Tools, KPIs, and kill switches

#The tool and vendor stack

Agency path. A broader service mix needs a broader stack than SMMA's single-channel tooling, but not Ad Agency's programmatic/ad-platform infrastructure: a CRM/project-management platform covering multiple deliverable types and client relationships at once (GoHighLevel, HubSpot, or a comparable all-in-one platform, priced on published vendor tiers), a content/design production stack, and an email/CRM automation platform if that function is in scope, which for this course's model it usually is. [Established] that vendor pricing tiers themselves are published and checkable; re-verify current pricing against the vendor directly before budgeting, since SaaS pricing changes without much notice.

Consulting path. Deliberately minimal, because the entire point of the model is near-zero delivery infrastructure: a scheduling and video-call stack, a document/deck tool for strategy deliverables, and — critically, per Where advisory-only fails — a contract and change-order process that makes the advisory/execution boundary explicit and easy to invoke in the moment, since the tooling gap in this path is process, not software.

#KPI gates, by path

Agency path, adapted from the concentration and utilization framework Ad Agency's own KPI lesson uses at a larger scale:

• Client concentration — no single client above roughly 25–30% of monthly revenue once past 4–5 clients; below that client count, concentration risk is structural and simply has to be managed by pipeline discipline rather than a ratio.
• Unbilled scope-creep hours, tracked explicitly per client per month — given the 57%-of-agencies-losing-$1,000–$5,000/month figure in Module 4, tracking this number at all (most agencies don't) is itself the KPI; a client trending upward month over month in unbilled hours is the leading indicator for a renegotiation conversation, not a client to simply keep serving harder.
• Team utilization in the 75–85% band, the same target Ad Agency uses — below it, capacity was hired ahead of a real bottleneck; above it consistently, the team is understaffed relative to committed scope, which is exactly how unbilled overtime and quality problems start.

Consulting path, structurally different because there's no team utilization metric to track:

• Hours-per-client against the retainer's priced scope, tracked weekly, not discovered at month-end. The 20–40 hours/month minimum-viable-engagement figure from The fractional-CMO model and market rates cuts both ways: fewer hours than that rarely produces real impact, and materially more hours than the retainer was priced against — the leading indicator of the failure mode named in Module 4 — should trigger a scope conversation immediately, not at renewal.
• A running log of any deliverable produced that wasn't in the original scope document, reviewed monthly. This is the direct operational version of the "if you've done execution work that wasn't priced, you're already inside the failure mode" test from Module 4 — the KPI exists specifically to catch the pattern before a full quarter has passed with the drift invisible.
• Renewal rate at the end of the minimum term, the honest measure of whether the strategic work was actually useful independent of whether extra execution work was quietly done to prop up the relationship.

#Kill switches

Stop and reassess the agency path if: fixed monthly burn exceeds a realistic 3-month revenue projection with no signed pipeline to close the gap; a single client crosses roughly 40% of revenue with no active plan to diversify; or unbilled scope-creep hours have grown for two consecutive months without a renegotiation conversation happening. [Directional], reasoned consistently with the concentration and burn-rate framing used across this platform's other agency-model courses rather than a single disclosed study for this specific tier.

Stop and reassess the consulting path if: actual hours delivered per client have exceeded the priced retainer scope for two consecutive months without a rate or scope adjustment (the direct, measurable version of the Module 4 failure mode); or the pipeline of prospective clients depends entirely on one referral source with no second channel developing, since the advisory-only model's near-zero delivery overhead is only an advantage if client acquisition is actually working, and a single-threaded pipeline is a different, quieter version of the same concentration risk the agency path tracks explicitly.

#What doesn't change between the two paths

Both are still, underneath the pricing and staffing differences, a services business that lives or dies on the trust relationship with the client's actual decision-maker — the same structural point Ad Agency makes about its own model applies here without modification: a technically excellent piece of work from a provider the client doesn't trust gets fired anyway, and a mediocre one from a trusted provider gets another quarter to improve. Nothing in either path's KPI framework substitutes for that relationship being real.