The market is noisy. Here’s how pipeline intelligence can help

Dave Kellogg called it a pipeline crisis in his recent piece, and he's not wrong. The Exit Five community is literally asking "what's working anymore?" And here's the harder truth: even when you find something that works, Andrew Chen's Law of Shitty Clickthroughs kicks in — whatever channel or tactic is working today gets saturated within weeks as everyone else piles in. You're running more campaigns, hosting more events, creating more content, yet somehow, it feels like nothing's working.

But here's what we realized: it's not that the tactics aren't working. It's that they're not working together. Every team has their own signals, their own tools, their own dashboards. Marketing knows who downloaded the whitepaper, sales knows who replied to an email, product knows who's active in the trial — but nobody has the full picture, so nobody can act with confidence.

We've felt the pressure too. Even with the advantage of getting more demo requests, the fundamentals of being at a hyper-growth startup don't change: the targets keep climbing, buying committees keep expanding, and the pressure to do more with what we have keeps mounting.

An overload of lead signals, in different places

Here's what was breaking: Our AEs were drowning in signals. They’d receive Slack notifications when someone signed up for our product, registered for an event, or even just visited our website. They had reports in Salesforce showing one piece of this picture, dashboards in Hex showing a complete puzzle, and Notion docs with summaries. The information overload made it impossible to focus.

Every signal felt urgent. None of them told our reps what to actually do. So they'd either chase everything (and waste time on cold leads) or ignore everything (and miss real opportunities).

Bringing the signals together for pipeline intelligence

That's when we realized: the pipeline crisis isn't just about "what channel works anymore" or "how do we get in front of people." It's about having too many signals with no way to make sense of them. You can't break through the noise in your prospects' inboxes if you can't cut through the noise in your own operations first.

Most organizations are treating this as a pipeline problem. I've come to realize it's actually a revenue intelligence problem. While executives debate attribution models and marketers scramble for budget, there's an opportunity sitting right there in your data team — or in your case, maybe it's sitting with you — if you're willing to evolve how technical and non-technical teams collaborate.

You probably have the pipeline — just solve for the signal overload

We stopped trying to send better alerts. Instead, we built one data app that brought every signal about each lead together in one place. This new data app became a live, prioritized view of accounts and the specific reasons they matter right now.

Every Monday morning, reps get a scheduled report, from this data app, in their inbox that tells them: here are your priority accounts for the week, ranked and explained. They can click back to the data app itself, which is live and always refreshable, but the weekly report gives them consistency. Focus means not chasing 10 different accounts every single day — it means committing to a set of high-priority accounts for the week and working them with intention.

The app pulls together:

• Product engagement: what leads are active in trial, what leads have dropped off

• Website visits: repeat visitors, pages viewed, depth of engagement

• Campaign engagement: event registrations, content downloads, webinar attendance

• Social interactions: LinkedIn engagement, community participation

• Hiring signals: new data leader hires, expanding data teams

• Opportunity data: previous conversations, past deals, who's involved

Then it runs everything through a scoring model and — this is the critical part — uses AI to analyze the results and provide actual recommendations. Not just "this account scored 87 points," but:

"This account is hot because they hired a new VP of Data last month, three people attended your webinar last week, and they've been actively exploring your documentation. Recommended action: reach out to the VP with relevant content about their specific use case."

The AI layer (we use Anthropic's Claude) looks at the combination of signals and explains why an account is prioritized. This reasoning is critical — it builds trust in the system. Reps aren't just following a black box score; they understand the logic and can personalize their outreach based on real context.

Why we built this instead of buying it

We looked at tools that promise to do this — intent data platforms, account scoring solutions, AI-powered sales intelligence tools. Some are genuinely good. But we realized something: we already had all the data. Product engagement, web analytics, CRM records, campaign interactions — it was all sitting in our data warehouse.

The question wasn't whether we could buy a solution. It was whether we should wait for a vendor to understand our specific use case, live within their platform’s confines, or build exactly what we needed ourselves.

Here's the other thing: I built this app on my own — and I’m not a data scientist. With Hex's Notebook Agent and my good-enough-to-be-dangerous coding skills, I built this in a day! That’s the amazing thing about our Notebook Agent. It was designed to write advanced SQL and help analysts build sophisticated data projects much faster and more accurately! I could ask the agent to help me write the scoring model in SQL, suggest Python functions for incorporating AI, and troubleshoot when things broke. I wasn't starting from scratch — I was collaborating with AI to build something tailored to our exact workflow.

