Suppose marketing reports a good month: more leads, lower cost per lead, campaigns ahead of plan. Sales sees fewer conversations worth pursuing. Someone has to decide whether to increase spending, change the campaigns, or fix what happens after an inquiry comes in.
Those reports can both be accurate. That’s what makes the disagreement worth investigating. A dashboard that puts the two numbers next to each other still leaves the important question unanswered: what happened between the response and the conversation?
I’d start by asking each team to show me the records behind its claim. Marketing may be counting submissions that arrived this month. Sales may be describing accounts it accepted, including inquiries from an earlier period. One person filling out three forms may appear three times in one report and once in the other.
Until those differences are visible, the team can spend a lot of energy defending numbers that answer different questions.
Follow one group of inquiries
Pick a campaign and an entry period. Choose a group old enough for the next sales steps to have happened, then follow those inquiries into the CRM. Keep the dates, the acceptance decision, the next action, and any recorded reason the account stopped progressing.
Keep unmatched records visible too. If an inquiry disappears between systems, that’s a finding. Quietly dropping the row makes the remaining report easier to read and the underlying problem harder to see.
Consider an illustrative account that submitted a form after clicking an ad. The campaign recorded a conversion. Sales marked the account unqualified because it needed a capability the product doesn’t offer. The original conversion was real. So was the mismatch. Now there’s something specific to inspect: did the ad or page imply that capability, or did the visitor arrive with an expectation the page never corrected?
A suitable account that waited a week for a response tells a different story. So does an account that sales accepted but the report missed because a field was never updated. These cases don’t belong in one bucket called “bad leads.”
Let the finding change the assignment
If the inquiries are outside the market you serve, look at targeting, search terms, and the offer. If suitable buyers aren’t getting a response, inspect the handoff and ownership. If the report loses track of accepted accounts, fix the record or definition before using it to cut a campaign.
A small review gives you hypotheses to investigate. It doesn’t establish how common each problem is. Once you see a plausible pattern, check it across more records and comparable periods before treating it as the explanation for the whole program.
That last part matters when buying cycles are long. A recent group hasn’t had the same chance to progress as an older one. Comparing this month’s leads with this month’s closed deals can make a healthy program look broken, or hide a problem behind deals that started long ago.
Give the next report a shared meaning
Write down what counts as an inquiry, an accepted account, and an opportunity. Preserve when each step happened. Decide who investigates missing or contradictory records. Then make the recurring report follow those definitions.
You may need an integration or a reporting rebuild to maintain that view. A manual pass helps specify what it has to do: preserve the account identifier, carry the right timestamps, and flag exceptions someone will actually review.
At the next review, I’d want the team to be able to name the problem precisely: unsuitable inquiries from a particular campaign, suitable accounts waiting for follow-up, or accepted accounts missing from the report. Each comes with a different owner and a different next step. That’s enough clarity to put the budget conversation on firmer ground.