[Un]Churned · 8 Jun 2026 · From the week of 8 June
Why Selling Software Isn't Enough Anymore ft. Chuck Ganapathi (Gainsight)
These are notes on the conversation, checked against its transcript. The episode itself has the full discussion.
In brief
Chuck Ganapathi, CEO of Gainsight, speaks with host Josh Schachter live at Pulse 2026 in Las Vegas. The episode centres on Gainsight's launch of Atlas, an AI-native managed service for renewals, and the company's shift from selling only software toward owning customer outcomes. Chuck argues that vendors should be willing to own the outcome, that build versus buy is a false choice, and that agents working with humans can make it economical to serve the long tail of customers. He also describes how humans should supervise agents and what pilot customers found hard, namely data readiness and the lack of playbooks.
For founders
- Chuck Ganapathi says selling software alone has made it hard to prove ROI, because the customer has to own and deliver the outcome, so he is moving Gainsight toward selling outcomes as a service.
- Chuck Ganapathi calls the build-versus-buy choice a false one and says customers should be able to build and buy on the same platform.
- Chuck Ganapathi says services can work when value comes from humans and agents collaborating, which he distinguishes from the low-margin labour arbitrage of traditional outsourcing.
- Host Josh Schachter quotes investor Jake Saper as saying domain expertise matters more in AI-native services than in software, because in services you are selling yourself rather than a product, and Chuck Ganapathi agrees.
- Chuck Ganapathi predicts that delivering outcomes through AI, rather than just providing software, will be the future of software.
For revenue leaders
- Chuck Ganapathi says most companies ignore the long tail of smaller customers, and that agents and humans working together can now cover that segment without adding headcount.
- Chuck Ganapathi's example is that if the long tail is 20 percent of ARR and renewals there improve by 10 percent, overall GRR improves by 2 points.
- Chuck Ganapathi says humans in an agent workflow should supply judgment, oversight and intervention when things go wrong, with guardrails refined over time.
- Chuck Ganapathi says today's smallest customers may become strategic accounts, so the long tail can be a source of future pipeline and expansion.
- Chuck Ganapathi says there is no established playbook for automating renewals in the long tail, and many customers are not yet ready for it.
What was said 15, most useful first
The build-versus-buy choice is a false one and customers should be able to do both. Listen
Chuck Ganapathi says the market frames the choice as either buying an inflexible piece of software that isn't tuned to your business or building it yourself, which is fun until you are stuck maintaining it. He says Gainsight lets customers build on its platform, which provides permissions, security, governance and data models, buy its prebuilt agents, or do both.
“we think that the right answer is that you should be able to build and buy”
Most companies ignore the long tail of customers and accept 60 or 70 percent churn there. Listen
Chuck Ganapathi says the usual pattern is that about 20 percent of customers drive 80 percent of revenue, and the best people focus on that high-touch group. He says the long tail is human-heavy and uneconomical to serve with people because contracts are small and renewals are numerous, so most companies ignore it and accept 60 or 70 percent churn, which he calls a huge waste.
“So what most people do is ignore the long tail and then accept a 60 or 70 % turn.”
Listen to the episode Retention & customer success Link to this Report a problem
Chuck Ganapathi gives an example in which a 10 percent renewal improvement in a long tail that is 20 percent of ARR adds 2 points to overall GRR. Listen
Chuck Ganapathi says the long tail may be only about 20 percent of ARR, but a 10 percent improvement in renewals there would be a 2 percentage point improvement in overall GRR, which he calls huge. He uses this to argue that long-tail renewals are worth managing even when individual contracts are small.
“if you can make a 10 % improvement in renewals in your long tail, that is a 2 % point improvement in your overall GRR”
Listen to the episode Metrics & finance Link to this Report a problem
Vendors that only sell software struggle to prove ROI because the customer has to own and deliver the outcome. Listen
Chuck Ganapathi says that in SaaS the ROI from a project has to be owned and delivered by the customer, and the vendor only supplies the software. He says most software companies have always found it hard to prove their product, and that this is why Gainsight is moving into services that deliver outcomes.
“the ROI has to be owned and delivered by the customer themselves”
Listen to the episode Strategy & market Link to this Report a problem
Chuck Ganapathi predicts that delivering outcomes through AI agents, not just software, will be the future of software. Listen
Chuck Ganapathi says the role of agentic software should be to deliver the customer's outcome, not to hand the technology to customers to execute. He says Gainsight launched Atlas, an AI-native renewals service that it sells as a managed service, and that Gainsight is no longer just selling software.
“We believe that's going to be the future of software.”
