Add up the leads each of your marketing channels claims to have produced last quarter. The ads platform reports its conversions. The SEO report counts its form fills. The directory listing tallies its calls. Social counts its messages. Now compare the sum to the number of new customers who actually appeared. For nearly every business that runs this arithmetic, the claims total two or three times reality.
Nobody is exactly lying. Every platform grades its own homework, counts generously at its own boundary, and takes credit for customers who touched three channels on the way in. But decisions made on inflated channel math are real: budget flows to whatever measures loudest rather than whatever produces jobs. The source audit is the correction.
What an honest audit traces
The audit follows individual leads, not channel summaries, through four checkpoints: where the lead first arrived, whether it became a real conversation, whether it became a booked appointment, and whether it became completed, paid work. The unit of analysis is the lead's journey, and the journey is reconstructed from your own records: CRM entries, call logs, booking data, invoices.
This sounds obvious and is rarely done, because the trail usually breaks at the seams. The call that was never logged with a source. The form lead that became a phone sale under a different name. The walk-in who actually found you on Maps. Part of the audit's value is discovering exactly where your trail breaks, because each break is also where your ongoing reporting has been guessing. Closing those breaks, with call tracking, form source capture, and CRM discipline wired together, is what makes next quarter's numbers trustworthy without another audit. That instrumentation layer is the foundation we install before we will stand behind any performance reporting of our own.
The patterns audits keep finding
Certain findings recur across businesses often enough to expect them.
Branded search subsidizes everything. A chunk of what channels claim as their production is people who already knew your name, searched it, and clicked whatever surface appeared. The channel that created the awareness, often reviews, referrals, or trucks and yard signs, gets no line on any report.
The unglamorous channels overperform. Google Business Profile activity and review presence routinely produce more booked work than channels receiving ten times the attention and budget, which tends to redirect investment toward local presence once the numbers surface.
Some spend produces leads that never become jobs. A channel can genuinely generate volume that genuinely does not convert: wrong geography, wrong job size, wrong intent. Without the audit, the volume looks like success and earns more budget. With it, the pattern is visible by source, and the money moves.
Referrals are undercounted everywhere. They arrive by phone, mention no campaign, and get logged as direct or not at all, while being the highest-closing source on the sheet.
How far back to look, and how honestly to sample
Scoping the audit is its own decision, and two mistakes bracket it. Too short a window, one busy month, and the sample mostly contains fast-closing emergency work, which flatters the channels that produce it and understates the slow-burn sources whose jobs close next quarter. Too long, reaching back years, and the audit describes a marketing mix that no longer exists.
The workable frame is a window long enough to contain your full sales cycle plus a season: for most trades, the trailing six to twelve months, with the most recent weeks explicitly marked immature rather than silently counted as losses. Leads are judged by the cohort of when they arrived, and a cohort is not finished being judged until its typical closing arc has had time to play out.
Sampling honesty matters as much as window length. The temptation is to trace the leads that are easy to trace, which systematically over-represents form fills, the best-documented lead type, and under-represents phone calls, the worst-documented and often largest. If the records only support tracing a fraction, draw that fraction randomly across all entry points and say so in the findings, error bars and all. An audit that quietly measured only the measurable channels will conclude the measurable channels matter most, which is the exact circular logic it existed to break.
The deliverable should state its own confidence: here is what the records could prove, here is what they could not, and here is the instrumentation that would make next quarter's version definitive. That honesty is what separates an audit from an opinion with a spreadsheet.
From audit to operating system
A one-time audit is a snapshot; the real prize is making source truth continuous. That means every lead entering with a source attached, every job closing with its source intact, and a monthly view of the funnel by channel that the whole leadership team reads the same way. Once that exists, budget conversations change character: less advocacy, more arithmetic.
It also changes what you ask of your agencies, including us. A partner confident in their channel should welcome measurement that traces to completed work. That is how we prefer to be graded, and the lead generation programs we run are built to be legible in exactly this way.
If your channel reports have never quite matched your bank account, the gap has an explanation, and it is findable. Book an audit conversation and we will scope what your records can already answer, what needs instrumenting, and what your last quarter actually looked like source by source.