Start Here: How to Read This Report

In Brief

This report runs to forty-nine pages across ten sections, and almost nobody should read it front to back. It makes one argument: retention weakened across seven survey editions while a cost correction made the financials look better. This page routes you to the part of that argument that lands on your desk, either by how much time you have or by what you are accountable for.

Key takeaways

The report is one argument, not forty-nine topics: the revenue engine weakened, the response to it made the weakening harder to see, and the two now reinforce each other.
Two things confuse most first readers, and both have a fix: every claim carries an evidence tier, and the two most recent years in every survey edition are the survey’s own forecasts rather than results.
If you have ten minutes, read the four findings in Key Insights and stop. If you have an hour, add the Integrated View and the Playbook.
Read your own function or seat first, then read the one you negotiate with most often. The second read is usually the more useful one, because it explains why your first ask keeps getting refused.
Route by decision authority, not job title: if you can change how cost is allocated, how people are compensated, or who is accountable for what, use the seat list rather than the function list.

What This Report Is

Seven editions of the KeyBanc and Sapphire private SaaS survey, read as one continuous record rather than seven annual snapshots. That is the whole method, and it is why the finding exists: each edition on its own looks stable, and the decline only appears when the editions are laid against each other.

The argument has two halves, and the order is deliberate. The first half documents a revenue engine losing strength. The second half documents the operational response to that weakening, and shows that the response made the underlying problem harder to see rather than smaller. The Integrated View is where the two halves are put back together, and it is the single most useful page in the report for anyone who wants the argument rather than the evidence.

The argument itself is in sections II and III. Everything else exists to support it, and two sections are safe to skip on a first pass for opposite reasons: section IV is the verification, which only matters once you want to dispute a figure, and section VII is the prescription, which is premature until you accept the diagnosis. The full map, in the same order and under the same names as the navigation beside you:

The report's ten sections: what each one carries, and whether it is worth reading on a first pass. The argument is in sections II and III. Section IV is the verification behind it and section VII is the prescription that follows from it, and both can wait. Section names match the navigation labels exactly.
Section	What it carries	On a first pass
I. The Claim	The thesis, and the eight findings at one line each.	Read
II. The Evidence: A Weakening Engine	Findings 1 to 4. Churn, downsell, expansion, net dollar retention.	Read
III. The Evidence: A Response That Made It Worse	Findings 5 to 8. The cost correction, and what it did to the thing it was meant to protect.	Read
IV. The Proof: Check Our Work	Every figure, source and correction behind the findings, published so they can be checked rather than trusted. Opens with the story of what checking found.	Read the opening story; the registers can wait until you want to dispute a number
V. The Mirror: Find Your Company	Your own position on five certified segment cuts.	Read
VI. The Mechanism: Why Nothing Has Changed	Why the problem survives being understood, and the half-measures that keep failing.	Read
VII. The Prescription: What Working Looks Like	The system redesign, and the staged playbook for getting there.	Skip, until the diagnosis lands
VIII. Your Move: The Role Assignments	What each function and each seat does about it.	Read your own
IX. Reference	Metric definitions and glossary.	Look up as needed
X. Resources	Source editions, further reading, disclaimer.	As needed

Two Things That Will Confuse You

Not every claim here is equally strong, and the report says which is which. Each quantitative claim carries an evidence tier, from a figure verified against a source page to an interpretation labelled as ours. Interpretation is never presented as finding. How the tiers work is explained in the Introduction, and it is worth two minutes before you start arguing with a number.

The two most recent years in any edition are forecasts, not results. The survey closes mid-year and publishes in November, so the last settled year always sits further back than the publication date suggests. Years marked with an E are the survey’s own estimates. This matters more than it sounds: those estimates have a record, and the record is that they come in optimistic. The KPI Scorecard is where that pattern is handled directly.

Three Paths, By Time

Ten minutes. Read Key Insights. It carries the eight findings at one line each, ordered by strength of evidence, and it opens with the one paragraph to read if you read nothing else on this site.

An hour. Key Insights, then Integrated View for how the findings interact, then The Playbook for what to do. That sequence is the argument end to end without the evidence apparatus underneath it.

The whole thing. Follow the navigation in order. It is built as an argument, so section I sets the claim, II and III prove it, IV shows the working, V puts you inside the data, VI explains why the problem persists, VII prescribes, and VIII assigns. Section IV is the part most readers skip and the part that makes the rest defensible.

