Turning reach into a pipeline you can report on
The five funnel stages worth tracking, and the ones that only look impressive.
The gap where trust goes
A marketing report that ends at reach asks the client to finish the argument themselves. Two point four million impressions, a four per cent engagement rate, eleven thousand new followers — and then a silence in which the client is supposed to conclude that this was worth what they paid. Some will. The ones who do not will ask at the next call, and if the funnel was never instrumented, the answer is an anecdote.
The fix is not more metrics. It is a small number of stages that each name a set of real rows — real sessions, real people, real deals — connected end to end, so the path from a post to money is a sequence you can walk rather than a leap of faith.
The enumerability test
Before adopting any stage, apply one test: could you produce it as a list? Not the number — the rows. If “qualified leads: 84” cannot be exported as eighty-four rows with names against them, it is not a stage. It is a figure derived from something else, and it will not survive a client asking which ones.
The test quietly disqualifies most of what appears in social reporting, and that is the point. It also explains why funnels built purely from platform dashboards fall apart in the middle: a platform can enumerate impressions and clicks, and nothing after that.
The five stages worth tracking
| Stage | What it counts | Where it comes from |
|---|---|---|
| Impression | Times the post was served. Context for everything below it, and a denominator. Never a headline. | Platform |
| Qualified click | A click that arrived carrying a campaign tag and did not bounce within a few seconds. | Your analytics, via UTM |
| Identified lead | Someone whose name you now hold: a form, a reply, a subscription, a booked call. | CRM or forms |
| Qualified opportunity | A lead a human has spoken to and judged real, against a written definition. | CRM, set by a person |
| Closed revenue | Money invoiced, attributed back to the campaign that started it. | Finance, matched to CRM |
Five is not a magic number, but it is close to the minimum that still tells a story, and every extra stage costs discipline at the point where discipline is already scarce. Steps between click and lead — sessions, page views, scroll depth — are diagnostic tools for a page, not stages of a pipeline. Keep them in the analysis and out of the report.
The one that is always wrong
Qualified opportunity is where funnels rot, because “qualified” is usually a feeling. Write the definition down in one sentence before the quarter starts — budget, authority, timeline, whatever your version is — and print it in the report next to the number. A stage defined by whoever last touched the record cannot be compared across two months, and a comparison across two months is the only reason to measure it.
What to stop reporting
- Impressions as a headline. They are a denominator. On their own they measure distribution spend, not interest.
- Engagement rate with no stage behind it. It tracks post format far more closely than intent — a poll beats a case study every time and means less.
- Follower growth. Worth watching for anomalies, worthless as a period-on-period line, and trivially bought.
- Video views. Most platforms count a view at two or three seconds. That is a scroll, not an audience.
- Reach with no denominator. Reach without the audience it reached out of is a large number with no shape.
None of these are lies, and none of them need deleting from the dashboard. They need demoting out of the report — because a report is a set of things somebody is expected to act on, and nobody has ever acted on follower growth.
Instrumenting it once
Everything above depends on a link carrying its campaign from the post all the way to the CRM. That is one piece of infrastructure, and it has to be got right once.
One taxonomy, written down
Decide the shape of source, medium, campaign and content, and write the rule for each in a sentence. Lower case everywhere, no spaces, no dates inside a campaign name.
One library, never hand-typed
Every link is built from a stored definition. Hand-typed tags produce Facebook, facebook and FB as three separate campaigns, and no amount of later cleaning recovers the split.
Capture the campaign at the form, not at the session
Store the campaign on the lead record when it is created. Session data expires, gets cleared, and does not survive the gap between a first visit and a form filled three weeks later.
Agree the attribution rule and print it
Last non-direct click is a defensible default for most teams. What matters is that the rule sits at the bottom of the report, so nobody relitigates it in the meeting.
Match spend to the same campaign key
Imported ad spend that cannot be joined to a campaign leaves cost on one side of the report and results on the other, which is where cost per lead goes to die.
Setting targets backwards
A target built by adding twenty per cent to last month is a wish. Build it the other way, starting from money:
- Start with the revenue the engagement is meant to produce this quarter.
- Divide by the average closed deal to get the number of opportunities that must be won.
- Divide by the win rate you have actually observed — not the one in the pitch deck — to get opportunities needed.
- Repeat up the funnel with the conversion rates you have: to leads, then clicks, then impressions.
The result is usually uncomfortable, and that is its value: it is the first honest statement of whether the plan and the budget agree. A required impression count the budget cannot buy is worth discovering in week one rather than in week ten.
Then divide by weeks, not months. A monthly target only becomes actionable in its final week, which is exactly when nothing can be done about it.
Reading it weekly
Watch the conversions between stages rather than the totals. A total falling can mean anything; a conversion rate falling names the problem.
- Impressions up, clicks flat. The creative is being served and ignored. A targeting or message problem, not a budget one.
- Clicks up, leads flat. The landing page is the constraint. The campaign is working and the page is not.
- Leads up, opportunities flat. Either the targeting brought the wrong people, or “qualified” is being applied inconsistently. Check the definition before blaming the channel.
- Opportunities up, revenue flat. A sales-cycle or pricing question — and usually the point at which marketing reporting has done its job and handed over.
Each of those has a different owner and a different fix, which is the entire argument for tracking stages instead of a headline number.
When a stage has no data
Say so, in the report, in the row where the number would be. A blank with a note — “no closed revenue attributed this month; two opportunities still open from July” — costs nothing and buys credibility for every other figure on the page. A stage quietly omitted in the month it looks bad is the fastest way to make a client re-read the previous six reports with suspicion.
Questions this raises
What are the marketing funnel stages worth tracking?
Five: impression, qualified click, identified lead, qualified opportunity and closed revenue. The test for including a stage is whether you could produce it as a list of rows rather than as a number. Steps you cannot enumerate — sessions, scroll depth, engagement rate — are diagnostics for a page, not stages of a pipeline.
Is engagement rate a vanity metric?
It is a useful diagnostic and a poor report line. Engagement rate tracks post format more closely than it tracks intent, so it rises when you publish polls and falls when you publish case studies, independently of whether anything downstream improved. Keep it on the dashboard and out of the report.
Which attribution model should a marketing report use?
Last non-direct click is a defensible default for most teams. The model matters far less than keeping it constant and printing it at the bottom of the report: a consistent imperfect rule shows a trend, while a rule that changes between months shows nothing at all.