Bet on the Team, Not the Tool: Three Converging Shifts Your GTM Motion Is About to Be Tested On

In late July, OpenAI launched Presence, a platform for deploying voice and chat agents inside large enterprises for support, outbound, and internal workflows. Within 48 hours, HubSpot had fallen 12.7 percent. Atlassian 11.8 percent. Workday 9.9 percent. Salesforce 7.7 percent. TD Cowen called the launch a likely major reason for a 3 percent drop in the IGV software index.

Right on cue, the hot takes followed. CRM is dead. SaaS is over. OpenAI is coming for your stack. All of it misses the point.

If your GTM motion depends on a specific piece of software to function, OpenAI is the least of your problems. The companies that panicked built their operation around a UI. The companies that did not spend five minutes on it built their operation around a result.

That distinction determines everything that follows: how you think about agent sprawl, how you show up to a buyer who has already shortlisted you, and whether the AI you deploy moves the number or just adds motion. Three things are converging right now. They look like separate stories. They are the same story.

The first sales call is now a validation meeting

94 percent of B2B buyers now use AI during the purchase process, according to Forrester's survey of nearly 18,000 global business buyers. Twice as many named generative AI or conversational search as their single most meaningful research source compared to any other channel, including vendor websites and salespeople.

Gartner's follow-up found 69 percent of B2B buyers now use sales reps specifically to validate insights they already got from AI. The rep is no longer the starting point of the buying journey. The rep is the endpoint of a process that already happened without you.

G2's research found 94 percent of buying groups have already picked their preferred vendor before speaking to a salesperson, and that vendor wins roughly 80 percent of the time. A salesperson now carries about half the shortlist influence of a chatbot. 69 percent of buyers chose a different vendor than the one they originally had in mind because of what an AI told them, and one in three bought from a vendor they had never heard of before an AI surfaced it.

Gong Labs adds the conversation-level view: since early 2024, a 250 percent increase in deals where buyers reveal they found the vendor through AI, a 280 percent increase in buyers disclosing they used AI as a buying advisor, and a 208 percent increase in buyers using AI to process RFPs and comparison matrices.

The implication is direct. When your rep walks into a discovery call today, the buyer has already consulted an AI about your product, your competitors, your pricing model, and your common objections. The rep who treats that as a standard discovery call, working through qualification frameworks and first-principles problem identification, is burning the one advantage they had coming in.

The rep who understands it comes in ready to validate, deepen, and address the specific things the buyer's AI research did not resolve. That is a fundamentally different conversation and it requires a fundamentally different preparation workflow.

INFUSE's 2026 research across 310 enterprise technology buyers found 47 percent now rank proven integration with their existing stack as the leading evaluation criterion, 32 percent want quantifiable ROI from comparable clients before committing, and 61 percent are evaluating, planning, or already cutting platforms as AI alternatives emerge. Established vendors are chosen over AI-native startups five to one when buyers state a preference. Buyers want the thing that integrates with what they already have and can prove it worked somewhere like them.

Agent sprawl is your sales productivity problem

Gartner's prediction: by 2028, AI agents will outnumber human sellers 10 to 1, and fewer than 40 percent of sellers will say those agents improved their productivity. Ten times more agents. More than half of sellers saying it did not help.

Gartner named the reason directly: AI agents are only as effective as the systems they operate within, and if those systems are fragmented, the agents will scale the fragmentation. That is agent sprawl. More tools, more prompts, more dashboards, more motion, less output. CSOs who overhaul data, automation, and user experience together are five times more likely to gain ROI from AI than those choosing quick fixes.

The Presence reaction is the same problem from the investor side. When a new AI product launches that overlaps existing SaaS categories, the market reprices the companies whose value was locked in the UI. Companies whose value is locked in the outcome are essentially unaffected, because the outcome does not care which tool produced it.

The teams winning right now are not loyal to tools. They are loyal to outcomes: pipeline velocity, conversion rates, revenue per rep. They pick whatever gets them there fastest and swap without flinching. The teams panicking built their whole operation around software instead of a result, so every new launch feels like a threat. Tools change. They always have. Teams that know what outcome they are solving for adapt in a week. Teams that do not spend six months evaluating the next thing and still do not move the number.

Three structural decisions before the convergence gets louder

Decision 1: Redesign call prep around buyer AI research

The standard call prep workflow is rep-centric. What do I know about this account, what questions should I ask, what objections should I prepare for. The new workflow is buyer-centric: what did this buyer's AI research tell them about us, what did it tell them about our competitors, what did it get wrong, and where is the gap between what they think they know and what would actually move them forward.

That gap is where the rep adds value in a validation-era first call. Practically: before any first meeting, run your own company, product, and competitive positioning through the tools buyers actually use, ChatGPT, Perplexity, Claude, and Gemini, and read what comes back. That is the pre-meeting briefing document that matters most right now.

Decision 2: Audit your agent stack for outcomes, not activity

For every AI tool deployed in your GTM motion, answer one question: what specific revenue or efficiency outcome is this responsible for, and do you have a baseline and a delta to measure it against?

If the answer involves activity metrics, emails sent, meetings booked, calls made, that is insufficient. Those are inputs. The outcome is pipeline influenced, conversion rate at a named stage, cycle length from first meeting to close, CAC per acquired customer. Count the tools. If the number is high and the shared data infrastructure is low, you have an agent sprawl problem, not an agent advantage.

Decision 3: Build your integration credibility story before the buyer evaluates you

With 47 percent of enterprise buyers making integration their primary evaluation criterion, your integration story needs to be findable before the first call, not delivered in a follow-up email.

Your website, your review site profiles, your partner directory presence, and your AI search footprint need to answer the integration question clearly and with evidence. What does your product connect to. What did integration look like for a customer similar to the one evaluating you. How long did it take. What changed after it was live. Buyers who shortlist you through AI will surface this before they book the call. The answer needs to exist in a form the buyer can find, not in a form you deliver after they ask.

The close

OpenAI Presence was not the first time the market repriced the tools and it will not be the last. The companies that spent that week figuring out what it meant for their CRM vendor made a category error. The companies that spent the week figuring out what it meant for their buyers made the right call.

The buyer is farther along. The validation moment is earlier. The agent sprawl is real. Organizations building around outcomes rather than tools will find every one of these shifts easier to absorb, because outcomes do not change when the tool does.