The meeting always starts the same way.
Sales are down.
Marketing says the leads are fine. Sales says the leads are garbage. The agency blames the algorithm. Google reports that conversions went up. The CRM says they didn’t. And the owner sits at the head of the table wondering how five intelligent people can look at the same business and tell him five different stories.
Notice what’s happening in that room: nobody’s being deceptive. The agency isn’t hiding something. Your salespeople aren’t making excuses. The platforms aren’t cooking their numbers. Each of them is reading their own report accurately. The reports just don’t agree—and no one, including the owner, can see enough of the system to know which one matches reality.
THE SOFTBALL PROBLEM
Friends of mine play in our local community softball league. An evening at that fence shows this exact problem. Some nights, the scoreboard is run by a volunteer. She’s on her phone half the night, updating the board between conversations and snacks. She’s not trying to fix the game, but the players have learned that if they want their outs and runs recorded right, they check in with the ump between plays.
And every ballplayer knows the kicker: even then, the correction doesn’t always get accepted. You can ask to take an out on yourself, and the scoreboard still stays the way it’s been. That’s why, on the teams that take the game seriously, somebody in the dugout keeps their own book.
The score everybody plays by is not always the game that happened.
That’s the game of marketing. The agency keeps its scoreboard. Each ad platform keeps its own. The CRM counts whatever reaches it. Phone calls come in that land on nobody’s scoreboard at all. Nobody is rigging your numbers. Rather, nobody is watching them—not all of them, anyway.
Every one of those scorekeepers is doing their job. But every one of them is a volunteer in the same sense the teenager at the ballfield is: keeping the true score of your whole marketing operation is not actually their job. It’s yours.
And here’s where the ballfield has it better than you do. The league at least has one official scoreboard and an ump with the authority to correct it. Your marketing has five scoreboards, no official one—and no ump.
TWO SCOREBOARDS IN ONE INBOX
Imagine opening your CRM tomorrow morning and finding that it disagrees with itself. Two fields, same system, two different answers about where your money came from. That’s where this next story ends up. It starts somewhere completely ordinary.
A shed builder recently shared his marketing agency’s monthly reports. Not because he distrusted them. His dashboard showed different numbers than theirs, and he planned to cross-reference the two. Think about that for a second: the business had two scoreboards for the same season, and reconciling them had become leadership’s job.
A few weeks later, after a single call about lead tracking, a written list of questions landed in my inbox. Every one of them was a version of the same question: where do leads actually come from?
Then came the question that summarized the whole mess.
The CRM’s own integrations page showed the ad platforms weren’t connected. But open an individual lead, and the record said, “Lead Source: Paid Search.” The closest thing the business had to an official scoreboard was disagreeing with itself, and everyone in the conversation quoted it.
The agency’s response was honest, and I want to underline that. They explained which lead sources they leave out of their efficiency numbers, admitted they weren’t certain how one source was being attributed, and acknowledged there were channels in play they had no visibility into. That’s not incompetence. That’s a scorekeeper telling you, accurately, what their book covers. Each party can see one slice. Nobody can see the whole game.
THE EXPENSIVE REFLEX
When marketing goes quiet, the response is almost always activity. Someone flips settings. Someone relinks a system. New campaigns launch. A vendor gets replaced.
I understand the reflex. Doing something feels like progress. It feels responsible. But every change made while you’re still guessing muddies the record further. The original problem gets buried under a pile of well-intended adjustments. By the time anyone looks systematically, there are three problems instead of one. Only the original was real.
Underneath the activity sits a belief that this is a marketing-skill problem. Try harder. Hire different. Find the agency with the better pitch or stronger promise. It isn’t—and it’s easy to see why the belief comes so naturally. Nobody engineered your marketing stack. It accreted. A website from one vendor. A CRM the sales team adopted. Tracking code from a campaign three agencies ago. Each piece was added by someone hired to operate one slice, and each slice works. The scorebook that spans all of them was never anyone’s assignment.
