Share with your CMO
HubSpot planned to fold business-card-level contact details from customer accounts into a shared enrichment dataset powering a new prospecting feature, then reversed course after a customer backlash that landed within days of the announcement. Chief product and technology officer Duncan Lennox admitted the company got it wrong and committed to a fully opt-in model for any future enrichment involving customer data. The vendor governance lessons from this reversal extend well beyond HubSpot, applying to any martech contract where “service improvement” language can quietly expand how customer data gets used.
What this means for your business
The HubSpot episode exposed a failure mode that lives in nearly every martech stack: the gap between the contract signed at procurement and the terms quietly governing the platform three years later. Marketing and revenue leaders are most exposed here because CRM and marketing automation sit at the center of their operations, hold years of accumulated customer data, and get renewed by people who rarely have the time or authority to flag a legal update buried in a community forum.
What makes this genuinely dangerous is the compounding effect of broad contractual language. Phrases like “service improvement” or “product analytics” look harmless at signing. But as vendors add AI features, acquire enrichment providers, and build cross-customer data products, that language stretches to cover uses no one anticipated. HubSpot’s original terms described business-card-level fields entering a shared dataset, with opt-out controls that were technically accessible but not prominently surfaced. That’s the pattern, not the exception. The distinction between opt-in and opt-out is not semantic: opt-out puts the burden on the customer to notice, understand, and act before a deadline, while opt-in requires the vendor to earn explicit permission before a new data use begins. Most martech contracts default to the former.
The internal governance gap matters as much as the contract language. Legal reviews the original agreement. A marketing manager gets the renewal notice. A billing inbox receives the legal update. Nobody is assigned to connect those three things, so a notice that qualifies as formal under the contract’s terms fails to reach anyone who can actually evaluate it. The fix isn’t a new policy document; it’s a named owner, a routed inbox, and a vendor inventory that maps which data each platform touches. If a vendor’s proposed data use can’t be toggled off independently from the core product, that’s a renegotiation trigger, not an administrative update to acknowledge and file.
The renewal conversation is the moment of actual leverage, and most organizations waste it by treating it as a formality. A CMO heading into a HubSpot, Salesforce, or Adobe renewal now has a concrete reason to demand clearer definitions around enrichment, AI training, and secondary data use, and to push for opt-in as the default rather than an exception carved out only after public pressure. I’d revise this view if HubSpot’s opt-in commitment proves durable through its next product launch, but the structural incentive for platforms to monetize aggregated customer data hasn’t changed, and a goodwill pledge in a community forum isn’t a contract clause.
Concept deep-dive: Data enrichment
Data enrichment is the process of supplementing existing records with information from outside sources, think of it as a CRM vendor filling in the gaps on a contact record using data pulled from third-party databases or aggregated from other customers. The business value is real: richer profiles mean better segmentation and outreach. The governance risk is that enriched data blurs the line between what a customer supplied and what the vendor contributed, and those two categories can carry very different ownership, consent, and deletion rights under privacy law.
Based on reporting from What HubSpot’s Customer Data Reversal Teaches About Reading the Fine Print, originally published 2026-07-21 16:03:00.
