How it works
Five numbers you already have, turned into the one number you can defend.
The Invisible Demand Diagnostic reads five signals off your own systems, scores how much of your buyers' journey you can actually see, and turns the result into a dollar gap built by subtraction from your own budget. Then it hands you a funded roadmap to close it. About 45 minutes. No new tools.
The four steps
01 · Report five numbers
From analytics, search console, CRM, your ABM platform. You enter them yourself. Nothing to install, no integration, no data project.
02 · Get your score
The five signals become your Demand Visibility Score, out of 25. Higher means more of the buying journey is visible to you. One deliberate rule inside the scoring: a dead signal drags the whole score down, because an entire channel you can't see is a blind spot, not a rounding error.
03 · See the gap
Your score maps to a benchmark: how much of your program budget belongs in influence work at your level of visibility. The benchmark runs from 15% at high visibility to 35% at low. The gap is that target minus what you spend today, in dollars. We also translate it into pipeline as a 2x to 4x range. Never a single tidy number.
04 · Get the roadmap
Eight influence-layer programs, ranked by your weakest signals and funded from the gap, with a funding level you choose (start small, build, or close the whole gap) and a 90-day leading indicator on every funded program.
The five signals
Each signal is read from a system you already own and scored against a published benchmark. They're listed here in weight order, heaviest first. The two at the top carry the most weight because that's where buying moved; peer influence carries the least because it's the hardest to measure precisely, and we'd rather under-weight the fuzzy signal than pretend it's exact. The specific weights and scoring thresholds are the model's working parts; clients get all of them in the workbook.
- 01
Anonymous ICP engagement
Right-fit accounts visiting and leaving without identifying themselves. The most direct evidence of buying you can't see. Source: your account-ID platform, or GA4.
- 02
AI / LLM discovery
Whether buyers researching your category in ChatGPT, Perplexity, and Copilot find you or a competitor. 94% of B2B buyers used AI in their most recent purchase (Forrester 2026). This one we audit live, on the call. Source: GA4 AI referrals plus a reproducible query audit.
- 03
Branded search gap
Brand interest rising faster than the demand your funnel captures. The visible tip of research happening somewhere else. Source: Search Console or SEMrush against your CRM trend.
- 04
Repeat evaluation
Buyers who come back again and again before they ever convert. Long evaluation means preference is forming off your funnel. Source: GA4, Mixpanel, or HubSpot.
- 05
Peer influence
How often your name gets passed around in communities, reviews, and referrals. Triangulated from three inputs so no single opinion moves it. Source: G2, LinkedIn, deal notes.
How the gap number works
The headline number is a subtraction. Target influence spend, minus current influence spend.
Target = your program budget × the benchmark for your score. Current = what you already spend on influence work, as you report it. Dedicated employee time counts at loaded cost; fractional, ad-hoc time doesn't, because it can't be benchmarked or relied on. Nothing is forecast. The only assumption is the benchmark.
And the benchmark is deliberately conservative. Research from the LinkedIn B2B Institute (Binet & Field) puts the optimal brand share of B2B budgets near 46%. Actual practice runs near 25%. Our targets of 15% to 35% sit between what companies do and what the research says they should do. If anything, the model under-claims.
If the gap comes out at zero, the report says so: you already spend enough, and the work is redeploying it, not adding budget.
What about pipeline and revenue? The gap translates to pipeline at 2x to 4x, always shown as a range. The report also shows the revenue at stake behind the gap, clearly labeled as the number with the most assumptions, which is why it comes last and never leads.
Measured vs. calculated
Measured
The five signals are measured. Read straight off your own systems. Not survey answers, not opinions. The AI audit is reproducible: run the same queries yourself tomorrow and you'll get the same result.
Calculated
The gap, the pipeline range, and the revenue at stake are calculated. Projections off numbers you gave the model, with every step visible in the workbook you keep.
You can trust the readings without swallowing the projections on faith, because you can see exactly how one becomes the other.
How to read the score
Out of 25, higher is better. 0 to 8 is Low Visibility: most of the buying that decides your deals forms where you can't see it. 9 to 16 is Moderate. 17 to 25 is High: your funnel captures most of what forms, and the job is protecting that.
Every signal is scored against a published benchmark, not a curve. Most companies don't score high, because most companies haven't built for how buying works now. Your report shows each benchmark and its source next to your reading.
What you should see by day 90
Score movement takes two or three quarters. So every funded program carries a leading indicator that moves inside the first one: AI query coverage rising, ICP engagement per post trending up, placements landed in target trade press, new verified reviews, branded search share holding or climbing.
That's the honest deal: we don't ask you to wait a year to find out if it's working. The quarterly re-score verifies the score is following the indicators.
What the model is not
- Not an attribution model.
- It doesn't claim to source pipeline to touchpoints.
- Not a forecast.
- The pipeline range is a one-period implication of the gap, not a multi-year projection.
- Not a tool sell.
- The recommended programs are mostly people, content, and cadence.
- Not opinion-led.
- Every recommendation is score-conditional and sequenced. Change the score and the recommendation changes.
- Not decimal-precise, and not pretending to be.
- The readings are exact. The projections are labeled estimates and shown as ranges.
See it on your own numbers
The first assessment is free while the founding cohort is open (25 companies; we publish the anonymized aggregate, which is why it's free). You enter your inputs through the diagnostic link, the model does the rest, and the full report lands within five business days. We walk through it together. The workbook comes with it, so every number is yours to check.