WHICH BOOK ARE YOU TRUSTING?
Before you go looking for a fix, run a quick test on yourself. Say sales drop tomorrow. Sitting where you sit, could you answer these with confidence?
- Which campaign produced your last five customers?
- Which lead source produces your highest-value jobs?
- Do your CRM and your ad platforms agree on where a lead came from?
- Do your agency’s numbers match your own?
- Where, exactly, does the trail go cold between a click and a signed contract?
If those answers come slowly, or come back as, “Well, it depends on who you ask,” the problem in front of you isn’t a marketing problem, yet. You’re coaching a game without knowing the score. Everything after this point assumes you’d rather fix that than keep guessing at it.
KEEP YOUR OWN BOOK
Remember the correction that doesn’t get accepted? Ad platforms work the same way. You can’t email an ad platform and argue its attribution into accuracy. Their book is their book. And since there’s no ump to walk to between plays, the only move that actually works is the one the serious teams already use: keep your own book and keep it well enough that it becomes the one everybody trusts.
Practically, that’s an audit, and the discipline is not complicated. Stop changing things. Map what connects to what. Take one lead and trace it end to end—click, page, form, CRM record, phone call, quote, sale—and find the exact point where the numbers stop matching what you’d expect. Reconcile the competing books against that trace. Then test one variable at a time.
Attribution sounds like agency jargon, but it’s just this: connecting an outcome to its source. You already do it in the shop. You wouldn’t tolerate a stack of CDX that’s disappearing faster than you’re building barns. Your marketing deserves the same rigor as your job costing.
One more thing about a good scorebook: it counts the outs against you too. Sometimes the audit shows the agency was doing fine and the break was on your side—a form change, a snippet nobody updated, a process gap in sales. You want that answer just as much. A book that only records what flatters you is no better than no book at all.
WHAT THE BOOK TURNS UP
Two examples of where this lands.
A client’s conversion tracking stopped reporting. Weeks of guessing followed—settings flipped, explanations offered in every direction, nothing improved. When we finally sat down and traced it systematically, it took about 30 minutes to find: one embedded snippet using syntax that used to work before an ad platform change. The fix took five minutes. The damage had been compounding for weeks, not because the problem was hard, but because everyone was reacting. Nobody knew where to read or what to read.
A company’s salesperson was failing. He felt it. His manager was starting to wonder too. So, we pulled the numbers. His close rate on his contacted leads was as strong as it had ever been. What had dropped was the contact rate on the leads coming in—and the ticket sizes with them. That’s not a problem with sales. That’s a problem with lead quality that traced cleanly to where the leads were coming from. He was working harder than ever against a thinner game. The record didn’t just settle the question. It changed who had to act—and it took a weight off a good man’s shoulders that never belonged there.
Notice what both cases have in common. The answer was sitting in the data the whole time. The scarce thing wasn’t marketing expertise or sales technique. It was a trustworthy book, and it took discipline to read it before reacting.
WHEN EVERYONE TRUSTS THE BOOK
When there’s one book everybody trusts, the arguments simply stop having anywhere to live. Your agency conversations shift from explanations to observations—and a good agency loves that shift, because now their best work is visible instead of debated. The vendors worth keeping do their finest work for clients who keep a clean book.
Marketing spend starts behaving like the rest of a well-run operation. You watch the leading indicators—lead source, contact rate, cost per sold job—instead of finding out at month-end that something broke in week one.
The builders pulling ahead treat lead attribution the way they treat job costing: not as an administrative chore, but as the thing that makes every downstream decision trustworthy.
Nobody was rigging the game. Everybody was keeping part of the score—and nobody was keeping the whole of it.
Marketing systems are complex, and complex systems break—that part never changes. What changes is what happens after the bad month. The teams that thrive aren’t the ones that argue best at the fence. They’re the ones holding the book everyone else has to check against.
