The Demand Engine · Ladder, rung two
Rung Two
Everyone gets stuck on the same rung. Free teaching is easy and the retainer is the goal, but the paid diagnostic in between is where most agencies have nothing built. This is the catalogue: 44 named engagement formats, what they deliver, and what they actually cost, with every published price attributed to the firm that published it.
5–15%
Price the diagnostic as a share of the work it leads to. Ten percent is the working default.
Baker & Enns · 2Bobs
3 options
Never deliver findings without priced implementation options in the same session. This is the whole conversion mechanism.
Every high-converting example
10–30×
What the name is worth. Outcome nouns price an order of magnitude above inspection nouns for the same work.
Observed across the catalogue
60–70%
Diagnostic to implementation, within 90 days. Below 50% means the wrong problem or the wrong price.
72Technologies
The name sets the ceiling
The sharpest finding in the research, and the cheapest to act on. Sort every published price in the catalogue by what the offer is called and the pattern is stark. Audit is a $500 to $2,500 word. Road Map is a $12,000 word. Same research, same client, same deliverable length.
Outcome and artifact nouns
- Road Map$12,000
- Positioning Sprint$10–30k
- Brand Sprint$15,000
- Messaging Sprint$25–45k
- Decision Pack£20–35k
- Stack Audit (exception)$15–25k
Inspection nouns
- Channel Check$499
- Accessibility Audit$1,250–2,750
- Content Audit$100–700
- Portal Audit£500–2,000
- AI Visibility Audit$500–2,000
- Productized SEO Audit$750
The martech stack audit is the informative exception: it keeps the word audit and still prices at $15–25k, because its headline finding pays for it. Thirty to forty percent of a typical stack is duplicate functionality, so the engagement funds itself out of cancelled licences before any implementation starts. When the finding is self-funding, the name stops mattering.
House rule. Name the engagement after what the client walks away holding, never after what we do to get there. An inspection noun is sometimes the only accurate word. When it is, the guarantee does the work the name cannot.
The catalogue
Filter by discipline or by price band. Every band is graded: A is a specific firm's published price, B is a named operator reporting their own numbers, C is an aggregator range worth treating as directional only. Starred formats are the ones worth copying structurally.
Benchmarks to hold ourselves to
| Metric | Target | Source |
|---|
| Diagnostic price as a share of the downstream project | 5–15%, default 10% | Baker & Enns, 2Bobs |
| Sell-through, of qualified opportunities | 40%+ | 72Technologies |
| Diagnostic to implementation, within 90 days | 60–70%+ | 72Technologies, DoodleWeb |
| Below this, something is broken | <50% | 72Technologies |
| Fee credited toward implementation, expiry window | 2–6 weeks | Sakas & Company |
| Deliverable length | 4–25 pages | Sakas, 72Technologies |
| Duration | 1–4 weeks | Multiple |
| Payment | 50 / 50, fixed fee | 72Technologies |
One number to sit with. Promethean Research surveyed 119 agency leaders in early 2026. Average net margin across the industry was 13%. Agencies that narrowed their service mix posted 30%, and grew 13% while doing it. That is the empirical case for a catalogue of defined offers over an open-ended capability list, and it is the reason this page exists.
How rung two fails
- Delivered without priced options. The single most common cause of the shelved deck. Findings alone ask the client "is this good value?", which nobody can answer. Three options at three prices asks "which is best value?", which anyone can.
- Scoped too broadly, cannibalising the project. Focus on specific initiatives, not comprehensive strategy. A diagnostic that answers everything has nothing left to sell.
- Priced below the seriousness threshold. The fee is a filter. Prospects deterred by it are usually poor fit, and a cheap diagnostic invites a cheap engagement behind it.
- Founder-dependent delivery. If only one person can run it, it is not a product. Two people must be able to deliver it before it goes in the catalogue.
- Commoditised by automation. One firm now sells a productized SEO audit at $750 delivered in 24 hours by stripping the scoping call and the presentation. Any audit whose inputs a machine can gather will meet that price. The defence is research a machine cannot do: interviews, buying-committee work, our own benchmark data.
- No proprietary method. A named, inspectable method is what makes the price defensible. Thirty-five checkpoints across six layers beats "we will review your account", and it can be delegated.
- Wrong problem chosen. Validate a new format against ten to fifteen prospect conversations before building it, not after.
The one that costs the mostAn audit delivered as a findings document, with the proposal to follow "next week", is the failure mode all seven of these collapse into. The findings and the priced options are one deliverable presented in one session. Every format in this catalogue that publishes a price also publishes what happens next, in the same artifact.
What we do not have
Honest gaps, so nobody mistakes silence for absence of demand. No credible published pricing exists for: attribution and measurement installs, incrementality testing, creative testing sprints, CRO sprints, executive thought-leadership programs, lifecycle and email audits, or analyst-relations work. In several of these the demand is visible and the pricing is not, which is an argument for us to publish first rather than a reason to skip the format.
AI search visibility is the clearest opening. Enterprises put an average of 12% of digital budget into it during 2025 and 94% plan to increase, yet the market's own stated obstacle is that it does not trust the measurement. Published audits sit at $500 to $2,000, which is exactly the scraped-dashboard tier CMOs are complaining about. A methodologically transparent diagnostic at $10–15k is not an aggressive price. It is the only price that answers the objection.