July 07, 2026
HubSpot’s Controversial New Customer Data Sharing Deal Didn’t Survive the Week
On 1 July, HubSpot rewrote its terms of service to let customer enrichment data flow into a shared pool other customers could draw on. By 5 July, the plan was dead. Chief Product and Technology Officer Duncan Lennox confirmed the data sharing reversal in a HubSpot Community post titled “We Got This Wrong. And We Are Fixing It”, four days after the change went live and roughly a month before the associated feature, Contact Discovery, was due to launch.
The mechanics explain the whirlwind speed of the climbdown. Customers who used HubSpot’s enrichment tools were opted into data-sharing by default. They had the option to withdraw split across three separate settings: enrichment participation, AI-model training, and tracking-code intent signals. Switching off one left the other two running. HubSpot framed it as business-card-level contact data, pooled to keep everyone’s records fresher. Customers read the development instead as a CRM they’d spent years building was quietly becoming raw material for a product sold back to the whole market, competitors included.
Gabe Larsen, CRO at Atonom, gave the backlash its sharpest line on LinkedIn:
“Somewhere along the way, SaaS companies forgot who the customer is.”
Dharmesh Shah, HubSpot’s co-founder, replied to him directly. “You are right. We made a mistake and have taken steps to correct it.” Founders don’t often concede a point that fast. That concession is arguably what turned a settings-page complaint into a story with legs.
“We made a mistake, and we’re reverting those changes completely.” Lennox’s community post went further than the LinkedIn apology, restating that CRM data belongs to the customer and promising that any future version of enrichment “will be fully and transparently opt-in”. Changes will now be flagged well in advance, across more than one channel.
Notably, what Lennox didn’t do is disown the underlying idea entirely. The word he used was “reassessing”, not “abandoning”.
Why HubSpot Wanted Data Sharing
The economics behind the plan haven’t gone anywhere. Contact and account data has been getting cheaper across the go-to-market stack for months. Clay has been undercutting on price, Unify has been opening up its platform, and many vendors are giving basic contact data away to compete on workflow instead.
In that market, a dataset that improves because customers actually use it is one of the few advantages that doesn’t erode. ZoomInfo built a business on exactly that model. HubSpot, sitting on hundreds of thousands of live CRMs, has a genuinely strong hand to play here, arguably better than most of its rivals. It just played it badly.
What the Reversal Means for CX and RevOps Teams
For HubSpot customers, nothing changes before 4 August. Enrichment and AI-training settings revert to whatever they were before 1 July, and no action is required.
The longer-term takeaway sits with whoever manages the vendor relationship rather than the tool itself. It’s naturally worth asking, at the next contract review, whether a shared dataset is on the roadmap, how many separate toggles would govern consent to it, and how far in advance you’d actually hear about it next time. HubSpot has said the next version will be opt-in from the start. Whether that holds true is absolutely a resolution worth pursuing.
