KPIs and failure modes

The operating dashboard that tells you whether an enterprise motion is actually working, and the specific ways it quietly stops working

3 min read

Modelled on the same discipline this platform's other courses use for their own operating-health lessons: the founder's own numbers, not the client-facing metrics reported in a QBR deck.


The core dashboard

MetricWhat it tells you
Sales-cycle length, by deal stageWhether deals are actually progressing at the pace Module 1's benchmarks predict, or stalling at a specific, identifiable stage (frequently security/procurement review — Module 2)
Win rate on RFPs reached the finalist stageWhether the product and pitch are actually competitive once a deal is being seriously evaluated, isolated from top-of-funnel volume
Gross revenue retentionWhether the switching-cost mechanism (Module 1) is actually showing up as real retention, or whether accounts are churning despite the theoretical stickiness
Net revenue retention (expansion included)Whether existing accounts are growing over time — for a healthy enterprise motion, this is frequently where the real economics clear, not first-year contract value alone (Module 1)
Customer acquisition cost, fully loaded against the real sales-cycle lengthWhether the long cycle (Module 1) is being correctly priced into the unit economics, not assumed away
Client concentration (top account, top 5 accounts, as % of revenue)Whether a single lost account is a setback or an existential event — the same concentration risk that governs any relationship-dependent B2B business

None of these numbers is meaningful read in isolation once a quarter — they're a running dashboard, checked against the benchmarks established in Modules 1 through 3, not a one-time health check. Confidence tags for the underlying figures this dashboard draws on are as stated where each was first introduced in those modules; this lesson consolidates them into one operating view rather than re-arguing or re-tagging them.

Failure mode 1: the sales cycle that never actually closes

A pipeline full of deals "in procurement" or "in security review" for months past the benchmark length in Module 1 is not a pipeline — it's a set of deals that have quietly stalled, and the difference matters because it changes what to do about it. A deal genuinely on pace looks like steady progress through the RFP stages named in Module 2; a stalled one has stopped moving for a specific, nameable reason — usually no real champion (Module 2's role breakdown) or an unresolved compliance gap (Module 2's security lesson) — and needs a direct conversation about that reason, not more patience.

Failure mode 2: winning deals the compliance and delivery layer can't actually support

A sales team that oversells scope, timeline, or capability to close a deal creates a delivery-operations crisis one module downstream, and, because of the switching-cost mechanism working in reverse during a bad first year, a badly-delivered first enterprise contract can poison the champion relationship the whole deal depended on (Module 2), directly undermining the renewal and expansion economics Module 1 and this lesson's own retention metrics depend on.

Failure mode 3: the compliance posture that quietly expires or falls behind

SOC 2 Type II (Module 2) is not a one-time achievement — it's an ongoing attestation covering a defined observation window, renewed annually. A vendor that lets its compliance posture lapse, or that adds product surface area (new integrations, new data types) faster than its compliance documentation keeps up, is building a gap that will surface at the worst possible moment: mid-RFP, in front of a technical buyer specifically screening for exactly this.

Failure mode 4: chasing the enterprise AI hype cycle without the scope discipline Module 4 names

Given Module 4's own disclosed data, the single most avoidable failure mode specific to the AI angle of this course is pitching, or attempting to build, a fully autonomous, unscoped agentic capability instead of the narrowly-scoped, clearly-measured automation the same research shows actually reaches production. This is a discipline failure, not a technology limitation, and it's checkable directly against Module 4's own criteria before a deal or a build commitment is made.

Enterprise Software · progress saved in this browser · sign in to sync across devices

Up next

The kill-switch framework

A falsifiable set of thresholds for when to stop — stated in advance, so a real decision doesn't get made under the emotional weight of a specific deal or a specific sunk cost

3 min