From Madhur Jain | Product & Market Analysis

Demand Generation Metrics: Who Still Discloses CAC Payback in 12 Filings

On this page

Of 12 public software companies whose latest quarterly filings we read, exactly 1 publishes a defined CAC payback figure. That company is Block, and the metric first appeared in its letters in the Q3 2026 shareholder letter. The demand generation metrics the other 11 offer are customer counts, retention rates and spending cohorts, and 9 of the 12 have added, dropped or redefined at least one of those since 2023.

Key takeaways

  • Only Block publishes a defined payback figure in a filing. Its Q3 2026 letter reports an estimated five-quarter payback for Square new volume and under three quarters for Cash App marketing, with both definitions in footnotes.
  • Wix names its TROI target but prints no number. Its Q2 2026 release says the Base44 target was raised, without stating the old or new figure or defining the metric.
  • 9 of 12 companies changed a customer or efficiency metric since 2023. Twilio dropped Active Customer Accounts in Q1 2026, Freshworks dropped its total customer count, and Atlassian and Samsara changed who counts as a customer.
  • The public software median of 30.5 months is an outside estimate. Meritech computes it from filings using non-GAAP spend and implied ARR, so it cannot be compared with a company's own payback definition.
1 of 12Companies in the sample that publish a defined CAC payback figure in a filed document. Source: company filings on SEC EDGAR, Q2 and Q3 2026.
30.5 monthsMedian CAC payback for public software in Q2 2026, computed by an investor, not disclosed by companies. Source: Meritech, September 2026.
9 of 12Companies that added, dropped or redefined a customer or efficiency metric since 2023. Source: company filings, compared by Zan Digital, November 2026.

Which demand generation metrics a public filing can contain

CAC payback is the number of months a company needs to earn back what it spent acquiring a customer. It is measured in gross profit, not revenue. A common company-level version divides sales and marketing spend by new recurring revenue multiplied by gross margin.

Every input in that formula is a choice. Which spend counts, which revenue counts, whose gross margin, and over which period. Two companies can both report "18 months" and mean very different things.

Public filings rarely carry the formula at all. What they carry is the raw material: GAAP sales and marketing expense, revenue, gross margin and some measure of customers or recurring revenue. Investors who want payback build it themselves.

This matters to anyone who owns a demand generation budget. Demand generation is spend that creates buying intent before a lead exists, as distinct from capturing leads already in market, a line drawn in the comparison of demand generation and lead generation. Payback is the one number that ties that spend to a return. It is also the number public companies are least willing to print.

Three families of metric appear in the sample. Acquisition efficiency covers payback and return-on-marketing targets. Spend intensity covers sales and marketing as a share of revenue. Customer outcomes cover counts, cohorts above a spend threshold and net retention. Only the first family answers the question a marketing budget owner actually asks.

The 12-company sample and the documents checked

The sample is 12 public companies that sell software or software-led services to businesses. They are Atlassian, Block, DigitalOcean, Freshworks, HubSpot, Klaviyo, monday.com, Samsara, Shopify, Toast, Twilio and Wix.

It mixes subscription SaaS with payments-heavy platforms on purpose. The payments businesses are where payback language has surfaced, and leaving them out would have produced a cleaner sample with a less useful answer.

For each company we opened the most recent quarterly earnings release or shareholder letter filed with the SEC as an 8-K or 6-K exhibit. For 10 companies that covered the quarter to 30 September 2026. Atlassian's fiscal Q1 2027 ends on the same date. Wix's latest release read was Q2 2026, and Samsara's was its fiscal Q2 2027, which ended on 1 August 2026.

To date each metric change, we compared those releases with earlier releases and annual reports back to 2023. Investor presentations and call transcripts were read only where the text says so. Where a figure comes from call coverage rather than a filed document, it is labelled as secondary.

The sample is not random. It leans toward companies with a visible marketing engine, which should make payback disclosure more likely rather than less. A result of 1 in 12 from a sample tilted in favour of disclosure is a strong signal.

CAC payback disclosure table: definitions, periods and metric changes

The table below is the core of this post. Read the third column for what you can verify about acquisition efficiency today. Read the fourth for how stable the surrounding metrics have been.

