You can’t operate a revenue team today without running into AI. It’s everywhere. And that’s not changing anytime soon because the tools are getting faster, cheaper, and more embedded into the systems teams already use. Which begs the question, should RevOps teams build, buy, or both?
You can’t operate a revenue team today without running into AI. It’s everywhere. And that’s not changing anytime soon because the tools are getting faster, cheaper, and more embedded into the systems teams already use.
Which begs the question, should RevOps teams build, buy, or both?
In this recap of a recent RevOps Co-op webinar, we explore this question in detail and break down how operators are thinking about that balance in practice. We also highlight where AI is delivering real productivity gains, where it is falling short, and unpack the people and process gap that is quietly undermining many AI initiatives.
Let’s dive in!
“What I’m seeing from some of our top customers and operations is the framework of owning your own data,” explains Ross Rich, CEO of Accord. “Then use other tools to operationalize it. Consolidate product usage data, conversation intelligence, and a deep understanding of why customers buy across segments, geographies, and products. From there, use that intelligence to power your outbound messaging, your pitches, your decks, and your deal reviews.”
At a high level, that approach looks like:
It’s also worth calling out that not every team is equipped to operate this way yet, and that’s okay. The companies doing this best tend to fall into two camps. Either they’re very small and have a technical founder or CTO building alongside the business, or they’re large enough to have dedicated resources focused on this full time. And most teams sit somewhere in the middle.
If you’re feeling left behind (especially after scrolling through LinkedIn and seeing posts about AI wins and transformational stories), you’re not alone. There’s a real sense in the business world today that everyone else is moving faster - but keep in mind, even among the most advanced teams, the gains are still incremental and you have time to catch up.
Before you decide what to build versus buy, it helps to zoom out and ask a more fundamental question: Where in the customer journey does AI actually make sense to invest?
As Ross puts it, it starts by identifying where the real opportunities are. “Are there specific parts of the customer journey that stand out as low-hanging fruit? Is it top of the funnel, post-sale, or deal execution in the middle?”
That lens matters, because not every part of the funnel carries the same level of risk or complexity. And the answer to that question should shape whether you build, buy, or leave something alone entirely.
From there, Matt Flotard, VP of RevOps at Gong, recommends pressure-testing your decision with a simple but practical framework.
That last question is often the tipping point. What looks like a quick internal build can quickly turn into something that should have been solved at a platform level. And this is where real world experience starts to shape how teams make these decisions.
For Matt DeLauro, President of SEON, the deciding factor often comes down to data quality.“If it’s a focused vendor that has high-quality first-party data, we’re far more inclined to buy,” he says. “Because the data has been the challenge when we’ve gone to build things.”
Matt shared an example of trying to build an AI-driven market research workflow internally. On paper, it made sense. In practice, it broke down quickly. The issue was not the model. It was the data.
“The quality of the data you’re going to use to train whatever outcome is super important,” he continues. “Because if it’s garbage in, you’re going to get garbage out of it.”
One of the biggest mistakes teams are making right now is assuming AI will have the same impact across every part of the funnel. But it doesn’t.
The teams seeing the best results are being deliberate about where it drives value and where it doesn’t. As Ross points out, the most effective teams are focusing their efforts on the parts of the customer journey where AI can meaningfully improve speed, scale, and decision making today.
Where AI is delivering results:
Where AI is still falling short:
As Ross puts it, “Are we going to have some agentic note go out to a key stakeholder in the middle of a deal? Probably not.”
That doesn’t mean AI has no role in deal execution. It just means its role is different. Instead of replacing human interaction, it works best as a support layer. It helps reps prepare, identify gaps, and move faster internally, rather than trying to automate the most critical moments with customers.
As AI adoption accelerates, another question is quickly becoming just as important as build versus buy: Should you cut a tool (or several) before adding a new one?
Instead of expanding their stack, many revenue operators are stepping back and asking a different question: Can we do more with what we already have?
As Matt DeLauro explains, much of the focus right now is on consolidation, not expansion. “We’re looking at whether we can do the same thing with fewer tools… and more features from the vendors we already have.”
This shift is being driven by two realities:
Every new tool adds:
And as Matt puts it, pulling tools out of your stack is rarely simple. “It can be like Jenga when you start picking apart a RevOps stack,” he says.
