The Top-of-Funnel Math Reboot — 60-Min Training

The folklore inputs versus the four rebuilt inputs

Most teams run the top of the funnel on inherited rules of thumb, and this Reboot exists to replace each one with a number the team derives from its own data. There are two competing ways to set top-of-funnel targets, and the whole session is a head-to-head between them.

Option A — the folklore stack. This is the default in most B2B SaaS orgs: "3x pipeline coverage is the standard," "we need more leads," "MQLs are top-of-funnel," and "SDR dial volume is the input." Every one of these is a constant someone declared, not a value the business measured. The 3x coverage figure in particular is a heuristic that Jason Lemkin at SaaStr has repeatedly flagged as misleading — it happens to be right only for a narrow band of win rates and cycles and wrong for everyone else. Folklore is cheap to adopt and expensive to run, because it silently under- or over-builds pipeline every quarter, and nobody notices until the forecast misses.

Option B — the derived stack, the four reboots. Instead of importing constants, the team recomputes four things from the last 90 days of its own funnel:

• Real coverage = quota divided by (win rate × average deal size), expressed as a multiple of quota — a derived ratio, not a declared one.
• Conversion-by-source — lead-to-won percentage for every channel, because the blended average is a fiction that hides 10x spreads between the best and worst source.
• Demand versus capture — separating the spend that manufactures new buyer awareness from the spend that merely harvests demand someone else created.
• Kill-the-wrong-leads — the counterintuitive move where cutting the bottom 30% of low-fit volume raises total wins.

The trade-off between the two options is real, not rhetorical. Folklore is faster to run and requires no analysis, but it misallocates budget and quietly leaves quarters underfunded. The derived stack costs an hour of the team's time each quarter and demands clean source-level data, but it exposes exactly where pipeline is short and which sources to cut. This is a build session for Option B, not a lecture — by minute 60 the team has rerun the Math against its own numbers and named the one source it is turning off. Pair it with the companion pipeline-math Training (st186), which covers what happens *inside* pipeline after opportunities are created; this session covers what flows *in*. The two are sequenced deliberately: fix the inflow first, because optimizing stage conversion on a funnel fed by the wrong leads just moves garbage through faster.

How to decide which reboot to run first

Not every team should start at the same reboot, and the first fifteen minutes of the Training are a triage exercise to decide where the biggest leak is. The decision hinges on one question: is the team's problem a demand problem or a capture problem? If real coverage comes out *below* the current ratio, the funnel genuinely lacks enough qualified opportunities and the demand-side reboots (demand-versus-capture, source mix) matter most. If real coverage comes out *above* current — meaning the team is technically carrying enough raw pipeline but still missing quota — the leak is conversion and quality, so the kill-the-wrong-leads reboot comes first.

The second branch is data readiness. A team with clean source-level attribution can run conversion-by-source immediately; a team whose CRM lumps every lead into one "inbound" bucket has to fix tagging before the numbers mean anything, so it starts with the coverage Math (which only needs quota, win rate, and ACV) while attribution gets cleaned up in parallel. The point of the triage is to avoid the common failure where a team spends the hour polishing a reboot that isn't its bottleneck. A facilitator who skips triage almost always defaults to conversion-by-source because it produces the flashiest table, even when the room's actual problem is that it is carrying half the coverage it needs.

The frame the facilitator opens with, said almost verbatim, sets the tone: "We are going to redo top-of-funnel Math from scratch. Four reboots — real coverage, conversion-by-source, demand versus capture, and which leads to kill. By the end of the hour each of you has new numbers and one source to turn off." No debate on the myths; the numbers settle them.

The concrete numbers behind each reboot

This is where the session earns its hour. Each reboot has a worked calculation the reps run against their own inputs, and the ranges below are the ones the Training uses as anchors.

Reboot 1 — real coverage. Take a team carrying a $2M new-ARR quota with a 22% win rate on a $60K average ACV and a 90-day cycle. Closed-won deals needed: $2M ÷ $60K = 34 deals. Qualified opportunities required: 34 ÷ 0.22 = 155 opps. Pipeline dollars required: 155 × $60K = $9.3M. Coverage ratio: $9.3M ÷ $2M = 4.65x — not 3x. Run the inverse and the danger is obvious: at the same 22% win rate, a flat "3x coverage" funds only $6M of pipeline, which covers roughly $1.32M of quota. The team is about $680K short before the quarter even starts, and nobody noticed because the slide said 3x. The rules the room writes down: lower win rate means higher required coverage (a 12% win rate needs closer to 8x); a longer cycle means coverage must be *built earlier*, not larger, a point Tomasz Tunguz has written about at length for SaaS; segment coverage because enterprise and SMB win rates differ and one blended number hides both; and recompute every quarter because win rates drift.

