Why a period funnel misreads lead generation
The textbook tree multiplies traffic by a conversion rate and stops at a count produced inside the reporting window. A lead programme breaks both assumptions. The count at the end of the chain is not the thing the business wants: a form submission is a request for attention, not a purchase, and its value depends entirely on what sales does next. And the window is wrong, because a lead captured this month turns into pipeline next month and into a contract after that. A tree that adds this month's leads to this month's pipeline is comparing different people, so a strong quarter for capture can sit beside a weak quarter for revenue and nobody can say which number to believe.
There is also a quality problem the generic tree does not face. Two sources can deliver the same volume while one produces buyers and the other produces students, competitors and people who wanted a free report. If the tree treats them as interchangeable, every cheap source looks like an improvement. The guide to lead generation versus demand generation sets out why capture and creation need different targets; the tree has to reflect that split rather than blend it. The lead generation objective hub lists the channels that feed the tree and the role each one plays.
Decisions that change for lead generation
| Tree step | Generic tree | Lead generation tree | Decision it forces |
|---|---|---|---|
| Root | Revenue or orders produced in the period | Sales-accepted leads or qualified pipeline value | How far downstream the root sits and who signs it off |
| Time | Every leaf in the same window | Leads by creation period, pipeline by maturity period | How long the lag is and which cohort each figure belongs to |
| Rates | One blended conversion chain | Stage-to-stage rates per source | Which sources are separated and at which stages |
| Ownership | Marketing channels own the leaves | Some leaves sit with sales development and sales | Which leaves marketing reports but does not own |
| Guardrails | Cost ceiling | Disqualification share and reasons from the CRM | Which quality signal can veto a volume gain |
Put the root past the form
Form fills are the easiest number to move and the least informative. A shorter form, a gated download or a pre-filled social form will raise them within a week without changing what the business earns. The root belongs at the first stage that a buyer-side judgement has touched: leads that sales has accepted, or the value of opportunities qualified from them. Accepted leads suit a team with a steady deal size and a clear acceptance rule; pipeline value suits a business where one enterprise deal outweighs many small ones. Whichever is chosen, agree the stage definitions with sales first, using the shared vocabulary in the guide to MQL, SQL and lead stages, and treat form fills as the first driver beneath the root rather than as the outcome.
Carry the lag as its own branch
Draw two clocks. One counts leads by the period in which they were created; the other counts pipeline by the period in which it matured. The branch between them is the lag: the typical time from creation to acceptance and from acceptance to qualified opportunity. Read each cohort of leads at the same age, so that last quarter's leads are judged on what they became after the usual delay, not on what had happened by the end of the quarter in which they arrived. The business-to-business measurement guide covers cohort reporting in detail; the tree only needs to show which clock each leaf runs on. Without this branch, a fall in pipeline is often blamed on the current campaign when it was caused by the campaign before it.
Split stage rates by source
A blended acceptance rate hides more than it reveals, because the population beneath it changes whenever the media mix changes. Split each stage-to-stage rate (CVRShare of visits or clicks that convert.) by source: paid search, native social forms, events, partners, organic. The LinkedIn targeting and lead gen forms guide explains why pre-filled forms tend to produce more submissions with weaker intent, which is exactly the kind of shift a blended rate would conceal. The decision is how fine to cut: separate any source large enough to move the blended rate on its own, and group the rest. The split also tells finance which cost per accepted lead to set against which source, rather than one average that flatters the cheap channel.
Name the owners on the sales side
Some leaves in this tree are not marketing's to move. Speed to first contact, the number of attempts before a lead is closed out and the acceptance decision itself sit with sales development. Put them in the tree with that team's name against them. Marketing still watches them, because a slow follow-up wastes leads it has already paid for, but a target on a leaf someone else controls produces argument rather than action. The choice to make is where the hand-off line falls; the lead scoring workflow guide shows how a score can route weaker leads to nurture so that the sales queue holds only what it can work quickly.
Feed lead quality back from the CRM
The tree needs one guardrail that the form cannot game: the share of leads that sales disqualified, with the reason, reported by source. If a new form layout lifts submissions while that share rises, the change is a volume gain at lower quality and is reported as such. The form design guide covers the qualifying fields that trade a little volume for a lot of clarity. Credit for the root is also an AttributionAssigning conversion credit to touchpoints. stance, not a fact; record which model the tree uses so that a change of model is not mistaken for a change in performance.
