TheShortlist^™

Disclosure is a choice. When a company goes public, it picks which numbers to put on the first page and which to bury in a risk factor. That choice is readable.

We pulled the public filings of 14 winning SaaS companies and the only two venture-backed failures that filed publicly. Every winner reported customer retention as a headline number. Both failures avoided one entirely.

What the winners disclosed

Each row is what the company itself reported, in its own words.

Atlassian is the most honest disclosure in the dataset. They wrote directly in the F-1/A: "given the volume of our customers, we do not track the retention rates of our individual customers." That is itself a disclosure. They built a viral self-serve product where individual-customer retention is the wrong measurement framework. Saying so is harder than reporting a clean number, and they did it.

What Bird disclosed instead

After Bird went public in 2021, the first annual report used the word "retention" nine times. None of the nine is a rider retention number. The headline operational figures are Rides per Deployed Vehicle per Day and Ride Profit Margin (49% in 2021, up from 20% in 2020). Both measure how much money each scooter makes per day. Neither measures whether riders come back.

The omission is not an oversight. A rider-side cohort retention curve is buildable from any ride-marketplace's transaction data. Bird chose to lead with the supply-side numbers because the demand-side numbers looked worse.

What Better.com disclosed instead

Better.com's 750-page IPO prospectus uses the word "retention" 28 times. Every single instance is about executive compensation. CEO Vishal Garg received a $6,000,000 retention bonus, structured as a forgivable loan at 3.5% interest. Other executives received cash retention bonuses totaling roughly $20M, paid on the day the deal closed.

Better is a mortgage company, and mortgages are usually a one-time purchase. But the same filing mentions "refinanc" 114 times, and the company had launched home insurance and a real-estate-agent product by then. They could have shown repeat customer rate, cross-product attach rate, or borrower NPS. They had the data. They chose not to.

Why disclosure choice is signal

The simplest read of the pattern: companies surface the metrics that flatter them. When a SaaS company has 140% Net Dollar Retention, every page of the S-1 leads with it. When a scooter company has rider cohort curves that look bad, the 10-K leads with vehicle utilization instead. When a mortgage company is operating in a rate cycle that destroys repeat-customer economics, the prospectus leads with refinancing volume.

That makes absence readable. If a company that should have a clean retention metric is filing a 750-page prospectus and never discloses one, the most likely explanation is that the number is not flattering. The same logic applies before public filings: a Series A deck with no traction slide is usually a deck where traction is the weakest section.

This is not the same as Peter Thiel's "no real competition" tell, where a founder claiming zero competitors usually has zero customers. The retention-disclosure tell is more specific: a company that has been operating long enough to measure customer retention, and chooses not to surface that measurement in a regulated filing, is making a disclosure-strategy decision.

What this means if you are applying to a fund

Traction is the section most founders write last and weakest. The retention pattern in this issue is the empirical reason to fix that.

1. If you have a retention metric, lead with it. Net Dollar Retention if you are SaaS. Repeat purchase rate if you are consumer. Cohort survival curves if you are marketplace. The specific framing matters less than the choice to put a quantitative loyalty number in front of the investor.
2. If you don't have one yet, say so directly, and say what you will measure when you do. A founder who says "we are pre-traction, we will track NDR starting at first paid cohort" reads better than a founder who buries the question.
3. If your business model genuinely doesn't have a single retention number, follow Atlassian and disclose that. Explain the measurement framework you do use and why the standard one does not apply. Defensive vagueness about loyalty is the failure pattern. Honest framework-shifting is not.
4. Do not invent a "retention" metric that means something other than retention. Bird's "ride profit margin" reads as a unit-economics metric, not a customer-loyalty metric, and rebranding it as the latter on a deck would have been worse than disclosing nothing.

The disclosure pattern is downstream of a deeper one: companies that retain customers know they retain customers and want investors to know too. Companies that don't, don't.

Further reading. David Skok, "SaaS Metrics 2.0," on Net Dollar Retention as a leading indicator of long-term enterprise value. Bessemer Venture Partners, "State of the Cloud" reports, on the 130%+ NDR benchmark for top-decile public SaaS. Mary Meeker, Internet Trends 2019, on retention curves as the cleanest predictor of consumer-product survival. Sequoia, "Crucible Moments," on the operational disciplines distinguishing winners from failures.