Reboot 2 — conversion-by-source. Aggregate top-of-funnel metrics lie, a theme Mark Roberge hammers in *The Sales Acceleration Formula* and David Skok documents in his For Entrepreneurs funnel analyses — lead-to-customer conversion can vary 10x within the same company. A representative 90-day table: inbound organic search at 800 leads → 18 wins is 2.25%; webinar at 1,200 leads → 4 wins is 0.33%; paid LinkedIn at 600 leads → 6 wins is 1.00%; partner referral at 40 leads → 14 wins is 35.0%. The diagnostic moment lands when the room sees that partner referral converts roughly 15x better than webinar, yet webinar often gets 30x the budget. That is not a volume problem — it is a mix problem. The rules: rank sources by lead-to-won percentage, not lead volume; measure cost-per-won, not cost-per-lead (a $50 webinar lead at 0.33% costs about $15K per won deal, while a $400 partner lead at 35% costs about $1.1K); and kill or shrink the bottom quartile, reallocating to the top. The uncomfortable follow-on is that scaling the partner channel is hard — 40 leads does not become 400 by spending more — so the room also learns the difference between a source that is *efficient* and a source that is *scalable*, and how to fund both without confusing them.

Reboot 3 — demand versus capture. The split most teams miss, popularized by Chris Walker and echoed in Bridge Group SDR research, separates *demand creation* (the buyer learns the problem exists) from *demand capture* (the buyer fills out a form). Branded search ads, retargeting, and gated whitepapers are capture — they intercept and harvest existing intent. Podcast sponsorship, founder LinkedIn content, and cold outbound to an ICP list are demand — they manufacture awareness that didn't exist. When more than 70% of spend is capture, the team is harvesting a pool someone else built, and when that pool dries, pipeline collapses. Trish Bertuzzi makes the parallel point in *The Sales Development Playbook*: point an outbound sales team at lists with no demand work behind them and conversion craters. The target for B2B SaaS at $25K–$500K ACV is roughly a 40/60 demand-to-capture split. The trap the room usually finds is that its "demand" spend is mislabeled — retargeting and branded search get counted as marketing wins because they show a great cost-per-lead, but they are only capturing intent that the under-funded podcast and content programs actually created, so the efficient channel gets the credit the expensive channel earned.

Reboot 4 — kill the wrong leads, and the wrong-lead tax. Cutting low-fit volume raises wins. Worked example: at 1,000 MQLs per month, 8% to SQL and 18% SQL-to-won yields 14.4 wins. Kill the bottom 30% by fit score and reps focus — 700 MQLs at 15% to SQL and 22% SQL-to-won yields 23.1 wins. That is roughly 60% more wins on 30% less volume. The hidden cost this exposes is the wrong-lead tax: every unqualified demo burns 45–60 minutes of AE prep plus a 30-minute call, and at a blended AE cost of $100–$150/hour that is $100–$200 of direct salary per bad meeting. A team booking 100 demos a month with 40% unqualified wastes 40 × $150 = $6,000/month, about $72,000 a year — frequently more than the demand-gen source that produced those leads cost in the first place. The second-order cost is worse: bad demos train reps to expect low-quality meetings, so discovery gets lazy, and that laziness then bleeds into the good meetings too. Killing volume is partly a Math exercise and partly a way to protect the quality of every remaining conversation.

Running the 60 minutes and sequencing the kill list

The agenda is a build, not a deck, and the sequencing matters because each block feeds the next. Minutes 0–15: pull the last 90 days of source-level data — leads, demos, pipeline created, closed-won — and calculate the real coverage ratio, cutting any source below 1.0x coverage. Minutes 15–30: run the wrong-lead tax per source, flag the bottom 20% by conversion, and estimate the AE time bleeding into them. Minutes 30–45: reclassify remaining sources as demand or capture and compute pipeline yield per dollar, not leads per dollar. Minutes 45–60: name exactly one source to turn off next Monday, assign an owner to pause it, document the expected impact, and set a 30-day re-check.

The kill-list itself follows a ranking so the cut is defensible, not emotional. The criteria, in order: fit score (ICP match) below threshold; source with lead-to-won under 0.5%; stalled MQLs older than 60 days with no SQL conversion; and job-title patterns that historically never close — students, vendors, competitors, interns. The diagnostic question that unlocks the room is blunt: "If we deleted the bottom 30% of MQLs by fit score, what would actually break?" The answer is almost always nothing, except that reps would finally have time to work the top 70%. The one guardrail the facilitator enforces is a single-variable change — turn off exactly one source, hold everything else constant — so the 30-day re-check produces a clean read instead of a tangle of overlapping changes nobody can attribute.

Each rep closes by stating three things out loud: my new real coverage ratio is X (not 3x); the one source I am turning off Monday is Y; my demand-to-capture split is Z and I am moving it toward 40/60 by a named action. The 30-day recheck matters because top-of-funnel Math is not a one-time recompute — win rates and source mix drift every quarter, so the Reboot is a habit, not an event. Teams that run this agenda consistently report a 10–20% lift in demo-to-close within 60 days, purely from starving the Funnel of bad leads rather than adding good ones.

