According to ScaleXP’s 2025 benchmarks, the median CAC payback period for B2B SaaS is 15 months, and for the enterprise segment it stretches to 18-24 months. Yet most companies make budget reallocation decisions based on CPL and lead volume - metrics that have no direct connection to how much revenue each channel actually generates. Investing in a channel for 15 months and only then discovering it brought misfit customers with high churn is the standard outcome for teams without cohort analysis.
The problem isn’t a lack of data: CRM systems accumulate everything needed. The problem is that data isn’t sliced correctly. Marketing looks at CPL, sales looks at win rate, and nobody connects the acquisition channel to the customer’s long-term LTV. As a result, budgets are allocated by gut feel rather than revenue data.
This article covers what cohort analysis means in the context of B2B pipeline, which metrics to track broken down by channel, and how to set this up in practice - without a complex BI system to start. If you’re interested in B2B attribution more broadly, see our article on multi-touch attribution.
The channel problem that standard reports don’t show
A standard marketing report says: LinkedIn delivered 40 leads, Google Ads delivered 60, SEO delivered 20. CPL on LinkedIn is three times higher than on Google. The obvious conclusion: LinkedIn is inefficient, redistribute the budget.
But what if LinkedIn leads close with a 35% win rate while Google leads close at 8%? And what if the average deal size from LinkedIn leads is twice as high? Then the conclusion is the opposite - and the budget should be cut in the other direction.
Cohort analysis lets you see exactly this: not surface-level lead generation metrics, but the long-term behavior of leads depending on the channel, the period they were acquired, and other parameters.
What a cohort means in the context of B2B pipeline
In SaaS, cohort analysis is usually applied to product metrics: retention, churn, LTV by activation month. In B2B sales the logic is the same, but it applies to the pipeline.
A cohort is a group of leads united by a common attribute. In B2B sales, the most common groupings are:
- Channel cohort - all leads that came through LinkedIn Ads in Q1 2025
- Period cohort - all leads that entered the pipeline in January 2025
- Segment cohort - all leads from fintech companies with 50-200 employees
The point is to track what happens to each cohort over time: how they move through stages, what win rate they close at, how much revenue they generated at six, twelve, and eighteen months.
Key metrics for pipeline cohort analysis
Win rate by channel. The most basic metric, which surprisingly few people calculate broken down by source. If your CRM has correctly filled source fields, this is a twenty-minute report.
Time-to-close by cohort. Different channels produce leads that make decisions at different speeds. An outbound cold lead might close in ninety days, while an inbound lead (who found you themselves) might close in forty-five. This affects cash flow and how you plan pipeline.
ACV (Average Contract Value) by channel. It often turns out that an “expensive” channel produces larger deals. For a detailed breakdown of how to calculate real CAC per channel, see our dedicated article on real CAC by channel. - and revenue per lead ends up higher despite the high CPL.
Churn rate by acquisition cohort. If leads from a specific channel churn more often after six to twelve months, that’s a signal of a quality mismatch. The channel is attracting the wrong people.
Pipeline velocity by stage. At which stage do leads from different channels get stuck most often? If LinkedIn leads reach the demo stage with 70% conversion but close less often - the problem is in sales or the product, not marketing.
How to set this up in practice
Cohort pipeline analysis requires three things: clean data in the CRM, correct source attribution, and a tool for the cohort report.
Data in the CRM - the source/channel field must be filled when the lead is created, not retroactively. If your sales team fills the CRM inconsistently, or the source field is optional, the data will be dirty and the analysis meaningless. This is an organizational problem, not a technical one.
Source attribution - how did the lead get into the CRM? Did they fill out a form with a UTM? Did they message you on LinkedIn? Did they come through SDR outbound? Each case requires its own attribution logic. Simple UTM-based attribution is often not enough for B2B with a long cycle.
Reporting tool - at the basic level this can be a CRM export plus a pivot table. In a more mature setup - a BI tool (Looker, Metabase, Power BI) with cohort dashboards. To start, Excel is enough: group leads by channel and quarter of entry, see what happened to each group.
What to do with the results
Cohort analysis often reveals several typical patterns:
One channel consistently delivers leads with a high win rate, but it gets a smaller budget because the CPL is “high” - redistribute the budget.
Leads from a specific period (for example, after a particular campaign) churn significantly faster - dig into what changed in that period: the offer, qualification, segment.
Pipeline velocity dropped sharply for a given cohort - this can point to a change in the sales process or a new competitor in that segment.
The main value of cohort analysis is not pretty dashboards, but the fact that it shifts the conversation from “how many leads did we deliver” to “how much revenue did we generate.” That changes the relationship between marketing and sales - and that’s worth it.