In a recent LinkedIn post, sales trainer and thought leader John Barrows issued a stark warning to startups in the AI and SaaS sectors, highlighting the potential obsolescence of products that function as standalone destinations rather than integrated connectors.
Barrows uses the recent development of Microsoft integrating legal AI capabilities into Microsoft Word, following the substantial funding of legal AI startup Harvey, as a prime example of this disruptive trend. He points out the speed at which a seemingly valuable startup could be rendered irrelevant by a platform update.
“If your product is a destination, meaning you need users to leave where they already work to come use your thing, you are one platform update away from irrelevance.”
The Platform Play: A Threat to Standalone Products
Barrows argues that the core issue for many startups is their reliance on users adopting their product as a primary workspace. He contrasts this with tools that integrate seamlessly into existing workflows, suggesting that the latter are far more resilient.
“This is not a legal industry story. This is a warning for anyone building a product right now, AI or SaaS,” Barrows emphasizes. He elaborates that the value of a product, or the amount of capital it has raised, becomes secondary if the underlying platform it relies on decides to incorporate similar functionality.
“It doesn’t matter how good your product is. It doesn’t matter how much you’ve raised. If the platform you’re built on top of decides to eat your functionality, your valuation can disappear overnight.”
The Rise of the AI Console and Connector Model
Looking ahead, John Barrows envisions a future where users primarily operate within a few major AI consoles, such as ChatGPT, Claude, or Gemini. In this model, other software tools would serve as connectors, feeding data and functionality into these central AI hubs rather than demanding users’ attention as distinct destinations.
Barrows shares his personal workflow as an example, noting that while he uses tools like Apollo.io for prospecting and Otter.ai for call recording, his primary interaction is now within an AI console. “Those tools work BECAUSE they connect. The ones that don’t connect are the ones I’ve stopped using,” he states.
“I’m already living this. I run prospecting sequences through Apollo.io and call recording through Otter.ai, but I’m not living in those platforms anymore. I’m working in Claude and telling it what to do.”
Implications for SaaS and AI Development
The implications of this shift are profound, according to Barrows. He stresses the critical importance of building direct backend connections to major AI platforms for any company aiming for long-term survival in the next few years.
“If you’re building a product, AI or SaaS, and you’re not building direct backend connections to the major AI platforms, I don’t understand how you survive the next 2-3 years. The destination model is dying. The connector model is what’s next,” Barrows asserts.
While acknowledging that some specialized products, like Harvey, might survive due to unique intellectual property, Barrows warns that many other AI and SaaS companies lack such defensible moats and are likely to face significant challenges.
He concludes by posing critical questions to both product builders and buyers regarding their strategies and the longevity of current point solutions, suggesting that the number of such solutions likely to remain relevant in three years is far lower than commonly perceived.
📝 About This Content
This article is based on insights shared by John Barrows on LinkedIn.
📅 Originally posted on May 18, 2026 | View original post on LinkedIn →