Quo vadis, qual?

 

“The mechanisation of qualitative research”

“The sanitisation of real human beings.”

These are a couple of pages from Wendy Gordon’s excellent Goodthinking A Guide to Qualitative Research, published in 1999. The latter section ends as follows:

We study consumer behaviour as David Attenborough studies gorilla behaviour - sometimes in hushed awe and sometimes with cruel laughter. By using the word “consumer”, we attempt to objectify human beings and persuade ourselves that we understand people as individuals, when in fact we have often drawn stereotypical and numerically based approximations.

David Attenborough is still out there studying gorilla behaviour but Wendy Gordon is sadly no longer with us. I wonder what she’d have made of qualitative research in the world of AI and synthetic respondents?

There’s certainly a lot going on in qual, which is in itself healthy. I’ve pulled out four examples of the directions various agencies and individuals are taking. What these have in common is that they’re all emphasising the uniquely human characteristics that are key to brand growth - so far, synthetic sales are not a thing, as they say.

REVOLUTION! The Billion Person Focus Group (Abi Awomosu)

From the claim upwards, outwards, downwards, everywaywards, this approach is mind-boggling however I look at it.

Abi Awomosu has worked at Meta, Uber, Apple and Co. , has written a book How not to use AI   and runs a consultancy called Data, Art and Soul. 

It’s a completely different perspective on AI, radical and dare I say it, political. Less about “extraction” (I had never thought about AI in those terms, but now I think about all that “scraping” ...) and more about listening and going where people speak unfiltered. This includes “dark data” and IRL (now there’s an idea Wendy would definitely approve of!).

This course (although it’s not really a course, more an induction into a way of thinking and working) will enable you to set up your “own OS with agents, workflows and cookbooks”. A first for me, reading through this is “digital body language” which includes "cursor hesitation" and "rage clicks”. 

DIGITAL DOPPELGANGER: Juno (Michelle Gilmore)

Juno is a little easier for me to get my head around. What Michelle Gilmore has done here is essentially created a doppelganger, encoding her craft as an experienced qualitative researcher into “Juno”, and training her to run expert interviews. So Juno doesn’t just run surveys, she carries out qualitative interviews - listens, thinks then acts. Juno “tracks emotion, detects hesitation and knows when to probe, and when to hold back. The way an expert researcher would.”

There are options for recruitment of respondents, including drawing from a global panel of 3m people. I assume none of these are synthetic ;)

Here’s an 8-minute video of how it all works.

EMPOWERING RESEARCHERS: Quallie.ai “Good work made easy"

There are undoubtably a few outfits like Quallie.ai about. This is a support system for qual researchers taking care of the “donkey work.” “Don’t let hours of qualitative research overwhelm you.”

The philosophy, is hey, let’s not pick sides, AI is here to stay. Our platform can help you speed up the process, leaving you with more time for the insight and brain work. The platform can transcribe (in 40 languages), summarise, interrogate the transcripts, organise and analyse. 

HYBRID: Source Nine

Most of the big research players have embedded AI in their products and services - and the smaller specialised qualitative agencies are doing that too. Source Nine have a product they’re calling Signal Insight - “for when you have no time or budget for full quality, and “synthetic AI” isn’t enough.”

It’s a combination of desk research, in-depths and Source Nine’s propriety tools. 

The trick here is that the AI isn’t an add-on - it’s embedded in the agency’s ethos, in this case “Building brand equity grounded in emotion.”

So there we go. There’s plenty going on. And what is cheering is that all of these, and others like them have gone beyond flogging improved efficiency. They are all looking for genuine human understanding - and I for one hope I’ll hear less disparging remarks about “focus groups” and more positive comments about the qual(ity) of market research in the future.

  

Driving me round the trend ...

 

My trend-weariness was setting in already over three years ago, I see. This year, I’ve even looked at some of the AI summaries of trend reports and, maybe predictably, haven’t seen much to inspire me. It really does seem to be a load of (crystal) balls.

A new-ish genre in the trend oeuvre - in addition to category trends, consumer trends, media trends, tech trends et al - is marketing trends aka The Future of Marketing. One such is the Cap Gemini report I commented on a few weeks back.

Another is McKinsey’s State of Marketing Europe 2026.  In case you’re wondering how McKinsey work out the future of marketing in Europe, they do it by looking in the rearview mirror. Via a survey of 500 senior marketing decision-makers across 5 countries, backed up with a few depth interveiws with CMOs and academics. 

The stunning conclusion is that “marketing leaders are returning to mastering the basics but simultaneously advancing with utilizing modern tools.” Gosh. 

To get on in 2026, marketers must: Be Trusted, Be Effective, Be Bold. As opposed to ...

Golly gosh indeed. Who’d have thought it? 

There’s a rather hectoring tone about the report - in reference to AI: “Unless European marketing leaders rapidly change their attitudes ...”

Yes, tut, tut. Watch out before the cruel VUCA world gobbles you up, or the Burning Platform turns into an inferno.

The 50-page report has no less than 10 named authors (which doesn’t include Claude or ChatGPT), plus a list of industry experts as long as your arm, plus numerous McKinsey experts and colleagues. Not to mention all the others named who offered “editorial support."

There’s something “too many cooks” about this. It’s a report on an opinion poll, where the questionnaire has been designed by McKinsey. The 500 senior marketing decision-makers, I’m sure, work on a fascinating and diverse range of brands. And are, most likely, a varied bunch with some marked differences in character, in experience - and I’m sure in the way they approach marketing. 

Yet they’ve all been reduced to data points to make generalisations of the LCD sort, rather than HCF. Standardised to fit McKinsey’s frameworks, questions, flywheels, playbooks, agenda and whatever else.

The result is a rather bland, self-fulfilling prophecy. Seek and ye will find.

This stuff is all bad enough, as far as I'm concerned, when McKinsey does it.

But where McKinsey sucks, so do many lesser mortals, out to make a quick buck.

The author/s of this “Industry Pulse Report" report will remain nameless. The industry in question is digital media and advertising. 220 UK digital media experts were questioned in the survey that forms the basis of this - ahem - sales pitch.

But just look at this one chart:

The “experts” were asked what was suitable content to be adjacent to their brand. 

Around a quarter of these experts think it’s not just OK but positively “suitable” for your brand to be next to content that contains inaccurate information, mis/disinformation, or hallucinations. Or content that provides an ad-spammy or cluttered user experience. 

With the rise of synthetic respondents, I am beginning to wonder about the “experts” who find time to take part in these surveys.