
How AI improves go-to-market for services teams: Fix how deals are built first
The AI tools are getting better. The deals coming out of them aren’t. We dug into why that keeps happening across many services organizations we work with.

The project was scoped correctly, pricing held up in review, and the client signed. But, somewhere between kickoff and the third delivery milestone, the margins started slipping, with no early signal warning that anything was off.
This is a pattern most services leaders recognize but struggle to catch in real time. The deal looked right when it left sales. The breakdown happened in the handoff, the space between how work was quoted and how it was delivered.
Nobody planned poorly. It was just that the internal systems weren’t connected.
Quoting happened in spreadsheets or siloed tools. Scoping logic relied on senior people’s experience. Delivery teams inherited deals without the context or constraints behind them. Eventually, the misalignment caused a wrong resource mix, underestimated effort, and margin erosion mid-project.
This is the scope-to-delivery gap. And in a recent episode of Kantata’s podcast, Anish Udayakumar, SVP of Product and Customer Experience at Provus, joined Banoo Behboodi, SVP of Enterprise Customer Success at Kantata, to break down exactly where it forms and what fixes it.
Together, they walked through:
If you’ve watched well-sold projects lose margin mid-execution, the full conversation is worth watching. See it below, or scroll past it to read about the highlights.
Most services teams assume that if the deal was priced correctly upfront, margins should hold through delivery. But as Anish explained, the erosion almost always happens during execution, quietly, and without a clear signal until the damage is already compounding.
Project managers know their target margin. What they lack is detailed visibility into how that margin shifts over time, across resource changes, scope adjustments, and delivery realities that move week to week.
“From a process standpoint, it’s very difficult to isolate and fix the root cause of margin erosion. When you add all of these up across projects, it widens the gap between what was sold and what was delivered, adding more uncertainty to the organization as a whole.”
That’s the compounding problem. One project losing a few points of margin is a manageable issue. A portfolio of projects all losing margin for different, untraceable reasons becomes a structural drag and a serious business risk.
Banoo and Anish spent time on a problem many services organizations treat as a cost of doing business: estimation inconsistency.
“Typically, two individuals estimating the same project can arrive at completely different results in terms of pricing and effort. We see this often in large organizations, especially when onboarding new sales teams or introducing new processes.”
Scoping involves both an “art component” and a “science component”. The art is deal-specific—client context, relationship dynamic, strategic judgement. The science is effort, duration, and pricing—the parts that should produce consistent outputs regardless of who builds the quote.
The problem is that most organizations never standardized that science layer. Every estimator works from their own template and their own assumptions. The fix is building a consistent foundation underneath the human judgment, so the repeatable elements stay locked while deal-specific flexibility stays open.
One of the sharpest threads in the conversation was the tension between sales and delivery, two teams aligned at the business level but operating under competing pressures.
Sales is incentivized to move fast and close revenue. Delivery needs enough context, resources, and realistic timelines to execute profitably. When they work in separate systems with separate data, neither has visibility into what the other is optimizing for.
Banoo explained that unless information flows continuously between scoping, delivery, and resources, one can expect predictability to drop.
Anish expanded by describing four functions that must stay aligned: sales drives revenue, delivery executes, operations ensures resource availability, and finance protects margins. All four have different priorities. Which is exactly why they need a system that connects them, ensuring that scope, pricing, and resource decisions are made with full context.
“At a high level, you need connected systems. But, more importantly, you need to learn from past execution.”
That line from Anish reflected on the fact that many services organizations treat quoting and delivery as sequential. The quote is built, the project is handed off, and the two processes rarely inform each other again.
When a project runs over, that information stays locked in the delivery system. The quoting engine never sees it. Then the same wrong assumptions get baked into the next estimate, and the same errors repeat.
On the other hand, when real delivery performance flows back into the quoting process, every new estimate calibrates against outcomes instead of untested assumptions.
This is where Anish’s “island of automation” concept came in. Many organizations have optimized individual tools in isolation—a CRM here, a PSA there, a spreadsheet in between. Each island works on its own. But without a connection between them, the intelligence each generates stays siloed, and the organization keeps making decisions on incomplete data.
Anish’s closing advice was practical. First, recognize that unpredictability in services delivery is a systemic problem. No single team or process is the sole cause, and no isolated fix will resolve it.
Second, start with people. Foster deeper collaboration between sales and delivery so both sides understand the other’s constraints and pressures.
Then, to make that collaboration sustainable, invest in a unified platform that connects every persona, from the seller building the quote to the project manager running the engagement. When all teams work from the same data, the handoffs that create misalignment start to disappear.
The goal is to create a scalable path to predictable and profitable growth, where what’s sold and what’s delivered are finally the same thing.
The podcast closed with a live demo of the Provus-Kantata integration, the connected system in practice. The attendees could see how a structured Provus quote converts into a fully configured Kantata project with a single action. Total amount, hours, margin targets, and required resources are transferred automatically.
What sales quoted is exactly what delivery inherited. And once the project is completed, the details flow back into Provus, informing smarter scoping and pricing on future deals.
The AI tools are getting better. The deals coming out of them aren’t. We dug into why that keeps happening across many services organizations we work with.
Enterprise CPQ was “overkill”. Spreadsheets weren’t enough. Our partners made that clear, so we built an AI-powered quoting tool that filled the gap.
The real work behind overhauling a global professional services sales org