You don't need to wait for another product demo, security review, or extensive onboarding to get your team the insights they need. You can ask just Hex (or whatever tool you use for agentic analytics) and learn as you build. The point isn't to start doing more — it's to start consolidating what you have and coordinating how you use it.

The results speak for themselves

The impact showed up fast: our lead-to-pipeline conversion climbed, which meant we were getting more pipeline from the same lead volume. Combined with other improvements across the sales process, the compounding effect was significant. We saw:

• 70% improvement in inbound conversion to sales opportunity using data apps that surface warm signals in real-time.

• 85% reduction in speed of execution on inbound follow-up, driving the increased conversion to opportunity because of a clear prioritization framework.

• We just hit our highest week of opportunity generation ever in the second week of December! It’s remarkable that during a season that we had expected to start slowing down, we sped up.

But beyond the metrics, there were some more fundamental shifts. The data team stopped being a service desk and started being strategic partners, our marketing team could be more prescriptive, and sales stopped chasing cold leads.

The most unexpected payoff? Instant credibility from industry-leading prospects

Aside from internal results with tangible pipeline velocity, we also saw an advantage that no one had seen coming: competitive, top-tier prospects were impressed with how quickly we followed up with them and told us that it increased their trust with us.

To directly quote a prospect:

”…Game recognizes game. Hex is clearly able to process a lot of data very quickly and that became clear to us when we visited your website and then immediately all three of us got emails asking to chat.”

This is the type of action that will help you cut through the noise.

Why this worked when everything else didn't

We'd tried a lot of things before this:

• Better Slack notifications (just more noise)

• Salesforce Reports (no strong way to summarize and prioritize)

• Training reps on how to use all the different tools (they didn't have time, and we couldn't get consistent adoption across AEs, especially as we were hiring quickly)

What made this new data app and its inbox notifications work, was combining consolidation with intelligence. We stopped expecting reps to be data analysts and instead gave them AI-powered recommendations they could act on immediately.

Instead of 47 Slack notifications, reps open one app in the morning and know exactly where to focus. Instead of guessing which account to call first, they have a priority list with reasoning behind it. And because it's the same coordinated view that marketing sees, everyone's finally working from the same playbook.

Now I maintain the scoring model and tune it based on what actually predicts conversions. I get feedback from the reps on what is helpful and what's not, and then I can easily make adjustments to improve the experience for everyone. The reps don't have to understand how it works — they just need to know who to call and why.

What you actually need to do this

You don't need to be a data platform company. But you do need three things:

1. Consolidated data in one place

Stop making reps check 7 tools. Pull your signals together — product data, intent signals, CRM data, engagement data. If you can't build it yourself, there are tools that can do this (we obviously use Hex).

2. A scoring model powered by AI, not just data

Scores alone don't tell you what to do. Layer AI on top to analyze the signals and provide recommendations. "Account scored 92" isn't helpful. "Account scored 92 because they just hired a data leader and attended two events" gives your rep something actionable. More importantly, it builds trust in the system because reps understand why they're being told to focus here.

3. Iterate and evolve it overtime

Your first version won't be perfect. What matters is building a feedback loop with your reps and then refining the model based on what actually drives conversions. I use Hex's Notebook Agent to analyze the app's performance — asking questions like "which signals are most predictive of closed-won deals?" or "are there engagement patterns we're undervaluing?" The agent helps me spot trends I'd miss manually and keeps the scoring model sharp without rebuilding from scratch.

What this means for the pipeline crisis

Dave Kellogg's recommendations are solid: focus on ABM, run better events, optimize for AEO, leverage partners. All of that works.

But every single one of those tactics requires knowing which accounts to target, when to reach out, and what message will resonate. That's where pipeline intelligence comes in — not as a replacement for good GTM tactics, but as the foundation that makes them actually work.

Because in a world where everyone's fighting for the same constrained pipeline, where every tactic gets saturated within weeks, and where buyers' inboxes are completely overwhelmed, the teams that win will be the ones that know exactly where to focus.

We made our signals talk to each other. We made our teams work from the same truth. We went from feeling like nothing was working to having a coordinated system that actually drives results.

And that makes all the difference.