Acquisition-efficiency disclosure and metric changes, 12 public software companies
CompanyPeriod readAcquisition-efficiency metric in the filingCustomer or efficiency metric change since 2023
BlockQ3 2026.Square Payback Period: estimated five quarters for new volume added, year to date through August. Cash App Payback Period: under three quarters, first nine months.Both payback metrics added in the Q3 2026 letter. Absent from the Q4 2025, Q1 2026 and Q2 2026 letters.
WixQ2 2026.TROI named. No definition or value in the release.Base44 TROI target raised in Q2 2026. Size of the change not stated.
HubSpotQ3 2026.None. Reports 311,704 customers and $12,038 average subscription revenue per customer.Net revenue retention recalculated in the FY2025 10-K to remove Solutions Partner commissions.
monday.comQ3 2026.None. Reports net dollar retention of 108% and paid customer cohorts.AI ARR of $36M added. Total customer count no longer reported as a metric, only in boilerplate.
KlaviyoQ3 2026.None. Reports customers, ARR cohorts and net revenue retention of 108%.Cohorts above $250k and $500k ARR added since Q3 2025.
FreshworksQ3 2026.None. Reports cohorts above $5k, $50k and $100k ARR and net dollar retention of 104%.Total customer count dropped from reported metrics.
TwilioQ3 2026.None. Reports dollar-based net expansion of 114%.Active Customer Accounts discontinued from Q1 2026.
AtlassianQ1 FY27.None.Customers above $10,000 Cloud ARR redefined from unique domains to distinct legal entities.
SamsaraQ2 FY27.None.Core Customer threshold raised from $5,000 to $10,000, then to $25,000 ARR. Affiliates now grouped as one customer.
DigitalOceanQ3 2026.None.Customer count redefined in Q4 2025. Net dollar retention last given as 101%, for Q4 2025.
ToastQ3 2026.None in the release.No change found in documents read.
ShopifyQ3 2026.None in the release. Reports MRR.No change found in documents read.

Source: each company's latest quarterly earnings release or shareholder letter on SEC EDGAR, read 10 November 2026, compared with earlier releases and annual reports. "None" means no payback, magic number or return-on-marketing figure in the document read. It does not mean management never discusses one on calls.

What 12 filings say about acquisition efficiency Latest quarterly release or letter on SEC EDGAR, read 10 November 2026 Payback value Payback language Metric changed Block Wix HubSpot monday.com Klaviyo Freshworks Twilio Atlassian Samsara DigitalOcean Toast Shopify Dark blue: in the filing. Light blue: named, no value. Red: changed since 2023. Source: SEC EDGAR filings.
The left column is almost empty and the right column is almost full. Metric definitions moved far more often than payback was disclosed.

Block is the one filer with a defined CAC payback period

Block's Q3 2026 shareholder letter, filed as an 8-K exhibit, is the only document in the sample that names a payback metric, defines it and gives a value. It does this twice, once for each of its two businesses.

Square payback, defined in a footnote

The letter defines Square Payback Period as the time it takes for the gross profit Block expects sellers to generate to recover the cost of acquiring them. For self-onboarded sellers, the cost side is Square marketing spend, which excludes personnel, plus travel and related costs.

The reported value is an estimated five-quarter payback for Square new volume added, year to date through August 2026. The letter adds that self-onboarding, field sales and partner channels each ran in line with or better than Block's own return and payback thresholds.

Cash App gets a parallel metric. It divides acquisition cost per new or reactivated active by the expected quarterly variable profit per active. The letter reports less than a three-quarter payback on performance marketing and referrals over the first nine months of 2026.

Two features of these definitions deserve attention. The return side is expected lifetime gross profit, so the payback is a model output, not an observed recovery. And the cost side for Square self-onboarding excludes personnel, which shortens the payback relative to any formula built on total sales and marketing expense.

Why the metric arrived in Q3 2026

Payback is absent from Block's Q4 2025, Q1 2026 and Q2 2026 letters. It appears in the same quarter that the letter reports sales and marketing up 15% year on year on a GAAP basis, driven by go-to-market investment. GAAP sales and marketing was $687.9 million against total net revenue of $6.72 billion.

Management had used the language before. Coverage of the Q3 2025 call quotes 4 to 5 quarter payback periods for Square, and Block's November 2025 investor day deck appears to show Cash App payback by channel. We could not verify either from the documents, so treat both as secondary.