But from Ross Rich’s perspective, this shift is not just about reducing tools. It’s about rethinking how those tools work together. For years, many platforms operated as closed systems, trying to own as much of the workflow as possible. But that model is starting to break down. The teams getting the most value today are not looking for one tool to do everything. They’re looking for tools that work well together and share data seamlessly across the stack.
“You don’t need to be everything anymore,” stresses Ross. “You can add a lot of value by solving a key part of the problem and working well with the rest of the stack.”
By now, you’re probably wondering if CRMs are still relevant. The short answer is yes, but not in the way you might think.
For years, the CRM has been expected to do two things at once. It has been the system of record, where all your data lives, and the system of action, where reps are supposed to spend their time. But that model is breaking down.
As Matt Flotard points out, most teams are no longer trying to drive more activity inside the CRM. If anything, they are trying to reduce it. The goal is not to get reps to update fields or log notes more consistently. It’s to give them more time to actually engage with customers.
That shift changes the role of the CRM entirely. While it’s still critical as a system of record. It’s where your data is structured, where your reporting lives, and where all of your systems ultimately connect. But it’s no longer where the most important work happens day to day.
Instead, that work is increasingly happening elsewhere. In tools that support real-time conversations, deal execution, and customer engagement, where insights are surfaced in the moment and action is easier to take.
As Matt DeLauro mentioned, systems like Salesforce and HubSpot still power a huge amount of the underlying business logic. The integrations, the rules, the financial connections, all of that still matters, and it’s not something most teams are going to rebuild from scratch.
What’s changing is how visible the CRM is in the day-to-day workflow. As Ross notes, teams are already shifting away from thinking of the CRM as the place where work gets done, and toward a model where it quietly captures and organizes data in the background.
The takeaway? The CRM isn’t going away. Instead, it’s becoming less central to have revenue teams operate, but it still remains foundational to how everything works behind the scenes.
So where does this leave RevOps teams trying to decide what to build?
One of the most practical takeaways from the discussion is that not every use case is worth building, and not every problem needs a new tool.
In fact, the teams seeing the most success are starting smaller than you might expect. They’re not trying to rebuild their entire GTM motion with AI. They’re focusing on specific, high-friction areas where automation can make an immediate difference.
As Matt Flotard suggests, a good place to start is simply sitting with your team and understanding where time is being lost. That starts with asking:
Those answers tend to point you toward the right opportunities.
The use cases that make the most sense to build internally usually have a few things in common. They’re narrow in scope, relatively low risk, and focused on internal workflows. Think things like automating research, cleaning up data, or removing manual steps from repetitive processes. These are areas where you can move quickly, test often, and see value without introducing too much complexity.
On the flip side, there are clear areas where building tends to break down. Anything that sits in the middle of a deal, where relationships, timing, and context matter, is still heavily human. Trying to automate those interactions too aggressively can do more harm than good.
The same goes for use cases that depend on messy or incomplete data. As Matt DeLauro pointed out earlier, data quality is often the limiting factor. “If it’s garbage in, you’re going to get garbage out.”
And then there is the hidden cost of maintenance. Even if you can build something quickly, the real question is whether you can support it long term. Integrations break. Requirements change. What starts as a quick win can turn into something your team is responsible for indefinitely.
If a use case depends on high-quality, proprietary data and is tightly tied to a specific workflow, it often makes more sense to buy. If it spans multiple systems or solves a unique internal problem, building can be the right path, but only if you’re prepared to own it over time.
The goal is not to build more, but to build selectively, because in this new landscape the advantage goes to the teams that know exactly where to apply their tools and workflows.
At the end of the day, AI isn’t a silver bullet for rev teams and it’s not going to transform every part of your funnel overnight. What it can do is create an opportunity for you to rethink what you build, what you buy, and how your GTM system operates.
The teams seeing the most success are making deliberate decisions about what to own, where they need support, and how those pieces fit together. They’re also realistic about where AI works and where it doesn’t, meaning they’re focusing on the parts of the funnel where it can drive immediate impact, and they are not forcing it into areas where human judgment still matters most.
Remember, AI isn’t a shortcut. It’s a multiplier. So the goal is to be intentional about where you invest, disciplined about what you adopt, and clear on how it actually drives outcomes for your team.
.png)
.jpg)