Synthetic research is coming, and it will stay.
Lots of great PR by Evidenza has gotten people talking. Is this the end of market research? Is synthetic research an existential threat to companies like Wynter?
Now I have not used Evidenza, so don't know the details of how they do things. The claims are that you don't need human data anymore, they create replicas of ICPs that you can survey.
What they don't say is where the source data for the LLM is coming from and how it is refreshed (in SaaS, a lot can and does change in a quarter!).
I have my doubts as pure AI/LLM-generated persona information is the average, it's where all the interesting edges have been filed off.
The problem they are solving is real, though: getting data through traditional market research is slow, and analyzing the data (e.g., 100 responses to 10 questions) takes lots of hours, even from professional qualitative researchers. These are the same problems we are tackling.
We at Wynter are already changing the market research game by sourcing research responses in under 48 hours (all humans). Now, we are also very close to launching our own take on synthetic research, which I believe is better.
My take:
First of all, human-based research is not going anywhere until humans buy products.
But synthetic research will make it better, and democratize access to insights (at least the way we at Wynter are thinking about it).
The way I see it, synthetic research is a companion to human-sourced, up-to-date data.
The best information is hi-fi: straight from the source, unbiased, unfiltered, authentic, and contains emotion. This is why qualitative research done on actual humans is where it's at.
Our AI/synthetic research product (launch is a couple of months away) is trained on constantly refreshed human-sourced data (needed to keep AI knowledge of the world current).
Wynter customers are running surveys and tests with real people every day.
Let's say 25 companies are researching the same ICP (e.g., talent attraction managers at large software companies), running surveys, and whatnot. We will feed AI (opt-in) with quarterly, up-to-date, anonymized knowledge on what those talent attraction managers think, need, and want.
So, what we have available is a large pool of data that we know about this ICP, and it's constantly getting refreshed. Now, synthetic research acts like a UI to query these insights.
You can ask all-new business questions about this ICP whenever you have a question. You can design surveys that poll the knowledge we have available about the ICP, and learn new insights in minutes without the need of a professional researcher at hand. You can conduct AI-based message testing as we already have data on their top pain point, desired gaines, etc. This will democratize access to insights.
Synthetic research coming, it's here to stay, and the quality of its output will depend on the quality and freshness of the source data.
