A Simple Example That Isn’t So Simple
Pulling on another thread from Conga Connect 2026, one of the more interesting conversations I had during the week was with Geoff Webb, VP Product and Portfolio Marketing at Conga. While much of the event focused on pricing, workflows, and AI embedded into revenue systems, our discussion kept circling a deceptively simple example that revealed a much larger issue.
We started talking about NDAs.
Not complex contracts or pricing strategy. Just NDAs. As Geoff pointed out, these are among the most common documents in any deal process, and yet they regularly slow things down. A rep sends one over, legal reviews it, questions come back, revisions happen, and what should be routine turns into delay.
This is exactly the kind of problem AI should be able to solve.
NDAs are structured. The risks are known. The variations are predictable. An AI system should be able to review the document, identify anything unusual, and allow most agreements to move forward.
But in practice, that acceleration rarely happens.
Not because the technology cannot do it, but because organizations are not aligned on when to trust it.
As Geoff framed it, the challenge is not whether AI can read the document. It is whether sales and legal agree on what “acceptable” looks like in the first place.
The Real Problem Isn’t Capability
This conversation was specifically about revenue workflows such as pricing, approvals, and contracting.
Across these workflows, AI is already capable of accelerating key steps. It can analyze deals, recommend pricing, summarize contracts, and surface risk in seconds as shown by AiMe at the conference. But the way those capabilities are used varies widely across teams.
Sales may push forward. Legal may slow things down. Finance may override. The same output leads to different decisions depending on who touches it.
That inconsistency is not a tooling problem.
It is a coordination problem.
From Zero Trust to Earned Trust
In response, many organizations default to a zero-trust mindset. Every output must be verified. Every decision requires oversight.
That instinct is understandable. Revenue workflows involve real risk, but when applied too broadly, it recreates the bottlenecks organizations are trying to remove. The more effective approach is not to eliminate trust, but to structure it.
If an AI system consistently evaluates NDAs correctly across hundreds of examples, and those outcomes fall within clearly defined boundaries, it becomes reasonable to allow that process to move forward automatically. Not blindly, but within guardrails.
This is where confidence matters.
If the system is highly confident that an agreement falls within known patterns, it can proceed. If confidence drops, or if something falls outside expected bounds, it is flagged and escalated.
Humans are still in the loop, but they are focused on exceptions, not everything.
That is how speed and control coexist.
What Narrative Alignment Actually Means
This is where narrative alignment becomes critical, and where most organizations struggle. Narrative alignment is not about messaging. It is about shared understanding across teams.
It is the agreement between sales, legal, finance, and operations on:
- what “acceptable” looks like
- where AI can act independently
- where human judgment is required
- how decisions should flow across the workflow
In the NDA example, narrative alignment means sales and legal agree that most agreements fall into a handful of acceptable patterns. Slight variations in language do not change the meaning. Only when something falls outside those patterns does it require escalation.
Without that alignment, every NDA becomes a legal review.
With it, most NDAs move forward automatically, and only the edge cases are flagged.
AI enables this by applying those rules consistently. But the rules themselves have to be agreed upon first.
This is the orchestration layer that sits across departments. AI can help carry context between teams, but it cannot define the boundaries of trust on its own.
Where Things Break Down
Problems emerge when organizations skip this step.
If every AI output requires validation, speed disappears. If too many decisions are automated without clear boundaries, risk increases. And if multiple AI-driven decisions are chained together without alignment, small errors can compound.
The issue is not whether to trust AI.
It is whether teams trust each other’s interpretation of how AI should be used.
Closing Thought
The conversation at Conga Connect reinforced something that is becoming increasingly clear.
AI is not just introducing new capabilities. It is forcing organizations to define how decisions move across teams.
The companies that move faster will not be the ones with the most advanced tools. They will be the ones that align on how those tools are used across workflows, where trust is placed, and where verification is required.
Because in the end, speeding up revenue execution is not just about automation.
It is about shared understanding.
And without it, even the simplest step in a deal, like an NDA, can still slow everything down.