My read is that payback disclosure works as a permission slip. A company prints it when it wants investors to accept a higher marketing line, and Block is spending more. That is not a criticism. It is the most useful thing a company can do when asking for that permission, and the other 11 have not done it.

SaaS S&M metrics you can still compute from filings

If payback is missing, the fallback every analyst uses is sales and marketing as a share of revenue. It is crude, but every company in the sample reports both inputs on a GAAP basis every quarter. The ratio below is our own arithmetic from the release figures.

GAAP sales and marketing as a share of revenue, latest quarter read
CompanyQuarterS&M share of revenueSame quarter a year earlier
SamsaraQ2 FY27.42.8%.Not compared.
HubSpotQ3 2026.42.4%.43.9%.
FreshworksQ3 2026.40.1%.47.0%.
monday.comQ3 2026.39.5%.52.3%.
KlaviyoQ3 2026.37.1%.41.1%.
WixQ2 2026.32.3%.23.1%.
AtlassianQ1 FY27.22.0%.Not compared.
TwilioQ3 2026.13.9%.Not compared.
ShopifyQ3 2026.12.5%.14.4%.
BlockQ3 2026.10.2%.9.8%.
ToastQ3 2026.8.6%.8.8%.
DigitalOceanQ3 2026.7.5%.Not compared.

Computed by Zan Digital as GAAP sales and marketing expense divided by total revenue from each company's release. Freshworks states its own percentage. Prior-year figures use the comparative column in the same or prior-year release. "Not compared" means we did not extract the prior-year pair.

Sales and marketing as a share of revenue, latest quarter GAAP, Zan Digital calculation from company releases. Red bars: the two companies using payback language. Samsara42.8% HubSpot42.4% Freshworks40.1% monday.com39.5% Klaviyo37.1% Wix32.3% Atlassian22.0% Twilio13.9% Shopify12.5% Block10.2% Toast8.6% DigitalOcean7.5% Median of the 12: about 27%. Payments revenue inflates the denominator for Block, Toast and Shopify.
The ratio sorts business models before it sorts efficiency. Notice that the two red bars are the only two companies in the table whose ratio rose year on year.

The median across the 12 is about 27%, the midpoint of Atlassian's 22.0% and Wix's 32.3%. That number means little on its own. Payments revenue swells the denominator for Block, Toast and Shopify, and expense classification moves the numerator, a problem examined in the review of sales and marketing spend across 12 IPO filings.

The two that raised marketing intensity are the two that talk payback

Eight companies have a year-earlier comparison in the table. Six of them cut sales and marketing as a share of revenue. monday.com moved furthest, from 52.3% to 39.5%, and its sales and marketing fell in dollars too, from $165.8 million to $149.1 million.

The two exceptions are Block, from 9.8% to 10.2%, and Wix, from 23.1% to 32.3%. They are also the only two companies in the sample that describe marketing in payback or return terms in a filed release. That is a sample of two and proves nothing statistically.

It is still the pattern I would bet on. Companies cutting marketing intensity do not need to defend it with a payback figure, because margin expansion defends itself. Companies raising it reach for the one metric that frames spend as an investment.

Wix names TROI and keeps the number off the page

Wix uses TROI, which call coverage expands to time to return on investment, as its internal guardrail for marketing. Its Q2 2026 release states that the company is increasing its TROI target for Base44 and raising the threshold moderately because of Base44's margin profile.

The release does not define TROI, state the old target or state the new one. Specific month figures appear only in call summaries. Coverage of the Q1 2026 call cites an expected seven to nine months on roughly $90 million of acquisition marketing, which we could not confirm in a filed document.

A raised payback target is a decision to accept slower returns on marketing. Investors were told the direction of that decision and not its size. I would treat an undisclosed target change as the weakest form of disclosure in the table, weaker than silence, because it invites trust without giving anything to check.

How customer metric definitions moved since 2023

If payback is the metric companies avoid, customer metrics are the ones they keep editing. 9 of the 12 companies added, dropped or redefined at least one customer or efficiency metric between 2023 and the latest quarter. Each change was disclosed. Each one also breaks a time series.

Dropped: counts that stopped being reported

Twilio's Q1 2026 release states that beginning in the first quarter of 2026, it discontinued Active Customer Accounts as a key metric. The last figure was more than 392,000 accounts at 30 September 2025. Dollar-based net expansion stayed, and was 114% in Q3 2026 against 109% a year earlier.

