The Pipeline Paradox
Here's what's happening across B2B right now, and I don't think enough people are connecting the dots.
Pipeline generation is up 23% industry-wide but win rates have dropped 18%. Optifai's 2025 benchmark across 939 companies found that enterprise deals over $100K ACV close at just 12 to 18%.
Source: Optifai 2025 Win Rate Benchmarks, 939 companies
The average B2B buying committee has grown to 8–13 stakeholders up from 6.8 in 2017 (attainmentlabs). 6Sense's 2025 Buyer Experience Report puts the average sales cycle at 10.1 months - and enterprise deals regularly stretch past 12 months.
So you've got longer cycles, bigger committees, lower win rates, and more pipeline than you know what to do with. Everyone's response? Generate more pipeline, run more campaigns and add more touchpoints.
But nobody is asking the harder question: do we even understand how our biggest deals actually close?
The 1,400 Touchpoint Deal
A demand gen leader I work with showed me why this matters. She pulled up a deal her team had just closed - a big enterprise retail company, and a major win.
She wanted to understand what actually worked. So she opened the customer journey in our attribution platform and started scrolling through 1,400 touchpoints across five years.
Her reaction was: "It was super interesting."
Journey Timeline view at Revsure across the span of 8 years
Think about that. "Super interesting" is what smart people say when their tools can't explain what they're staring at. It's the polite version of: I have no idea what happened here, but it looks important.
What she actually wanted wasn't to figure out which campaign gets credit. She wanted to understand what pattern repeats across their biggest wins - how to more systematically find and close enterprise opportunities like that.
We're optimising for more pipeline when we should be figuring out what conditions actually convert enterprise deals. Those are two completely different problems.
Your Model Wasn't Built for This
Normally attribution models - linear, time-decay, first-touch - were built for journeys with 5 to 20 touchpoints over 30 to 90 days. Prospect clicks an ad, downloads a whitepaper, talks to a rep, signs. Clean and traceable.
But when you stretch them across 1,400 touchpoints, the weaknesses become obvious.
Linear gives each touchpoint 0.07% of the credit, so the signal vanishes.
Time-decay overweights the last demo and erases years of relationship building.
First-touch credits a blog post from 2019 for a deal that closed in 2024. Your enterprise deals don't look like that.
Enterprise deals are messier and slower. They might begin with a conference booth conversation in year one, then a webinar series in year two, then a whitepaper download by a completely different stakeholder in year three. There can be six months of radio silence, a new VP who brings her own vendor preferences and restarts the whole evaluation, and then, finally, a budget cycle opens up and a peer recommends you at a dinner, and suddenly you're in procurement.
That kind of complexity is exactly why the numbers break down. Forrester found that 86% of B2B purchases stall because a stakeholder's concerns went unaddressed. Optifai's pipeline study across 939 companies showed enterprise deals progress through stages at just 55%, compared to 75% for SMB.
Photo: Excavation team at work. American Museum of Natural History (AMNH).
These aren't journeys. These are geological formations. You can't attribute a geological formation unless you do archaeology.
What I'd Do About It
I spent years in supply chain and ERP before building RevSure. In that world, you don't ship an item or take a purchase order until everything is connected end-to-end. A tweak to a warehouse in Barcelona can impact a hypermarket in Boston. Similarly to understand what worked and what are the surest way of success, you had to see the full system of enterprise GTM.
Here's how I'd apply that thinking:
1. Reconstruct the Full Timeline.
Take your 10 biggest closed-won deals from the last 3 years and reconstruct the full timeline.
Every touchpoint, every dormant period, every stakeholder change. Don't clean it up.
If your CRM only shows 12 months, pull from marketing automation and ad platform history. You want the raw sediment.
2. Look for the Sequences That Repeat.
Was there always a conference interaction before a demo request? Did the same content type show up 60 to 90 days before contract?
I've seen teams do this and find that three of their biggest wins had nearly identical engagement patterns in the 90 days before close.
Patterns invisible in a standard attribution report. Those aren't causes, they're simply conditions. And conditions can be engineered.
3. Look at the Dormancy.
This is where most teams throw away their best intelligence. When a deal goes quiet for 3 to 6 months, everyone deprioritises it.
But dormancy isn't absence of intent. It's intent waiting for a trigger. It could be a peer referral, news about a competitor getting acquired or may be a regulation that changed the urgency.
If you can name the trigger, you can manufacture it next time.
4. Build a Behavioural Signature.
When there is journey similarity to your historical wins. When a current account starts looking like your past big deals, that's when you deploy resources.
5. Run a Real Test.
Take your top 20 accounts showing early-stage versions of that pattern. Architect the missing pieces: the conference touchpoint, the deep-dive content sequence, the peer introduction.
Measure whether they convert faster than a control group you left alone.
That's the shift. You stop trying to attribute the past and start engineering the conditions for the future.
One Question Before You Close This
Do this today: Pull up your three largest closed-won deals from the last two years. Can you reconstruct the full timeline? The dormant periods, the stakeholder changes, what brought the conversation back each time?
If you can't, then you're part of the paradox. You're generating more pipeline while understanding less about how your most important deals actually close. The SDR who booked the final demo gets credit in Salesforce. The five years of formation that made the demo possible gets none.
Fixing that doesn't start with a better model or more pipeline. It starts with excavating what your data already knows but nobody has thought to ask.
Share this with someone who's still measuring enterprise deals with velocity-deal instruments.
P.S. At RevSure, when we onboard enterprise customers, we ingest up to five years of historical data across CRM, marketing automation, ad platforms, and sales engagement systems. It helps us to reconstruct what actually happened at the account level and surface the patterns that repeat across wins. If you want to see what your archaeological record looks like, mail me at [email protected]