Decision points in a worked lead generation tree
An invented industrial-software vendor, Ferrowell Systems, whose form fills rise while qualified pipeline falls
Ferrowell sells maintenance-planning software to factories. Its root is qualified pipeline value. In the third quarter form fills rose from 1,200 to 1,650, yet qualified pipeline fell from £2.16m to £1.67m, and the board asked how more leads could produce less pipeline.
| Source | Fills, Q2 | Accepted, Q2 | Acceptance, Q2 | Fills, Q3 | Accepted, Q3 | Acceptance, Q3 |
|---|---|---|---|---|---|---|
| Paid search | 500 | 200 | 40% | 480 | 180 | 37.5% |
| Social lead forms | 400 | 80 | 20% | 900 | 72 | 8% |
| Webinars | 200 | 50 | 25% | 170 | 34 | 20% |
| Partner referrals | 100 | 30 | 30% | 100 | 30 | 30% |
| Total | 1,200 | 360 | 30% | 1,650 | 316 | 19% |
| Sales-side leaf | Q2 | Q3 | Owner |
|---|---|---|---|
| Median hours to first call | 4 | 19 | Sales development lead |
| Accepted to qualified opportunity | 50% | 44% | Sales development lead |
| Qualified opportunities | 180 | 139 | Sales |
| Average opportunity value | £12,000 | £12,000 | Sales |
| Disqualified as student, job seeker or competitor | 9% | 41% of social form leads | Guardrail from CRM |
The source split puts the whole volume gain in one row: a gated benchmark report promoted to a broad audience more than doubled social form fills while their acceptance fell from 20% to 8%. The acceptance leaf then explains the second loss. Sales development spent the quarter calling those leads, median time to first call on every source rose from 4 to 19 hours, and the conversion of accepted leads into opportunities slipped from 50% to 44%. The lag branch adds context: closed revenue in the same quarter rose by 6%, because it came from first-quarter pipeline, and that figure had been read as proof the new campaign worked.
Decisions: the report's audience is narrowed to plant and maintenance job functions; the social form gains one qualifying question; leads scoring below the threshold go to nurture rather than the call queue; and the sales development lead owns returning median first-call time below 6 hours within the next quarter.
The example shows why each decision earns its place. Rooted in form fills, the tree would have reported a record quarter. Without the source split, the fall in acceptance would have looked like a general decline in lead quality rather than one campaign. Without a named owner for first contact, the slower follow-up would have been treated as a marketing problem. And without the lag branch, the rising revenue figure would have kept the campaign running for another quarter.
Measuring a lead generation tree: KPIs and cadence
Read the tree on three rhythms. Weekly: form fills, cost per lead and speed to first contact by source, because these move quickly and a broken routing rule shows up within days. Monthly: acceptance and accepted-to-opportunity rates by source, read for the cohort of leads created a lag-length earlier rather than for the calendar month. Quarterly: qualified pipeline value against target, cost per accepted lead and cost per opportunity by source, and the disqualification guardrail. Closed revenue belongs in the quarterly view as a lagging check matched to the cohorts that produced it. Platform conversion tracking counts the form submission reliably; everything after it has to come from the CRM, so the tree is only as good as the stage dates sales records. The cost-per-acquisition and conversion-rate calculators in the tools section help sanity-check the arithmetic between levels.
Where a lead generation tree misleads
Small numbers make the lower branches noisy: a vendor that closes a handful of deals a month will see opportunity rates swing for reasons no leaf can explain. Acceptance criteria drift as sales teams get busy or quiet, so the same lead can be accepted in one month and rejected in the next, and the tree will show a marketing change that never happened. Long cycles mean a cohort is not finished when it is first read, and the temptation is to judge it early. Buying committees put several contacts from one account into the tree, which inflates lead counts while the number of accounts in play stays flat; count accounts as well as people. And the tree describes capture only. Demand that is created now and harvested later never shows up as a leaf, which is why a lead tree should sit beside, not replace, the measures of demand creation.
Common mistakes
- Rooting the tree in form fills. The cheapest number to inflate becomes the target.
- Adding this period's leads to this period's pipeline. Different cohorts, one misleading ratio.
- A single blended acceptance rate. One noisy source drags it down and hides which.
- Targets on leaves sales owns. Marketing cannot shorten a call queue it does not run.
- No quality guardrail. A form change that doubles volume is celebrated before the disqualifications arrive.
- Counting contacts, not accounts. A committee of five looks like five leads and one deal.
