Dispatch
The reports SaaS operators read before the board reads them.
Weekly analysis of earnings calls, churn cohorts, and GTM pivots — published before the LinkedIn consensus forms. Numbers you can cite. Narratives you can steal.
Reports Published
47
issues
Median NRR cited
112%
Series B–D
Avg churn delta
−1.4pp
Q1→Q2 2025
Readers this week
6,200+
operators
The data your CFO will ask about in Q2 planning.
Aggregate churn cohorts across 84 anonymised Series B–D SaaS companies, fiscal 2024–2025. Source: Dispatch Primary Research Panel.
Cohort Churn by Segment
n=84 companies| Cohort | Segment | M1 | M3 | M6 | M12 | YoY Δ |
|---|---|---|---|---|---|---|
| Q1 2024 | SMB <$10K ARR | 3.2% | 8.1% | 14.3% | 22.8% | +2.1pp |
| Q2 2024 | SMB <$10K ARR | 2.9% | 7.4% | 13.1% | 20.2% | −2.6pp |
| Q3 2024 | Mid-Mkt $10–50K | 1.1% | 3.2% | 5.8% | 9.4% | −0.8pp |
| Q4 2024 | Mid-Mkt $10–50K | 1.4% | 3.8% | 6.6% | — | TBD |
| Q1 2025 | Enterprise $50K+ | 0.3% | 0.9% | 1.8% | — | TBD |
Highlighted row: Mid-market cohort showing strongest improvement in logo retention.
Median ARR Bridge
FY 2024"Expansion revenue is doing more heavy lifting than new logos for the first time since 2021. The PLG cohort is the exception."
— Dispatch Analysis, Issue #47
Full breakdown: 12 additional cohort segments, regression analysis, and the three companies dragging the median down.
Continue readingThe NRR compression hiding inside every 'healthy' retention number.
When Salesforce reported 124% NRR last quarter, every VP of CS in your Slack celebrated. They shouldn't have.
The figure is calculated on a trailing-twelve-month basis using a cohort that excludes churned accounts after month three. That's a methodological choice that inflates the headline by an estimated 6–9 percentage points against the IAS 38-aligned calculation used by Workday and ServiceNow.
We ran the same adjustment across 23 companies that reported in the last 45 days. Median NRR on a like-for-like basis: 108%. That's still good. But it's not 118%.
The gap matters for three reasons: your board is benchmarking against the unadjusted number; your comp plan is tied to a metric that's not measuring what you think it's measuring; and your Series D pitch deck is about to get stress-tested by a tier-one fund that does this adjustment automatically.
"Every fund we spoke to runs the churn-adjusted NRR calculation before the first partner meeting. Most operators don't know it exists."
Marcus Oyelaran
Former VP Revenue at Intercom. Covered 200+ SaaS earnings calls. Now writes the macro lens for Dispatch.
−9pp
Median NRR overstatement when churn-adjusted using IAS 38-aligned methodology
Most-cited finding
NRR inflation of 6–9pp is systematic across SaaS companies using TTM cohort exclusions. The adjusted median for Series B–D is 108%, not 117%.
Priya Krishnamurthy
Ex-Goldman TMT analyst. Reads every 10-K footnote so you don't have to. Dispatch's resident contrarian.
$40M
Median ARR at which PLG self-serve CAC exceeds enterprise motion CAC in our panel
Most-cited finding
The PLG ceiling consistently appears at $38–44M ARR. Companies that pre-build enterprise sales infrastructure at $25M ARR show 34% higher NRR at $80M.
The PLG ceiling is real. Three companies hit it this quarter and no one said anything.
Product-led growth works until it doesn't. The inflection point — where self-serve CAC begins rising faster than ACV — is consistently appearing around the $40M ARR mark in the cohort we track. Not $100M. Not $200M. $40M.
Figma hit it at $38M and quietly built a 60-person enterprise sales team before anyone noticed. Notion hit it at $44M. Linear is at $31M and the signals are already visible in their job postings if you know what to look for.
The tell is always the same: average contract value starts compressing in the self-serve tier while enterprise deals above $25K ACV begin appearing in the pipeline. The company is bifurcating. The GTM motion is bifurcating. The comp plan hasn't caught up yet.
What makes this quarter notable is that three companies in our panel crossed the threshold simultaneously. The tactical implication for operators: if your PLG motion is generating more than 60% of new ARR and your average deal is below $8K, you have 18 months before this becomes your problem.
"PLG is a acquisition strategy, not a revenue strategy. The companies that figure this out before the board does win."
The archive.
Why three unicorns quietly fired their SDR teams in January
"Pipeline velocity, not volume, is the new north star metric."
Zendesk's hidden NRR problem buried in footnote 14
"The cohort data tells a different story than the headline number."
SMB SaaS churn is normalising — here's what the distribution actually looks like
"Median M12 churn dropped 80bps. The tail got fatter."
The ARR multiple compression narrative is wrong. Here's the correct read.
"Comps are distorted by two outliers everyone keeps citing."
What your board actually wants to see on slide 7 of the QBR
"Rule of 40 is table stakes. Bring the cohort waterfall."