HubSpot announced earlier this month that it was updating its terms of service to allow enrichment data (business contact details, employer information, email deliverability signals) to be shared across customer accounts. Customers were opted in by default. Four days later, after a sustained and very public backlash on LinkedIn, co-founder Dharmesh Shah and chief product officer Duncan Lennox both admitted the company had made a mistake. The reversal post was titled, simply: "We got this wrong, and we are fixing it."
What happened is worth unpacking, because this isn't really a story about a botched terms-of-service announcement. It's a story about where legacy SaaS companies are in their attempt to stay relevant in the age of AI, and the flawed thinking that's leading some of them off a cliff.
The logic behind HubSpot's original plan wasn't hard to follow. Years of customers enriching their CRM contacts (cleaning data, verifying records, building reliable contact databases) represents an enormous amount of cumulative value. If HubSpot could pool that enriched data across accounts, it would have one of the richest prospecting datasets in the industry, which it could then use to power AI-driven "Trusted Prospecting" tools and, ultimately, differentiate its platform.
That's a classic SaaS moat strategy: aggregate the data your customers pour into your product, use it to make the product better for everyone, and create a competitive advantage that's hard to replicate. For a long time, it worked. Customers had limited options, limited visibility, and limited leverage.
Customers pushed back hard and fast. The objection was straightforward: they had spent years building, cleaning, and maintaining their CRM data, and the opt-out default treated that work as HubSpot's to redistribute.
SaaS vendors who believe they can build moats around the data their customers store in their platforms are going to find that strategy increasingly difficult to sustain.
There's a broader shift happening that makes the old moat strategy not just ethically questionable but strategically fragile. For most of the SaaS era, customers adapted to how software vendors built their workflows. You wanted a CRM? Here's how HubSpot thinks a sales process should work. You wanted marketing automation? Here's Marketo's model of a campaign. Customers accepted constraints on how they worked because the software delivered enough value, and because the alternative was worse.
That dynamic has inverted. Customers today are increasingly accustomed to working with tools that adapt to them. AI tools like Claude let users describe what they want to accomplish in natural language and get a result that fits their specific workflow, their specific data, their specific context. The flexibility customers now experience every day with AI is recalibrating what they're willing to tolerate from the rest of their stack.
When HubSpot announced it was going to treat customer-built data as an asset to be pooled and redistributed, even at the "business card level," customers didn't see a platform improvement. They saw a vendor trying to extract value from the work they'd done, using a contract change as the mechanism. The opt-out default didn't help. It made the extraction feel automatic, frictionless on HubSpot's end and consequential on the customer's.
For most of the SaaS era, customers adapted to how vendors built software. Now software vendors have to adapt to how customers actually work. That's not a temporary disruption. It's the new baseline.
HubSpot will almost certainly come back with a revised version of Trusted Prospecting, one that's opt-in, transparent, and framed as a value exchange. That's the right move. But the companies that will lead in the AI era aren't the ones who figure out better opt-in mechanics for extracting data from their customers. They're the ones who stop thinking about their customers' data as an asset and start thinking about their customers' outcomes as the product.
The rest will keep finding themselves in the position HubSpot was in last week: issuing apologies and reversals while their customers question whether the relationship has changed underneath them.
If you want to go deeper on how AI is reshaping the SaaS landscape and what it means for martech practitioners, Marketing Demo Lab (MDL) covers exactly this territory. MDL is a podcast co-produced by StoryAZ Studio where we dig into martech products, demos, and conversations with experts who are navigating these shifts in real time. Visit the MDL page https://www.storyaz.studio/marketing-demo-lab to learn more and find out how to come on the show.