Listen to the episode Strategy & market Link to this Report a problem
Chuck Ganapathi describes three ways companies can answer the retention question: build, buy, or hire the vendor. Listen
Chuck Ganapathi says the question CEOs, CFOs and boards are asking is how to use agentic AI to drive higher retention, and that there are three answers. Build means creating agents on Gainsight's platform, buy means using its prebuilt agents such as its expansion agent, and hire means having Gainsight run the work as a managed service through Atlas.
“we think there are three ways that question can be answered”
Listen to the episode Strategy & market Link to this Report a problem
Chuck Ganapathi agrees that domain expertise, a technology platform and willingness to own the outcome are the criteria for an AI-native services business. Listen
Host Josh Schachter relays investor Jake Saper's view that these are the key criteria for AI-native services, and Chuck Ganapathi says Jake is absolutely right. Chuck says Gainsight meets them through 15 years of customer success work, an Atlas agent platform it has been working on for a little over a year with pilot customers, and a willingness to own the outcome and be paid on it.
“in an AI Native service business, having domain expertise is even more important than in a software company”
Listen to the episode Strategy & market Link to this Report a problem
The long tail can be a source of pipeline and a growth engine for the future. Listen
Chuck Ganapathi says that as new logo acquisition gets harder, companies must go back to their base and sell more, and small customers are a growth engine and can be a great source of pipeline. Host Josh Schachter adds AI-native companies as an example: they may be small in raw volume today, but their growth rates suggest they will become premier accounts.
“your long tail may actually be a great source of pipeline”
Listen to the episode Retention & customer success Link to this Report a problem
Tolerance for agent mistakes is very low because the work involves real revenue. Listen
Chuck Ganapathi says these agents are not handling customers of Gainsight's own, but real revenue and real dollar numbers, so the stakes are high. He says guardrails are refined and reinforced continuously because the agents keep learning, so the company has to rely on human intervention to make sure agents do the right thing.
“The tolerance for mistakes is very little”
Atlas pilot customers found that data readiness and agent-human workflow design were not yet in place. Listen
Chuck Ganapathi says that over the past year working with Atlas pilot customers, everyone wanted to use agentic AI to drive renewal outcomes, but the answer was not clear. He names the open questions as having the right data, knowing how agents and humans should work together, and deciding where the agent stops and the human picks up.
“People don't have the right data. People don't know how to get agents and humans to work with each other.”
There is no established renewal playbook for the long tail to automate. Listen
Chuck Ganapathi says that if you ask customers to automate renewals with a renewal agent and ask for their playbook, there is none for the long tail. He says the long tail has been ignored, and that many customers are not ready for this kind of automation.
“In the long tail, there's no playbook.”
Listen to the episode Retention & customer success Link to this Report a problem
A prebuilt expansion agent can handle the common part of an expansion analysis, with customers layering their own workflows and data on top. Listen
Chuck Ganapathi says Gainsight's expansion agent gives you 80 percent of the answer, and a customer can add its own specific workflow, its own data and third-party sources such as ZoomInfo or LinkedIn to take it further. He describes this as agents built on top of agents, similar to a mixture of experts, where each agent is an expert in one thing.
“gives you 80 % of the answer”
Customers can use Gainsight through external agentic tools such as ChatGPT, Gemini or Claude without logging into Gainsight. Listen
Chuck Ganapathi says Gainsight launched its MCP interface about a month before Pulse, and all its products now have one. He says this lets customers work in their preferred agentic tool and use Gainsight without ever having to log into it, and that the Gainsight CLI lets admins configure the product through coding agents.
“you can use Gainsight without ever having to log into Gainsight”
AI-enabled services differ from traditional outsourcing because they create value through humans and agents rather than labour arbitrage. Listen
Chuck Ganapathi says services have always been a low-margin business when all the work is done by humans, and that BPO companies have long done outsourced work. He says what is different now is that Gainsight's services combine humans and agents in value creation, which a traditional outsourcing company cannot match.
“what we're doing is not labor arbitrage”
Listen to the episode Strategy & market Link to this Report a problem
In human-agent collaboration, humans supply judgment, oversight and intervention when things go wrong. Listen
Chuck Ganapathi says the human relationship remains central to customer relationships, and in Gainsight's Atlas model humans will handle judgment, oversight and strategic intervention. He says humans in the loop let the company make the system's reliability predictable and put guardrails around what the agents do, and that agents and humans train each other in a learning loop.
“the humans will drive the judgment, the oversight, the strategic intervention when things go wrong”
Listen to the episode Retention & customer success Link to this Report a problem