Related questions

Why is "3x pipeline coverage" wrong for most teams?

Because 3x is a declared constant, not a derived one. Real coverage = quota ÷ (win rate × ACV). A 22% win rate needs about 4.65x and a 12% win rate closer to 8x, so a flat 3x quietly underfunds most quarters before they begin.

Does killing leads reduce revenue?

No — you pause low-yield sources and low-fit segments, not existing live deals. Redirecting SDR time from the bottom 30% of MQLs to the top 70% typically raises SQL and win rates enough to grow total wins on lower volume, as the 14.4-to-23.1 example shows.

Who should attend this Training?

SDR leaders, managers, and RevOps — anyone who can actually adjust source budgets and coverage targets. Reps benefit from seeing the Math, but the recalculation exercises assume the attendee has access to pipeline and conversion data and authority to pause a source.

What is the demand-versus-capture split?

It separates spend that creates new buyer awareness (podcasts, founder content, cold outbound) from spend that harvests existing intent (branded search, retargeting, gated assets). Over-weighting capture past ~70% leaves pipeline dependent on a demand pool someone else built.

How is this different from st186?

This session fixes what flows *into* pipeline — sources, coverage ratios, capture efficiency. st186 optimizes what happens *inside* pipeline once deals exist — stage progression and velocity. Run this first to fix inflow, then st186 to improve conversion through the stages.

FAQ

What exactly does "coverage is a function of win rate and cycle" mean? It means the "3x pipeline coverage" rule of thumb ignores your actual conversion data. A team with a 20% win rate and a 90-day cycle needs a different multiple than a team at 30% and 60 days. The Training walks you through calculating your real coverage number from your own win rate, ACV, and quota rather than a generic benchmark.

How do you decide which leads to "kill" without losing revenue? You analyze conversion-by-source to find the bottom 10–20% of channels that consistently produce low close rates or long cycles. "Killing" means pausing or shrinking investment in those sources and redirecting SDR capacity toward higher-converting ones — not deleting existing leads. The diagnostic uses your own historical data, not arbitrary thresholds.

Is this session for SDRs only, or can managers and RevOps attend? It is designed for SDR leaders, managers, and RevOps — anyone who owns top-of-funnel metrics and can adjust source budgets and coverage targets. Reps benefit from understanding the Math, but the hands-on recalculation exercises assume access to pipeline and conversion data.

Does it cover the outbound-versus-inbound split? Yes — the demand-versus-capture reboot handles it directly. You separate leads from active demand generation from those captured off existing intent, because the Math and expected coverage ratios differ for each. The goal for most B2B SaaS teams is roughly a 40/60 demand-to-capture split.

Do I need to prepare data beforehand? You get the most value bringing 6–12 months of pipeline data by source — lead counts, opportunities created, and closed-won amounts. The session includes a template so you can estimate ranges if exact numbers aren't available. The Math works with honest ranges, not just precise figures.

How is this different from st186, the deal-stage pipeline Training? This session focuses on inflow — lead sources, coverage ratios, and capture efficiency. st186 covers what happens inside pipeline after deals are created, like stage progression and velocity. They are complementary: fix inflow here, then optimize stage conversion with st186.

Sources

• https://www.saastr.com/ — Jason Lemkin essays on pipeline coverage heuristics and SaaS sales metrics.
• https://tomtunguz.com/ — Tomasz Tunguz analyses of SaaS pipeline coverage and sales-cycle timing.
• https://www.forentrepreneurs.com/sales-marketing-machine/ — David Skok on SaaS funnel conversion by source.
• https://www.markroberge.com/ — Mark Roberge, *The Sales Acceleration Formula*, on source-level versus aggregate conversion.
• https://www.tenbound.com/ — SDR and sales development benchmarks and qualification research.
• https://www.linkedin.com/in/trishbertuzzi/ — Trish Bertuzzi, *The Sales Development Playbook*, on qualification and quality versus volume.
• https://refinelabs.com/ — Chris Walker / Refine Labs on demand creation versus demand capture.
• https://hbr.org/2017/03/a-refresher-on-marketing-roi — Harvard Business Review on marketing ROI and cost-per-outcome thinking.

Related on PULSE

• [The Pipeline Math Reboot — 60-Min Training](/knowledge/st186)
• [The Gatekeeper Navigation Reboot — 60-Min Training](/knowledge/st257)
• [The Sales Org Health Check Reboot — 60-Min Training](/knowledge/st239)
• [The Enterprise Land-and-Expand Reboot — 60-Min Training](/knowledge/st238)
• [The PLG Sales Motion Reboot — 60-Min Training](/knowledge/st237)
• [The Founder-Led Sales Transition Reboot — 60-Min Training](/knowledge/st236)