HubSpot Backed Down on Sharing Your Data. The Clause Underneath Didn't Move.

Written for anyone overwhelmed by AI in marketing. Not here: why it works underneath. The Trust Algorithm

On 25 June 2026 HubSpot posted a notice in its own community forum headed "July 1, 2026, Contact Discovery is Coming Soon + Legal Update". Its terms would change on 1 July, and on 4 August a prospecting feature called Contact Discovery would launch, so a sales team could find, verify and add new contacts without leaving HubSpot. The quality of those contacts, the notice said, was only possible because of a shared dataset: "When you opt into enrichment, some of your business contact data helps keep that dataset current." In scope was what fits on a business card, name, job title, company, work email and employer, plus signals from HubSpot's own email delivery about whether an address is live. Notes, deals, call recordings and custom fields were out. An account that used enrichment contributed to the pool, and the way to stop contributing was for a super admin to turn enrichment off.

It didn't last the fortnight. On 5 July Duncan Lennox, HubSpot's chief product and technology officer, posted "We Got This Wrong. And We Are Fixing It" in the same forum. It opens "We made a mistake", says HubSpot will not move forward with the terms changes it communicated on 1 July, and says a customer's contacts, notes, deals, call recordings, custom fields and records belong to the customer. He committed that new enrichment capabilities using customer data will be "fully and transparently opt-in", and that HubSpot will communicate early and across multiple channels before future changes that affect your data. The June notice now carries a line at the top saying it is no longer in effect. That is a real retraction, made in days, and it deserves to be reported as one.

This page has needed correcting twice, which is a small lesson in its own subject. The first version, published on 5 September, said HubSpot will share your enrichment data. I had read the investigation that broke the story on 2 July and not the vendor's reply three days later, so the page was a snapshot of a record that had already moved. The second version, on 9 October, reported the retraction but still took its dates and the wording of HubSpot's email from that investigation, and Lennox's words from a trade report, because I hadn't read HubSpot's own posts. This version reads them. The notes at the bottom keep the history, which is the habit this whole site argues for.

I don't want to hate on HubSpot; I'm picking on it because I know it. At BookPrint I was an accredited HubSpot marketer on the free CRM and the paid marketing automation, and I once exported my own form data through Zapier to import it back into the same product, because the form responses couldn't reach the automation unless I upgraded the CRM. What the Zapier loop and the enrichment episode have in common is not a villain. It is architecture. A vendor's data structure and a vendor's business model are the same thing.

What the retraction didn't touch

The retraction withdrew a rollout. It did not withdraw the permission underneath it. HubSpot's Product Specific Terms, last updated 16 September 2026, say in section 6.5 that you acknowledge HubSpot "may use Enrichment Data to provide, maintain, append, improve, enhance, and develop our commercial dataset", and that "only the Enrichment Data you transmit to HubSpot via the Subscription Services may be copied to our commercial dataset and processed further as HubSpot Content." The same sentence about copying sits in the archived version dated 18 September 2024, nearly two years before Contact Discovery was announced.

HubSpot's June notice said as much in its own words. It said the terms were being updated "to better explain this practice", which is a description of something already happening, not something new. What July added was a feature that would put the pool in front of other customers, and the feature is what the backlash stopped. Lennox's post doesn't give up on the idea either. HubSpot still believes, he wrote, that there is a better way to prospect than the status quo, and the plan now is to build it with clearer opt-in.

Enrichment works because of a pool. Every contact your account enriches is drawn from it, and what you send in can flow back into it. A platform sitting on that kind of pool is sitting on an asset, and in 2026 the companies holding large datasets are looking at what those datasets are worth to AI providers and as enrichment data.

Depending on noticing

So the backlash worked, this time, because customers noticed within days and said so loudly in HubSpot's own forum. A vendor can apologise for how it rolled something out, mean it, and still hold the asset. The useful question for a business is not whether HubSpot can be trusted. It is whether the business depends on noticing, on someone reading a legal update in a community forum in the week it is posted, and on there being enough of them to make a vendor change course.

That question has a smaller answer than leaving HubSpot: where does the truth about my business live, and who else can read it? If the answer is inside a vendor's database, the honest description of that arrangement is a rented core, and the rent includes whatever the vendor's terms allow next year. Own the core is the alternative, and it does not start with a platform migration. It starts with a simple system that writes down what the business actually decides, then a data layer the rented tools read from, so that the day a vendor's terms change, the tool is the thing you replace and not the thing you are.

Sources: HubSpot Community, "July 1, 2026, Contact Discovery is Coming Soon + Legal Update", posted 25 June 2026, for the plan, its dates, its scope and the opt-out route. HubSpot Community, "We Got This Wrong. And We Are Fixing It", posted 5 July 2026 (6 July in New Zealand), for the retraction and its commitments. CX Today, 7 July 2026, for Lennox's title. HubSpot Product Specific Terms, section 6.5, last updated 16 September 2026, and the archived version of 18 September 2024. All read on 13 October 2026.

Correction, 9 October 2026: the first version of this article, published 5 September 2026 under the headline "HubSpot Will Share Your Enrichment Data. That Was Always the Business Model.", reported the July plan as going ahead and did not report HubSpot's 5 July retraction.

Correction, 13 October 2026: the 9 October version dated the plan to an email of 2 July and said the sharing took effect on 4 August, quoting that email, from BLACKOUT Advisory's investigation BTI-2026-0001 (2 July 2026). HubSpot's own notice was posted on 25 June for terms effective 1 July, and 4 August was the launch date of Contact Discovery. I have not seen the email itself, so its wording is no longer quoted. The 9 October version also said customers were in unless they turned a switch off; HubSpot's notice describes enrichment as opt-in, with contribution to the dataset ending when enrichment is turned off, and this version says that.