The Bowtie Funnel: Instrumenting the Post-Close Half

Most SaaS teams can recite their acquisition funnel to the decimal: visitor to signup, signup to qualified, qualified to closed-won. Ask the same team for their onboarding-to-adoption rate, or their adoption-to-expansion rate, and you usually get a shrug and a gesture at "customer success." That asymmetry is the exact problem the bowtie funnel exists to fix. In a recurring-revenue business, most of the value is earned after the first sale — yet the post-close journey is typically the least instrumented part of the entire lifecycle.

The bowtie reframes the funnel as symmetrical: a wide-to-narrow acquisition funnel meeting, at the closed deal, a narrow-to-wide retention-and-expansion funnel. The right half — onboarding, adoption, expansion, advocacy — is a funnel with real, measurable stage-to-stage conversion, not a vague status field in a CRM. This post covers how to instrument that post-close half: what each stage's entry event is, how to measure the transitions between them, and how to join product behaviour to revenue so the funnel spans both. It extends, rather than replaces, the acquisition instrumentation in the full-funnel instrumentation guide.

What the bowtie is and why it exists

The bowtie funnel is a closed-loop model that mirrors the pre-sale funnel after the deal closes. As one framing describes it, the model "gets its name from its shape: a wide-to-narrow funnel for acquiring new customers connected at its narrowest point — the closed deal — to a narrow-to-wide funnel for retaining and expanding those customers." The closed deal is the centre of the bow, not the finish line. Everything to the right of it is the part of the customer relationship where subscription businesses actually make their money.

It exists because the traditional funnel's shape encodes a false belief: that the sale is the goal. For a one-time transaction that is true, but for recurring revenue the first sale is a down payment on a relationship whose value compounds through renewals and expansion. The bowtie makes that structural, forcing you to model and measure the post-close stages instead of treating them as an operational blur. It pairs directly with the way the open-versus-closed funnel guide defines conversion windows — the bowtie simply keeps counting past the point where a conversion funnel normally stops.

The post-close stages, defined as events

The right half of the bowtie usually breaks into four stages, each of which must become a concrete entry event before you can measure anything. Onboarding begins at the account's first login after close and ends at first value. Adoption is sustained, habitual use of the core action over a rolling window. Expansion is any seat, tier, or usage-based increase in what the account pays for. Advocacy is the account renewing and, ideally, referring others. Each of those is a timestamped event you can emit, not a subjective label a CSM assigns.

Defining them as events is what turns "customer success" into a funnel. Once onboarding-complete, adopted, expanded, and renewed are real events keyed to an account, you can compute the conversion rate between each pair — the share of onboarded accounts that reach adoption, the share of adopted accounts that expand — exactly as you compute lead-to-close on the acquisition side. The activation and adoption events here are the same ones covered in the activation-versus-adoption guide, now positioned as the first two stages of the post-close funnel rather than standalone product metrics.

Measuring stage-to-stage transitions

With each stage defined as an event, the post-close funnel becomes a sequence of conversion rates you monitor for account cohorts. Group accounts by close month, then track what fraction of that cohort reaches each subsequent stage and how long each transition takes. The onboarding-to-adoption rate tells you whether first value turns into habit; the adoption-to-expansion rate tells you whether habit turns into spend. A cohort that onboards well but stalls before expansion has a very different problem from one that never gets past onboarding — and only stage-level transition rates let you tell them apart.

Time-to-stage matters as much as the rate. An account that expands in month three is a healthier signal than one that expands in month eighteen, and watching the transition timing lets you forecast expansion revenue instead of merely reporting it after the fact. This is the same cohort-and-conversion discipline the acquisition funnel already uses, applied to the half of the journey most teams leave uninstrumented. The stages also connect to referral mechanics: advocacy is where the loop feeds back into acquisition, which is why it belongs next to the viral-coefficient instrumentation guide.

Joining product behaviour to revenue

The hardest and most valuable part of instrumenting the bowtie is joining two data worlds that usually live apart: product-usage events (logins, core actions, feature adoption) and revenue events (renewals, expansions, contraction, churn). The post-close funnel only works when a stage transition can be defined across both — expansion, for instance, is a revenue event that should be predicted by a product-usage pattern. If your usage data and your billing data never join at the account level, you can measure adoption or you can measure expansion, but you cannot measure the transition between them.

The practical build is an account-level identity spine that lets every user event roll up to the account and sit alongside that account's contract and billing state. That join is what makes "which usage patterns precede expansion" an answerable question rather than a guess. It is the same behaviour-to-revenue join problem covered in the full-funnel CRM-revenue join guide, here extended across the post-close stages so the bowtie spans both product and money end to end.

Why post-sale deserves pre-sale rigour

The core argument for the bowtie is economic: in a recurring-revenue model, the post-close half is where lifetime value is actually created, so under-instrumenting it means flying blind over your most valuable stages. Contentsquare's SaaS funnel research found that the highest-performing SaaS companies "treat post-sale marketing as equal to pre-sale marketing, measuring activation rate, feature adoption, NRR, and expansion ARR with the same rigor they measure pipeline and win rate" (Directive). The gap between those companies and everyone else is largely a measurement gap.

Applying pre-sale rigour means the post-close stages get named owners, target conversion rates, and dashboards — not just a health-score colour. It means net revenue retention and expansion ARR are read as funnel outcomes with upstream drivers you can instrument, rather than lagging totals you discover at renewal. The expansion-revenue mechanics themselves are covered in the expansion-revenue and NRR guide; the bowtie's contribution is to place them in a staged funnel so you can see what feeds them, not just what they sum to.

Building the post-close funnel without over-engineering it

You do not need a warehouse-native semantic layer on day one to start. Begin with the four stage events — onboarding-complete, adopted, expanded, renewed — emitted at the account level, and a single cohort view of the transition rates between them. That alone surfaces where accounts stall and gives customer success and growth a shared, quantitative language instead of anecdotes. Resist the urge to model a dozen micro-stages before the four core transitions are trustworthy; the same onboarding instrumentation from the onboarding-checklist analytics guide is enough to power the first stage.

Once the basic funnel is stable, layer in timing and prediction: how long each transition takes, and which usage patterns precede expansion so you can intervene before renewal rather than after. That is where the bowtie starts paying back beyond reporting — it becomes an early-warning system for churn and an opportunity radar for expansion. The discipline throughout is the same one that governs the acquisition side: define each stage as an unambiguous event, measure the transitions for cohorts, and keep the definitions frozen so the funnel stays comparable release to release.