Freshworks took a similar step. Its Q4 2025 and Q3 2026 releases report only cohorts above $5,000, $50,000 and $100,000 ARR, with the total count left to the boilerplate. DigitalOcean last gave a net dollar retention value, 101%, for Q4 2025. Its Q2 and Q3 2026 releases omit it, without a statement that it was discontinued.

Dropping a total count usually follows a deliberate move upmarket, where small customers matter less. That is a defensible reason. It also removes the one metric that showed whether the bottom of the funnel was still filling.

Redefined: thresholds and who counts as a customer

Samsara raised its Core Customer threshold from over $5,000 ARR to $10,000 in its FY2024 annual report, then to $25,000 in its FY2026 report. The FY2026 report also grouped affiliated entities under one parent as a single customer, and restated prior counts.

Atlassian's fiscal Q1 2027 letter redefined its count of customers above $10,000 in Cloud ARR. The old basis was organisations with unique domains, at 57,334 in Q4 FY26. The new basis is distinct legal entities, at 59,877 in Q1 FY27. The letter we read gives no bridge between the two.

HubSpot recalculated net revenue retention in its FY2025 10-K to remove the effect of Solutions Partner commissions. Under the new method 2024 was 101.8%, against 102.2% under the old. For 2023 the figures were 103.0% and 103.9%. The spread of current retention figures is covered in the comparison of net revenue retention across 12 Q2 filings.

None of this is misconduct. The problem is downstream: a series with a definition break is two series, and most investor dashboards draw it as one line. If you track competitors, mark every break before you plot anything.

Outside payback estimates: Meritech's 30.5 months

Because companies do not publish payback, investors compute it. Meritech publishes a quarterly median for public software in its Software Pulse. The Q2 2026 figure is 30.5 months, up from 28.4 months in Q1 2026.

Median CAC payback, public software, months Meritech calculation from company filings, Q3 2021 to Q2 2026 15 25 30 17.2 31.5 30.5 Q3 2021 Q1 2025 Q2 2026 Formula: prior-quarter LTM non-GAAP S&M divided by LTM net new implied ARR times gross margin, times 12.
The median rose 13.3 months, about 77%, over five years. The Q1 2025 peak of 31.5 months shows the latest reading is high but not a record.

The methodology is printed under the chart. It divides prior-quarter trailing twelve-month non-GAAP sales and marketing by current trailing twelve-month net new implied ARR, multiplied by non-GAAP gross margin, then multiplies by 12. Implied ARR is derived from reported revenue, not from a company's own ARR figure.

That formula differs from Block's on every input. Meritech uses all non-GAAP sales and marketing, Block's Square figure excludes personnel for self-onboarding. Meritech uses trailing recurring revenue growth, Block uses expected lifetime gross profit. Do not put Block's five quarters next to Meritech's 30.5 months on one slide. They measure different things and the comparison flatters Block.

The trend still carries information. Over the five years Meritech charts, the median moved from 17.2 months to 30.5 months. Public software now takes far longer to earn back a dollar of acquisition spend, measured consistently, than it did in 2021. That is the context in which 11 of 12 companies decline to print their own figure.

Where this analysis is weakest

The finding is narrow and it should be read narrowly. Three limits matter, and one of them argues the companies are right.

12 companies and one document each

Twelve companies is a sample, not a census of public SaaS. A different 12 could produce 2 or 3 disclosers. We also read mainly earnings releases and shareholder letters. Parts of the Shopify and DigitalOcean 10-Qs were not fully read, and Wix reported Q3 2026 results on the day this was written, which we could not check against the filing.

Filings are also not the only channel. Coverage of Toast's Q4 2024 call quotes a payback in the mid-teens of months, and coverage of Shopify's Q2 2024 call describes an 18-month payback guardrail for marketing. We could not confirm either in a filed document. If you count calls, the true number of companies discussing payback is higher than 1 in 12.

The case against disclosing payback at all

A single blended payback figure can mislead more than it informs. It averages fast self-serve channels with slow enterprise sales, and the inputs are judgement calls that management could tune quarter to quarter. Block's own definitions show how much room there is.

There is also a commitment problem. A metric printed every quarter becomes an obligation, and dropping it later draws more scrutiny than never printing it. Given the definition churn above, a cautious CFO has reasons to stay silent. I still think silence costs investors more than a well-footnoted number would, but the counter-case is real.

