← All posts

The Broken Feedback Loop in Sales Orgs

Kevin Kho

Kevin Kho

Founder ·

There's a broken feedback loop in sales that nobody's fixed yet.

Until recently, I was an AI Engineer at Drata, building document processing pipelines, vector indexing, and agentic workflows into a compliance product at scale. A few projects touched the sales and marketing side too: automating competitive intel, pulling product gaps out of Gong calls. These were side projects, but we felt like there was so much potential there.

So a colleague and I started talking to product marketers outside the company. Dozens of conversations over a few months, originally just poking at whether there was something worth building around competitive intelligence. There was. But the more PMMs we talked to, something else kept surfacing. Content creation was painful, sure, but that wasn't the real issue. They had no idea what happened to their content after they hit publish. They'd spend weeks on a case study or a battle card, hand it off to sales, and then... nothing. No signal on what landed, what got ignored, what got adapted. Almost every one of them said the same thing: "I think the reps have their own version of this, but I have no visibility into what they're actually doing."

That was enough to send us to the AEs. And it turned out they had their own version of the same problem, just in reverse. They couldn't find the right content, didn't know what was current, and had zero visibility into what was working in other deals. Every deal started from scratch. Whatever they learned rarely made it back to the team in any usable form. And the further up you went, the less visibility anyone had into any of it.

Which brought us to the sales leaders. They had the dashboards, the CRM data, the pipeline reviews, but none of it told them what was actually happening. Lagging indicators, all of it. Close rates, won/lost, maybe a few notes in Salesforce. What they couldn't see was what moved a deal forward, where buyers got stuck, what a winning deal looked like compared to a losing one. By the time they had enough signal to act on, it was already over. They were managing outcomes, not opportunities.

It became clear: this wasn't three separate problems for three different personas. It was one loop, broken at every handoff. Content went out, nothing came back, and the next deal learned nothing from the last one.

What the Loop Actually Looks Like When It Breaks

We've heard this story enough times to recognize the shape of it.

A PMM builds a new slide deck for an upcoming product launch. They upload it to the shared drive, post it in the sales Slack channel, and move on, assuming this asset will find its way into new deals. With no real way to verify either.

A couple weeks later, they find out that most of the team is still using an older version. One rep has been working off a copy they saved to their desktop months ago. Another put together a trimmed-down version, aiming for something that would move faster. A third is using the new deck but has been swapping in custom quotes and tweaking slides on their own, because getting a tailored asset from the PMM usually means a couple days turnaround, and when deal momentum is on the line, it makes more sense to just handle it themselves.

But no one actually knows if their version is performing better, or worse. Not the reps, not the PMM. Nobody finds out which version closed deals, what content other reps are using, or what buyers actually spent time on, what questions it raised, or where it fell flat.

Why Deal Rooms Haven't Fixed This

Digital sales rooms were supposed to close this loop. One centralized, branded space where the buyer can find everything, instead of scattered email threads with attachments nobody can track. Buying committees are bigger now, deals are more complex, and the old way of selling through PDFs and follow-up emails can't keep up. The pitch was right. The execution wasn't.

In practice, the rooms became a chore. Some of the newer tools have gotten faster at setup, but the core workflow hasn't changed. Pick a template, drag in content, hope the rep actually uses it. And every time, the AE does the mental math: is this deal worth the setup? For the $200K enterprise opportunity, probably. For the $15K inbound lead that could still close this quarter? No chance. So the tool that's supposed to cover the pipeline only gets used on the biggest deals.

But here's the deeper problem with how these tools are built. Even when the rooms work as designed, buyer opens the link, browses the content, spends time in the room, the signal stays trapped in that single deal. Some tools surface engagement data to the rep, which is useful. But it doesn't tell you that every enterprise deal where buyers spent more than five minutes on the ROI calculator closed 30% faster. It doesn't tell the PMM that the case study they almost cut last quarter is the single most-viewed asset in won deals. It doesn't tell the sales leader which content is actually moving pipeline and which is just taking up space in the room. The signal exists. It just never goes anywhere.

And that's the fundamental design flaw. Deal rooms are isolated containers. They hold content for a single deal, and when that deal wraps, whatever happened in that room stays there. Nothing compounds. The tenth deal room your team builds isn't any smarter than the first one.

The Real Gap

The tool that fixes this won't be the one with the flashiest room. It'll be the one that actually learns.

Every deal your team runs generates signal: what content buyers spend time on, what questions they ask, where deals stall. Right now, all of that signal evaporates. It lives in rep anecdotes, Slack threads, and the occasional call debrief that never gets read.

The broken feedback loop isn't just a visibility problem. Every deal your team runs without learning from it is a missed chance to get better, and those add up. The teams pulling ahead aren't doing more deals. They're getting smarter between them.

That's what we built KnitKnot for. Deal rooms that close the loop, not just hold the content. More on that in Part 2.


Schedule a demo at knitknot.ai