For all their pitches promising one thing new, AI startups share most of the identical questions as startups in years previous: How do they know once they’ve achieved the holy grail of product-market match?
Product-market match has been studied extensively over time; total books have been written about easy methods to grasp the artwork. However as with so many issues, AI is upending established practices.
“Truthfully, it simply couldn’t be extra totally different from all of the playbooks that we’ve all been taught in tech up to now,” Ann Bordetsky, a associate at New Enterprise Associates, advised a standing room-only crowd at TechCrunch Disrupt in San Francisco. “It’s a very totally different ball sport.”
High of the record is the tempo of change within the AI world. “The know-how itself isn’t static,” she stated.
Even nonetheless, there are methods that founders and operators can consider whether or not they have product-market match.
Among the finest issues to observe, Murali Joshi, a associate at Iconiq, advised the viewers, is “sturdiness of spend.” AI continues to be early within the adoption curve at many corporations, and a lot of their spend is targeted on experimentation somewhat than integration.
“More and more, we’re seeing individuals actually shift away from simply experimental AI budgets to core workplace of the CXO budgets,” Joshi stated. “Digging into that’s tremendous crucial to make sure that this can be a software, an answer, a platform that’s right here to remain, versus one thing that they’re simply testing and attempting out.”
Techcrunch occasion
San Francisco
|
October 13-15, 2026
Joshi additionally instructed startups take into account traditional metrics: day by day, weekly, and month-to-month energetic customers. “How ceaselessly are your clients participating with the software and the product that they’re paying for?”
Bordetsky agreed, including that qualitative knowledge may also help present nuance to a few of the quantitative metrics which could counsel, however not affirm, whether or not clients are more likely to follow a product.
“When you discuss to clients or customers, even in qualitative interviews, which we do are likely to do rather a lot early on, that comes via very clearly,” she stated.
Interviewing individuals within the government suite will be useful, too, Joshi stated. “The place does this sit within the tech stack?” he suggests asking them. He stated that startups ought to take into consideration how they’ll make themselves “extra sticky as a product when it comes to the core workflows.”
Lastly, it’s essential for AI startups to consider product-market match as a continuum, Bordetsky stated. Product-market match is just not kind of one cut-off date,” she stated. “It’s studying to consider the way you perhaps begin with a bit of little bit of product market slot in your house, however then actually strengthen that over time.”
