Scoped projects that work from the data you already hold and come back with a specific, ranked set of moves, sized in dollars. From a focused search allocation study to a full omni-channel audit. C3 Metrics sells no media and earns nothing when a dollar moves between channels, so the recommendation follows the data rather than the revenue.
Greg Collins leads the analytics practice at C3 Metrics, and every engagement is led by him, across search, omni-channel measurement and media performance.
Forty years across four domains that rarely sit in one person. A Big 8 management consultant, then an investment banker with more than seventy completed transactions, then an operator: Senior Vice President at Reynolds & Reynolds, and chief executive of Basho Technologies. Enterprise software, automotive retail technology, adtech. Darden MBA, Williams College. That range is why a finding arrives already tested against the different questions a CFO, a CMO and a head of data will each ask about it, and why it holds up outside the analytics team. It is also the difference between a number that is technically correct and one that is useful.
He also leads Cape Fear Advisors, C3 Metrics’ sister firm, which provides independent structural analysis and operating advisory to private equity and management teams.
Every engagement opens with a review of existing relationships, and anything that could bear on our position or our findings is disclosed before the work starts. Beyond the fee for that engagement there are no other incentives, and no part of it depends on what the measurement finds. Engagements stay isolated from one another, and findings belong exclusively to the client, never pooled, benchmarked, or disclosed to third parties.
Three things set the size of an engagement: how much of the program it covers, how far down the stack it reaches, and what it has to rely on for data. The Search Allocation Analysis reads one channel at one layer, entirely from exports you already hold, which is why it costs what it costs. A continuous program covers every channel at all four layers, on signal collected independently rather than reported by the platforms being measured. Everything between those two is a position on those three axes, and the fee follows the position rather than the other way round. The entry engagements are narrow by design and run on a method already built and already validated, which is what makes them affordable. The same person conducts all of them.
Two ways in
Independent cross-platform analysis using your existing search exports. The SAA identifies which platform is more efficient, tests that finding for consistency across months, diagnoses whether your allocation reflects active management or default behavior, and delivers a specific, testable reallocation recommendation with three scenarios grounded in market data. Includes an AI test blueprint funded by the efficiency gain itself. No attribution infrastructure required.
A measurement read on what your stack can and cannot see about AI and agents. Which AI surfaces show up in your data and which are landing in “direct,” whether your window spans the journey where AI actually operates, and whether an agent-completed purchase could be told apart from a human one. You get what is visible today, what it would take to see the rest, and what we would measure once it can be seen. C3 Metrics sells no visibility optimization, which is what makes the verdict on it worth having.
Three ways deeper
Report-level analysis of the most recent three months across all paid channels — search, social, programmatic, CTV, affiliate. Identifies invalid traffic, saturation, and platform self-report discrepancies. The fastest path to a complete cross-channel picture.
Broader analysis window for more reliable saturation and trend findings. Preferred for programs with seasonal spend variation or recent budget changes. The same full-channel scope with the additional period needed to distinguish pattern from anomaly.
Measurement holds up only when the definitions do: what counts as a conversion, which spend belongs to which channel, how the sources reconcile against each other. This engagement settles those definitions with your team and delivers a one-time structured load of your own data, ready for analysis. It makes the tiers above possible where the data has never been organized for measurement, and it is the on‑ramp to a continuous program for clients who want one.
Why reliance matters. The third axis sets what a finding can claim, not only what it costs. An engagement working from platform exports is working in self-reported units, which compress the very gaps it is looking for, so it returns direction with confidence and magnitude with a stated floor. An engagement working from independently collected signal carries no such ceiling. Both are worth having. They are different claims, and we say which one you are getting before the work starts.
A word on fit. These projects are sized by what the spend is made of rather than how large it is. Roughly $4 million a year across digital, search, social, programmatic and CTV gives the analysis plenty to work with. The same $4 million in linear television gives it very little, because the signal is not there to read. The Search Allocation Analysis wants about $1 million in annual search for the month‑level comparisons to hold up, which is not a high bar for a program where search is doing real work. Tell us the channel mix and we will tell you directly whether there is enough to find.
Everything above runs as a project: scoped, delivered, finished. A program works differently. It stands up as its own entity inside the measurement system and collects continuously, with log‑level invalid traffic verification and a delivery cadence agreed at the start. Suited to ongoing performance monitoring, AI bidding programs that need independent measurement, and deal‑time commercial due diligence.
Because a program carries a standing cost in compute and management every month, it is sized against total spend rather than composition. Around $10 million in annual media is where it pays for itself comfortably. Below that the arithmetic gets tight, and we will say so. The project tiers above usually do more for less.
Invalid traffic (IVT) inflating platform-reported conversions. Industry rates: 25% desktop and mobile web, 39% mobile app, 25% CTV (Pixalate, Q2 2026, United States).
Channel saturation — spend past the point of diminishing return. In one engagement, saturation analysis confirmed approximately half of YouTube investment had crossed that line.
Platform self-report discrepancies. Signal verification for a major automotive advertiser surfaced a traffic anomaly platform reporting had not flagged — triggering a formal third-party investigation.
Search allocation inefficiency. Independent MTA reveals 6–23% more conversions at the same spend, validated across financial services, automotive, and food service. Platform self-reported data tells one of three stories: no gap visible at all, a gap too small to act on, or a gap that understates the true opportunity. The SAA identifies which scenario applies to your program and provides the directional finding. Full magnitude requires independent attribution.
Opportunity identified has typically run 15–25% of analyzed spend, as a combination of effectiveness and efficiency weighted to the objectives the client sets. The deliverable is not a single number. It is a set of specific opportunities, sized and ranked for testing and implementation against those priorities. Where the analysis finds little to move, that is a finding too: the program is already close to right, and you know it rather than assume it.
CPM benchmarking establishes whether the rate was competitive — not whether the spend produced a real return.
Agency audits are conducted by parties with a direct financial stake in the channels being measured. The conflict is structural, not incidental.
Platform-reported attribution adds up to more than 100% of actual conversions across any multi-channel program. The numbers can't all be right simultaneously.
Verified data is the prerequisite for any AI bidding initiative to produce reliable output. An independent audit establishes that foundation before the program runs.
"Traditional auditing asks whether the rate was fair. C3 asks whether it worked."
Every engagement is conducted from data the client already holds. No platform access credentials required. No attribution infrastructure. The Search Allocation Analysis requires only paid search campaign exports.
| Source | What's Required |
|---|---|
| Paid Search Google Ads, Microsoft Advertising |
Campaign-level exports, branded and non-branded segments. Minimum six months for consistency testing across time periods. Required for Search Allocation Analysis and all full-audit tiers. |
| Paid Social, Programmatic, CTV / Video, Affiliate |
Platform exports and DSP delivery reports at campaign level. Full-audit tiers only. |
| Direct Mail + Offline Linear TV, Radio |
Drop schedules and post logs where available. Full-audit tiers only. |
| Required Supporting Sources — Full Audit Tiers Only | |
| GA4 + Media Plan + Agency Reports | Conversion signal baseline; planned vs. actual spend; cross-reference for platform-reported figures. |
Most engagements begin with a 30-minute call to confirm fit and scope. The Search Allocation Analysis can begin within a week of receiving data. There's no obligation beyond the one-page agreement that starts the work.