as they should do at this moment in time because of economic issues – so I think it’s causing a bit of a training gap.”

With a variety of information available, there is also a risk of people accessing less credible resources. Employers can support by streamlining training, says Tarbotton.

“People feel like they are often having to train themselves and there are so many different ways of doing that. You can attend an event online, either at the time it’s running or on demand, you can watch a YouTube video and there are resources on LinkedIn. I think streamlining training is something that people can do to support, because otherwise, if people don't have the information, they are going to look elsewhere for it – and then you have all these questions about credibility and whether it’s giving the right advice.”

There is also a need for senior researchers to take up the mantle and coach junior researchers.

Hansom says young people who participated in the Delphi research were more likely than seniors to want to spend time in the office. While not advocating for making office attendance mandatory, Hansom says leaders need to create conditions for juniors to learn. “They wanted to learn and be able to ask those silly questions. You need to find an opportunity to be able to answer the silly questions.”

The bigger picture, according to Hansom, however, is that nobody really understands fully what is shifting with AI because more junior people in organisations know more about it than the seniors in terms of day-to-day use, while junior researchers may not have the broader context of implications on the industry.

Hansom says: “I was speaking to somebody, and they were saying that they use AI almost every day in their work as a junior researcher. But the more senior you get the less involved you are in that day-to-day stuff, and I think that juniors are using it secretly, because we haven't caught up with knowing how to deal with that internally, and the data risks and all that sort of stuff.”

Learning by doing

The continued push for faster, cheaper research creates pressure across the sector, but there is a risk that this could affect the development of junior researchers if they are not given the time and space to bloom.

Putting in the hard yards is part of the process of learning in any discipline, and there is some concern that by allowing AI to streamline research, those in the early stages of their careers are missing out on the chance that more experienced colleagues had – to learn by doing.

Tom Woodnutt, founder at Feeling Mutual, was part of a group of Independent Consultants Group (ICG) members to conduct a survey of around 100 of its members about AI. One of the findings was that 72% fear that new researchers’ development of critical skills and expertise are stilted.

There is a risk that AI reduces the opportunities for experiential learning – and efficiencies should not trump the next generation’s development, according to Woodnutt.

He says: “There are positives to having generative AI tools, but it comes at a cost, and one of those costs could be, if not managed carefully, the development of those young researchers. Great research is ultimately about great judgement and that is a skill that’s forged in the trenches, if you like, over time – judgement to choose the right method, to decide on the data that matters and to confidently commit to a conclusion.

“This skill is something that you get through experience, trial and error, and critically, through feedback from senior people guiding you, which requires iteration and time and space to have that dialogue, to get that feedback.”

With AI offering time-saving opportunities throughout the process, Woodnutt says senior researchers should offer guidance and feedback.

“If we leave young researchers to blindly default to what AI says, that’s going to stunt their development and it’s also going to stifle the creativity across the industry. So, although AI might offer convincing shortcuts for junior researchers, whether that’s the design, analysis or even doing the moderation for them, that might save time in the short term, but it removes that opportunity, or at least reduces the opportunity for experiential learning, which is a key way to learn to embed information in the brain and to turn it into a long term skill that you can adapt to different contexts. And that does require feedback and senior guidance.”

There is also the issue of pressure. Srivastava feels the use of generative AI tools can increase the expectations – even if only subconsciously – of junior researchers, citing the example of brainstorming or initial thinking for a project. Leveraging such tools, he says, can end up with “junior researchers feeling like they have to come armed with more information or more research, or just more of a starting point than they would have maybe previously.”

He adds: “I would say it adds a little bit more pressure in that you feel like you need to have a bit of a unique perspective because you don't want to just default to the kind of generic stuff that that you can end up getting out of AI platforms.”

Future skills

With the research sector undergoing such transition, the current context offers an opportunity to rethink what the researcher of the future will look like.

If the sector is using AI as its junior, it raises the question of how juniors are treated, according to Hansom. “Is a junior someone who edits my report for me? Is that a good use of someone with a degree? People have been living independently, paying bills, having a life, and then they join us and we give them really menial tasks.”

Hansom is hopeful that the industry will be able to use AI to help juniors to have more impact – “by applying their knowledge, thoughts and individuality” – but she cautions “that can only happen if we teach them our trade. Unless we teach them our trade, they won't be able to do it. At the moment, they don't get to learn a lot because they are told to do these tasks and they don't get to see the big picture.”

Looking ahead, Hansom sees the skills needed to succeed as a researcher in future as two things: firstly, the ability to be good at learning and never stop learning; she references Google DeepMind chief executive Demis Hassabis, who recently championed the concept of ‘learning how to learn’. Secondly, she says: “Nothing that we're doing is going to stay the same in terms of our processes. We need to be able to be OK with that. We need to be resilient to that change.”

Tarbotton reiterates the need for solid grounding in fundamental research skills: “We all know that AI can have issues with hallucinations, seeing things that are not there. If you're solely relying on AI with no functional background to lean on in terms of being able to double-check the insights yourself, I think that has the potential to be a huge issue. That’s a big concern for me. I think it’s all about making sure that we're moving with the future of technology, but also not completely letting go of all of the really important fundamentals that give us the grounding to understand how to actually do research.”

• The MRS &more conference, focused on the next generation of research and insights professionals, will take place on Thursday 23rd October.