Frequently asked questions

What is a good CAC payback period for public SaaS?

There is no single standard, because every company defines payback differently. Meritech's median for public software was 30.5 months in Q2 2026, using non-GAAP sales and marketing, implied ARR and gross margin. Private-company surveys often report lower medians, but they use different formulas and samples. Compare a figure only with others built on the same definition and the same period.

Which public SaaS companies disclose CAC payback?

In our sample of 12 public software companies, only Block publishes a defined payback figure in a filed document. Its Q3 2026 letter reports about five quarters for Square and under three quarters for Cash App marketing. Wix names a TROI target in its releases but gives no value. The other 10 report customer counts, cohorts and retention instead.

How do you calculate CAC payback from a 10-Q?

Take sales and marketing expense from the prior quarter. Divide it by the increase in recurring revenue, annualised, multiplied by gross margin, then convert the result to months. If the company reports no ARR, derive it from quarterly subscription revenue times four. State every choice you make, because each one moves the answer by months.

Why do SaaS companies stop reporting customer counts?

The usual reason is a move upmarket, where a few large accounts matter more than many small ones. Twilio discontinued Active Customer Accounts from Q1 2026, and Freshworks dropped its total count from reported metrics. Companies replace the count with cohorts above a spending threshold. The cost is losing the clearest signal of whether new small customers are still arriving.

Is sales and marketing as a percent of revenue a good efficiency metric?

It is a useful starting point and a poor final answer. In our sample it ranged from 7.5% at DigitalOcean to 42.8% at Samsara, and much of that gap reflects business model rather than efficiency. Payments revenue inflates the denominator, and expense classification changes the numerator. Use it to compare a company with its own past, not with a different model.

What demand generation metrics should a SaaS board see?

A board should see payback by channel, with the formula written down and unchanged between meetings. Add sales and marketing as a share of revenue, new customers by cohort size, and net retention on a stated basis. When any definition changes, show both versions for at least two quarters so the break is visible rather than hidden in a trend line.

Where to start this week

Rebuild one competitor's payback yourself. Pick the closest public company to your model, take two consecutive quarterly releases from EDGAR and apply the Meritech-style formula above. It takes under an hour, and you will learn more about the inputs than any benchmark report will tell you.

Then write your own definition before your next board deck, not after. Name the spend you include, the revenue measure, the margin and the period. Block's footnotes are a usable template, and a definition fixed in advance is the only thing that makes a payback number worth reporting. If you report to a board quarterly, compare it with how public companies handle guidance in the SaaS earnings guidance scorecard.

Last verified: 10 November 2026. Figures will be re-checked after the Q4 2026 reporting season.

Reader and next step

Written for founders and finance directors who set their own marketing efficiency definitions. The single action: write your payback formula down in one sentence and keep it unchanged for four quarters.

References

  1. Block, Inc., Q3 2026 Shareholder Letter, 8-K exhibit 99.1, November 2026. Used for Square and Cash App payback definitions and values, and sales and marketing expense.
  2. Wix.com Ltd., Second Quarter 2026 Results, 6-K exhibit 99.1, 4 August 2026. Used for the TROI target change and Q2 2026 expense figures.
  3. Meritech, Software Pulse, 25 September 2026, September 2026. Used for the median payback series and methodology.
  4. Twilio Inc., Q1 2026 Earnings Release, 8-K exhibit 99.1, 2026. Used for the discontinued Active Customer Accounts metric.
  5. HubSpot, Inc., Form 10-K for fiscal 2025, February 2026. Used for the net revenue retention recalculation.
  6. Atlassian Corporation, Q1 FY27 Shareholder Letter, 5 November 2026. Used for the redefined customer count and expense figures.
  7. Samsara Inc., Form 10-K for fiscal 2026, March 2026. Used for the Core Customer threshold and customer definition changes.
  8. DigitalOcean Holdings, Q4 2025 Earnings Release, 24 February 2026. Used for the redefined customer count and last net dollar retention value.

The weakest part of this source base: sales and marketing figures for HubSpot, monday.com, Klaviyo, Freshworks, Toast, Shopify and Twilio come from their Q3 2026 earnings release exhibits on SEC EDGAR, which are not listed individually, and all call-transcript figures are secondary.

MJ
Madhur Jain
Writes for Zan Digital on B2B software economics and what public filings do and do not disclose.

Related reading