Learning from history, maybe
If you talk to a veteran in the insights industry about the AI technological advancements, you’ll potentially hear about when internet access changed market research.
Market research used to be conducted via phone banks or mailing out surveys or the classic “mall-intercept” to get results. Studies were crafted carefully, using probability sampling methods and sample sizes calculated to determine the minimum number of responses needed to achieve a certain confidence level, and results being reported with margins of error (“plus or minus X%”).
Market research was also time-consuming and expensive. It took weeks and months to get research completed. And, as phones became mobile and caller ID became more common, reaching people by phone became more difficult, which was making response rates drop, and timelines extend further.
Then, the internet boom. Surveys moved online, but they were still slow for a bit as companies had to either use their own customer databases to field their studies or else buy lists from list brokers. The first proprietary panels were built slowly and methodically using the old quotas so they could be representative panels. But this meant they were expensive to use.
As technology improved, it became easier to build panels. But as it became easier to build panels, it also became easier to sign up for panels, and the “professional respondent” emerged - someone who signed up for every panel possible to gain points and rewards. And as more panels were being built, research was becoming faster, and as research became faster, another differentiator was needed: price.
Soon, companies were expecting research that wasn’t only fast, but cheap.
And in the process, we sacrificed quality.
Because in order to meet demand, we stopped asking for sampling methods and confidence levels as often as we used to. We stopped slowly releasing emails to get survey responses and checking the data quality as completes came in, because we needed to get data back out to customers quickly. So we did what bits of quality checking were possible in the time allowed, maybe a quick glance for anything glaringly weird, and passed along the data. Or we didn’t check as we started to make data available “as it comes in!”
Now, we have AI tools that are promising to decrease time-to-insights even further. Now, the demand is for data to take hours to come in, not days. Or, if possible, for data to be instant.
Here’s my hypothesis: the same way that technology made building panels easier and a proliferation of panel providers followed by differentiation driven by price, AI technology is making these “instant insights” tools more accessible, and we will see a race to the bottom on price again. Already, we see “research slop,” where data from tools is readily available but not insightful or useful, and I’ve seen presentations to market researchers about tools that tout delivering professional-level market research in little time for customers who knew nothing about how to do market research.
Can we stop the momentum? Honestly, I don’t know. But here’s what I DO think.
It’s time to actually reimagine ourselves.
We can exert influence on tech choices. We can be selective about the tools we buy and use, evaluating them not just on a single use case application or feature set, but on core principles like usability, flexibility, and data security, so that we invest in tools that will amplify our expertise.
We can own the voice of the customer. We can combine our expertise with the tools available and own the voice of the customer to product teams, marketing teams, and executives.
We can be strategists. We know how to ask questions to get data. We turn data into strategy by stopping with the 30-slide decks and focusing on the key recommendation the business needs to move forward - not the data needed to not lose money, but the data needed to increase revenue.
AI isn’t like a new wheel on a bike. It’s not even like a new bike. It’s a new way of getting around town. We’re uniquely positioned with expertise to pair to it to help organizations use it in ways that will help them get around faster, so let’s lead the way!
Mental Health Moment
As the end of the year approaches, so, too, do holidays. And holidays can be a dicey time for some. Not everyone finds holiday seasons to be full of coziness and joy. I have friends for whom holidays bring sadness, loneliness, and loss.
For so many this year who have been struggling through unemployment and a severely difficult, ever-changing hiring landscape, it can be particularly tough to be facing holidays with tighter financial situations.
Let’s just be aware of allowing people space to spend the holidays as they need, but let’s also check in on each other to be sure that we’re okay during this season.
Wherever you’re at, may you find some time to rest this month before the turn of the year.