AI Sped Up Sales Prospecting. It Didn't Fix the Data Problem Underneath It.

Gong Enterprise AE Brian LaManna’s argument, in a new Gong Labs post: AI sales tools solved the wrong half of productivity. Generative AI makes drafting and research faster, but it hasn’t touched the actual bottleneck — reaching real prospects — so when contact data is wrong, “the extra speed just helps us fail sooner.” Bain research puts the ceiling on how much time reps spend actually selling at about 25%, a number AI could in theory double. But HubSpot data shows roughly 22.5% of B2B contact data decays and goes invalid every year, and from his own pipeline, LaManna estimates one in five to seven contacts are simply wrong — costing dozens of wasted account touches every quarter before a single AI-drafted email even gets read.

The cost compounds through a second mechanism that’s easy to miss: cognitive drain from switching between multiple data providers to verify what the CRM already got wrong. Apollo’s own research quantifies a version of this “toggle tax” — sales reps switch apps roughly 1,200 times a day, burning nearly four hours a week, and a Forrester study ties reduced app-switching to $5.1 million in value over three years at a 5,000-person org. Aviso AI makes the same diagnosis from the memory side: “Revenue teams don’t need another assistant to complete individual tasks… they need a unified intelligence layer that connects every signal,” arguing that context scattered across marketing, CRM, product, and CS tools is what actually breaks revenue AI, not a lack of models.

LaManna’s prescribed fix tracks both: not another bolted-on point tool, but integrated workflows where verified contact data and account context stay in view without tab-switching. The uncomfortable implication for anyone budgeting AI spend this year — a faster drafting tool sitting on top of a broken contact list isn’t a productivity win. It’s a faster way to burn a list.