"Why CAC Payback Is More Useful Than LTV to CAC" — CJ Gustafson (Mostly Metrics)

Why this is in the vault

Direct follow-on to the 2026-05-11 CAC Payback mechanics piece, this time argued through a named practitioner (Datadog's SVP of FP&A) rather than CJ's own voice — the case for CAC Payback as the operating metric and LTV:CAC as the investor-facing metric you begrudgingly still report.

The core argument

CJ interviews AJ Ljubich (Datadog SVP of FP&A) on why CAC Payback beats LTV:CAC for actually running a business. LTV:CAC compounds too many sensitive inputs (churn, retention, gross margin) into one number that "can get results that are just silly and sort of hypothetical" for a high-retention business like Datadog — and when the number is bad, there's no single owner to fix it. CAC Payback ("Sales & Marketing costs / Revenue Additions, adjusted by Gross Margin") is tangible, benchmarkable against public comps, and decomposable — Datadog uses it by sales channel (enterprise field sales vs. high-velocity inside sales) and even to pressure-test operating decisions like letting top reps hold accounts longer. The concession: LTV:CAC still matters because it drives valuation multiple (a16z found 3x LTV:CAC businesses carry ~3x the multiple of 2x businesses), so it's worth tracking for investors even though it's "kinda useless" day-to-day.

Curation section

Run the Numbers podcast — new episode, "How Datadog Built a World-Class FP&A Team" (the AJ Ljubich interview this issue draws from), YouTube/Apple/Spotify.
Weekly Valuation & Efficiency Metrics — the standing Sunday public-comp dataset (Koyfin-sourced): NTM revenue multiples, CAC Payback Period, Revenue per Employee, Rule of 40, and OPEX-as-%-of-revenue across 9 named sector cohorts (Security & Identity, Data & AI Infra, Dev Tools & Observability, Horizontal SaaS, GTM/MarTech, Vertical SaaS, Take-Rate Platforms, Payments, Consumer Fintech/Crypto — ~130 companies total). Same recurring weekly-metrics block as prior issues; no new companies or definitions this week.

⚠️ Sponsorship

Two disclosed relationships this issue: (1) Abacum (FP&A/budgeting software) — lead sponsor block pitching reforecast automation, standard partner-content placement, no editorial bend into the CAC Payback argument. (2) Koyfin — recurring "data partner" affiliate credit on the weekly valuation/efficiency charts (?via=metrics affiliate link), same as every prior weekly-metrics issue. Both are on the known rotating sponsor pool; no new sponsor to flag.

Mapping against Ray Data Co

Reinforces, doesn't extend, the MAC discovery script decision from 2026-05-11. That note already concluded "what's your CAC Payback Period, gross-margin-adjusted and sales-cycle-lagged" beats "what's your CAC" as a discovery question. This issue adds the argument for why — LTV:CAC's sensitivity to compounding assumptions makes it useless as an internal fix-it lever, which is exactly the failure mode MAC would hit trying to coach a portfolio company off a single blended LTV:CAC number. No new action; strengthens the existing script rationale with an authoritative practitioner voice (Datadog FP&A) worth citing directly if the script ever needs a credibility anchor.
Channel-level CAC Payback segmentation is the more novel idea here, and RDCO doesn't have the surface area to use it yet. Datadog's field-sales-vs-inside-sales payback comparison assumes multiple paid acquisition channels running in parallel. Squarely and Sanity Check are both still organic-acquisition (per the 2026-05-11 note), so there's no channel mix to segment CAC Payback across — this is a "revisit when paid spend starts" pointer, not an immediate application.
LTV:CAC-drives-valuation-multiple data point (a16z: 3x LTV:CAC ≈ 3x the multiple of 2x LTV:CAC) is the one number worth banking if RDCO ever fundraises or is valued on Sanity Check/Squarely/MAC — it's a concrete argument for why the "investor-facing metric we don't operate off of" framing from the 2026-05-11 note still earns its keep in a pitch deck.

Mapping is medium: reinforces an existing decision rather than creating a new one, and the most novel piece (channel segmentation) doesn't apply until RDCO runs paid acquisition.

[[2026-05-11-mostlymetrics-cac-payback-calculation]]
[[2026-05-11-mostlymetrics-ltv-cac-nickelback]]
[[2026-05-04-mac-product-shape-decisions]]