That is half two of a three-part sequence on how HubSpot reworked with AI. Half one covers how we construct with AI. Half three is how we function as an AI-first firm.
Over the previous three years, now we have systematically rebuilt how we appeal to, have interaction, and delight clients by creating a brand new go-to-market mannequin. With AI, now we have added a whole bunch of hundreds of corporations to our whole addressable market, grown certified leads from reply engines by 1,850%, and now e-book over 10,000 conferences per quarter by way of customized outreach, with a 13% improve in win charge for offers the place guided promoting is used.
In the present day, we run an Agent-first GTM: a flywheel the place brokers are doing actual work at each stage and people are working with greater affect, connecting extra deeply with clients.
Appeal to: Discovering the best clients sooner
The highest of our funnel seems nothing prefer it did three years in the past. The place we as soon as relied on type fills, content material leads, and inbound chat groups, we now run a requirement engine powered by AI.
Rebuilding it required three bets: discovering the best corporations, changing those who confirmed curiosity, and displaying up the place patrons have moved.
The primary guess began with the Demand Agent. It identifies our Best Buyer Profile (ICP) and finds new corporations that match it. The agent enriches these contacts with indicators from a wide range of knowledge sources, and generates a prospect worth rating for each account: a prediction of each probability to shut and anticipated ARR. Final 12 months, Demand Agent added 345,000 accounts to our whole addressable market – accounts that reps would in any other case have lacked enough knowledge to pursue.
Subsequent we checked out automating the method as soon as a prospect reveals curiosity. We constructed Inbound Agent, a chatbot on our web site that handles 82% of all inbound chats with zero human involvement. The agent qualifies guests, handles aggressive questions, makes use of propensity scoring to establish actual shopping for intent, books conferences with our gross sales reps, and closes what it might probably. It’s now starting to promote HubSpot Starter when there’s a clear match.
The third guess was a few totally different form of purchaser completely, one who hasn’t raised their hand but, however is asking questions elsewhere. We moved early on Reply Engine Optimization (AEO), and constructed AEO Agent to make HubSpot seen and credible in AI-generated responses from instruments like ChatGPT and Perplexity. HubSpot is now essentially the most seen CRM in LLMs. Certified leads from AI-generated solutions grew 1,850% between Q1 2025 and Q1 2026. These leads convert at as much as 3x the speed of conventional search.
Interact: Enabling deeper buyer connection
Changing curiosity into pipeline is the place now we have invested closely. We constructed brokers and assistants at each stage of the gross sales movement, every instructing us one thing we didn’t anticipate.
The primary lesson got here from our Prospecting Agent. We assumed electronic mail sequences would do many of the work. They didn’t. Solely a small share of conferences get booked by way of electronic mail alone. So we rebuilt the agent to orchestrate throughout all channels: monitoring intent indicators, producing customized multi-touch sequences, and creating duties for reps on the proper second. In the present day, AI-personalized outreach books over 10,000 conferences per quarter.
The subsequent lesson got here from energetic offers. We began by constructing a single place the place reps may see every part a few deal like danger scores and similar-won offers. It was helpful, however we realized reps didn’t simply desire a dashboard. They wished to ask questions. So we constructed Guided Gross sales Assistant, a local conversational interface that lets reps interrogate their pipeline the way in which they’d ask a colleague: what’s the danger on this deal, how did we win in related conditions, what ought to I do subsequent? This context helps drive outcomes: we’re seeing a 13% improve in win charge for offers the place AI steering is used.
We additionally constructed for the analysis stage within the purchaser journey. A pre-sales agent handles advanced technical questions that will in any other case require a specialist. A Demo Agent spins up a tailor-made demo setting on the spot, personalized to the prospect’s particular business, geography, and firm dimension. These options take away friction at moments that used to gradual offers down.
Delight: Scaling success and help with AI
Once we introduced AI into the delight stage, help discovered product-market match nearly instantly. Prospects obtained sooner solutions, our group obtained capability again, and CSAT elevated. Our Buyer Agent now resolves roughly 60% of our inside help inquiries with out human intervention. For companies fascinated about the place to begin with AI, we suggest help. The outcomes are predictable and the trail to worth is the quickest.
However buyer success was a special story. The trail to AI was much less apparent and the result extra stunning. As a substitute of changing human connection, brokers have enabled deeper connection.
The issue we had been fixing was consideration. CSMs carry giant books of enterprise and might’t give each account the main focus it deserves. The query they ask most frequently is: which accounts ought to I concentrate on right this moment?
Our Buyer Success Assistant solutions that query. It identifies who wants consideration and why, drafts the outreach, and flags alternatives to drive deeper engagement. Greater than 80% of our CSMs use it each week. What we observed wasn’t simply effectivity. CSMs utilizing the assistant are having higher conversations, grounded in context, centered on outcomes, and taking place earlier than clients even must ask for assist. The outcome: a 7-point greater buyer save charge.
For patrons with out an assigned CSM, our Digital Success Agent guides them by way of the product, recommends subsequent steps based mostly on utilization patterns, and surfaces new capabilities.
The Agent-first GTM flywheel
That is what Agent-first GTM seems like in observe, and it continues to evolve with new experiments day-after-day.
-Apr-27-2026-04-39-33-1796-PM.png?width=650&height=366&name=image%20(8)-Apr-27-2026-04-39-33-1796-PM.png)
An important half is what it means for our groups. Entrepreneurs now attain every buyer with one thing related. Reps stroll into each dialog with full context. CSMs know precisely who wants them and why. Each group will get higher at engaging in their targets.
That’s as a result of no agent is working alone. Demand Agent expands who we are able to attain. The Inbound Agent converts those that have expressed preliminary curiosity. AEO captures patrons who had been by no means in our funnel in any respect. Prospecting Agent and Deal Assistant speed up the offers that matter. Buyer Agent and the Success Assistants preserve clients longer and assist them get extra worth. That’s the flywheel. And it will get stronger with each interplay.
We constructed the Agent-first GTM with twenty years of information, a whole bunch of hundreds of consumers, and a front-row seat to how one of the best groups on the planet market, promote, and serve. That’s why we’re sharing our method.
A notice on availability: Lots of the capabilities described on this sequence are a mixture of HubSpot-built instruments, third-party brokers, and customized integrations. Some can be found in HubSpot: AEO options, knowledge enrichment to drive demand, and brokers for buyer help and prospecting. Assistants like Guided Gross sales Assistant and our Buyer Success Assistant can be constructed as customized assistants.
We’re a bigger firm than most of our clients, and a few of what we’ve constructed displays that scale. That’s at all times been our mannequin: show it on ourselves first, then construct it for patrons. Our dedication to innovation means we be taught sooner, and people learnings go immediately again into the product so our clients don’t must determine it out alone.
