Ed Zitron, the CEO of tech public-relations agency EZPR, recently published an article making a sad prediction about AI innovation, “AI Is Becoming a Band-Aid over Bad, Broken Tech Industry Design Choices.”
His argument is simple: the tech industry’s appetite for growth incentivizes endless feature bloat rather than elegant usability. That is, products like the iPhone or Google Search continuously add features to appeal to all conceivable customers and to head off niche competitors.
The result? Apple is incentivized to allow nearly any app into its App Store, Amazon to invite every obscure vendor to sell on its platform, and Google Search to cram in interminable ads and irrelevant links.
Within these existing ecosystems, AI can feel like a solution to bloat. More intelligent natural language search, for instance, means that you can sometimes find the needle in the haystack without combing through each piece of hay. But the problem is that the whole haystack is still there–and growing every day.
What Zitron calls “the Rot Economy” is hardly limited to consumer technology, though. If anything, the problem is more pervasive in the world of B2B SaaS software, including the market research space that Glimpse is part of. And the incorporation of generative AI-based innovation into market research platforms could make things much worse–or potentially much better, as we’ll see.
Some of the leading survey platforms, like Qualtrics, are amalgamations of every feature hardcore market researchers could desire. Few of us, however, would say that Qualtrics has placed a priority on elegant design, usability, or accessibility to non-traditional users–like marketers and content creators–who might benefit from the collection and analysis of first party data.
(To learn more about how Glimpse is using generative AI to simplify complex research and business challenges, take a look at our recent Harvard Business Review or Adweek articles.)
Lessons for avoiding “The Rot Economy” trap in market research
Here’s what NOT to do with generative AI:
- Don’t automate research approaches that are already broken: We all know that 90-question surveys offer a terrible respondent experience. Automating the creation of 90-question surveys with generative AI doesn’t fix the problem.
- Don’t focus on data analytics more than the quality of inputs or the usefulness of outputs to business challenges: Sure, applications like ChatGPT are amazing; but they’re only as good as their training and input data. Most of the work to make generative genuinely useful to organizations depends on the careful incorporation of high quality proprietary data. And on the translation of gen AI outputs into real business contexts.
- Don’t add gen AI capabilities on top of a pile of existing capabilities: For gen AI to be truly valuable, it must be transformative rather than additive. Focused experimentation will reveal challenges uniquely suited to gen AI-based solutions. For instance, it’s never before been possible to instantly capture the nuances of unstructured (human language) data at scale, instantly.
- Don’t create ever more complex dashboards: If your goal is to expand the accessibility of the data that matters to organizations (and especially to non-specialist users), the path forward is not to add more layers of data to your dashboard. Instead, focus on making the important data more visible and easier to engage with.
It may sound obvious but the only real way to avoid “The Rot Economy Trap” is to focus maniacally on the needs and behaviors of users and respondents, as elegantly as possible. The recent fervor over–admittedly astonishing–advances in gen AI can easily distract us from this imperative. After all, it’s much easier to add some gen AI features to a bloated platform than to change anything fundamental.
At Glimpse we strive every day to follow this principle. Take a look at our platform and see if you think we’ve succeeded!
.svg){kind=small}
-20.png)
-8.png)