Read By Role

Two lists follow, and they are not the same list twice. The function list is for the people who run the work: it points at what the survey does and does not measure about your team. The seat list is for the people who can change the terms: budget, compensation, allocation, accountability. Route by what you can authorize rather than by what your title says. A founder without a formal C-suite is usually best served reading the CEO and CFO entries together.

If You Run a Function

Customer Success

then read Support

What you own here: Where your cost is booked, and whether anyone can see it.

Your function is defined as a revenue function and measured, seven editions running, as a capacity line inside someone else’s budget.

Sales

then read Customer Success

What you own here: Contract term, and what it does to churn after you have moved on.

Every instrument pointed at your function stops measuring at the signature. The strongest retention lever in the dataset is one you set and are not compensated on.

Product

then read Support

What you own here: Why customers leave, which no edition of this survey has ever decomposed.

You are measured on what shipped. Nothing in seven editions measures what happened to the customer afterward.

Marketing

then read Sales

What you own here: What share of your output goes to replacing revenue that already left.

You have the widest gap in the report between budget size and retention accountability. Your metrics stop at the close; the churn tax starts there.

Support

then read Product

What you own here: The earliest retention signal in the company, and the fact that nobody collects it.

Across every edition your function appears twice: as an accounting question about where its cost sits, and as an automation ranking. Neither is a measurement.

Professional Services

then read Customer Success

What you own here: The steepest churn gradient in the dataset, and no seat where it gets discussed.

Attach rate moves churn further than almost anything else measured here. Your function has no margin line, no headcount line, and no owner in the survey.

If You Hold a Seat

CEO or Founder

then read the CFO

What you own here: The mandate. Every fix below this line needs authority the function does not hold.

The half-measures in this report all fail the same way, and they fail at an altitude your team cannot reach. The case has to be made to you because the change can only be made by you.

CFO

then read the CEO

What you own here: How cost is allocated, which decides what the rest of the company can see.

The survey’s respondents are overwhelmingly finance. Its blind spots are a fair portrait of what finance functions can currently see, which makes this report a mirror before it is a benchmark.

CRO

then read the CCO

What you own here: A net number, measured with gross instruments.

You are accountable for retention-inclusive revenue and shown acquisition-only efficiency. The gap between the two is where the argument in this report lives.

CMO

then read the CRO

What you own here: Pipeline quality, currently validated at the wrong end of the funnel.

Quality is judged against close rate, never against what those customers do twelve months later. Three of the changes that would fix it are not yours to grant.

CPO

then read the CRO

What you own here: The roadmap, which is a retention document whether or not it is read as one.

The pricing-unit question is being handled as a pricing decision. It is a product build, with metering and packaging attached, and the window on it is not open indefinitely.

CCO

then read the CRO

What you own here: An outcome you are accountable for and roughly half the inputs that produce it.

Working harder inside that trap is the default move and it does not close the gap. The renegotiation is about accountability terms, not headcount.

Investor or Board Member

then read the CEO

What you own here: The valuation multiple, and the composition question behind it.

Two companies can report the same net retention and be opposite businesses underneath. Diligence that stops at the headline number cannot tell them apart.

Then Locate Your Own Company

Everything above routes you through an argument about the industry median. The median belongs to no one in particular, which makes it easy to agree with and hard to act on. Find Your Company puts your own numbers against five certified segment cuts, so you can see which of the eight findings apply to you with the most force before you read what to do about them. If you only click one link on this page, make it that one.

Frequently asked questions

How should I read The Retention Reckoning report?

Start with Key Insights for the eight findings at one line each. Add Integrated View and The Playbook if you have an hour. Read the full navigation in order only if you want the evidence apparatus underneath the argument.

Which parts of the report matter for my role?

The report routes two ways. Six GTM function entries cover Customer Success, Sales, Product, Marketing, Support, and Professional Services. Seven seat entries cover the CEO, CFO, CRO, CMO, CPO, CCO, and investors or board members.

Should I read by job title or by what I control?

By what you control. If you can change how cost is allocated, how people are compensated, or who is accountable for an outcome, use the seat list. If you run the work day to day, use the function list.

What should I read if I only have ten minutes?

Read Key Insights. It carries all eight findings at one line each, ordered by strength of evidence, and opens with the single paragraph the report would keep if it could keep only one.

Why do the two most recent years show as forecasts?

The survey closes mid-year and publishes each November, so the two most recent years in every edition print as the survey estimates rather than results. Those years carry an E suffix and have historically come in optimistic.