L5. Lead generation

Outputs doc: outputs.md, L5 section. Fill it in as you work through the steps below. Raw funnel-test data and response conversations go in captures.md. L6 reads the L5 section before starting.

Jump to: Diagnostic · Step 1 · Step 2 · Step 3 · Step 4 · Step 5 · Step 6 · Step 7 · Summary · Assumption sweep

What this layer is: The repeatable system that turns presence in the L4 channel into captured leads. It is the offer that gives the buyer a reason to respond, the path that carries them from seeing the message to becoming a known contact, and the numbers that say whether the path works. This is the first quantitative layer: everything above it was about being right, and this is where being right gets measured. The output is a working system with a defined offer, a mapped conversion path, real conversion rates at each stage, and a clear definition of what counts as a lead worth handing to L6.

Why it comes after L4 and before L6: L4 put you in the channel where the buyer is, with a message they will meet. L5 is the machine you build inside that channel to convert attention into contact. It cannot be built before L4, because a lead-generation system is channel-specific: the offer, the call to action, and the capture mechanism all depend on which channel and what the reader’s intent there is. And it must come before L6, because follow-up has nothing to follow up without captured leads; L5 produces the contact that L6 nurtures toward sales-ready. This is the layer where the error messages are loudest, the low conversion rates, the dead funnels, the cost-per-lead that does not work. And precisely because the symptoms scream here, it is the layer where most people start tinkering first, rewriting the ad, changing the button colour, while the real fault sits in a vague audience, a consensus value proposition, or a channel the buyer is not in. The numbers are only meaningful if L1 through L4 hold up. When the funnel underperforms, the first move is to check upstream, not to optimise the form.

What finishing this layer produces: A written lead-generation system in the L5 section of outputs.md: the offer and why it fits the channel’s awareness and intent, the step-by-step conversion path, a funnel-math table with real or tested conversion rates and targets traced back to L1 through L4, a definition of a lead and the qualification bar that separates a captured contact from one ready for L6, and the identified bottleneck stage with its fix. Backed by a live small-scale run, not by projected rates you wish were true.

Diagnostic: is L5 actually done?

Before building anything, answer these questions in writing. Vague answers mean the layer is not finished.

1. Is there a specific reason for the buyer to respond, matched to the channel’s intent, not just “contact us”? A call to action with no offer behind it converts the few who were already ready and loses everyone else. If the only reason to respond is that you would like them to, you do not have a lead-generation system, you have a hope. The offer is what turns attention into action.

2. Can you draw the full path from a stranger seeing the message to becoming a captured lead, every step? Not “they see our post and reach out.” Every step: what they see, what they click or send, what they land on, what they give you, what they get back. If there is a gap in the path you cannot name, that gap is where leads disappear.

3. Do you have numbers at each stage, with targets tied back to L1 through L4? This is the numbers-driven layer. If you cannot state the conversion rate from reach to response to lead, even as a tested estimate, and the volume you need to hit your goal, you are running on “we will see what happens.” A funnel without numbers cannot be diagnosed when it underperforms.

4. Have you defined what a lead actually is, and the bar that separates a captured contact from one ready for L6? A captured email is not the same as a qualified lead. If everyone who downloads a thing is called a lead and passed on, L6 and sales drown in people who will never buy. The qualification bar, built from L1 signals, is what keeps the handoff honest.

5. Have you run it at small scale and found the actual bottleneck, or are you assuming the funnel works? Projected conversion rates are fiction until a live run replaces them. If no real traffic has moved through the path and you cannot name the weakest stage, the system is a diagram, not a system.

If you answered all five clearly and in writing, L5 may already be done. Jump to the L5 section of outputs.md, fill in the fields, check the checklist, and move to L6. If not, work through the steps below.

Step 1. Pull the inputs from L4 (and above)

Duration: 30 minutes

L5 does not start from a blank page. It starts from the channel L4 chose, the message that channel carries, and the value proposition the whole thing has to deliver. The fastest route to a working funnel is to treat L4’s channel and baseline as the ground you build on, and L2’s value as the thing the offer must echo.

What to do:

1. Open the L4, L3, L2, and L1 sections of outputs.md and copy the following into the “Inputs from L4 and above” field in the L5 section:

   • From L4: the beachhead channel(s), the message form per channel, the reader and awareness level per channel, the baseline response from the Step 6 test, and the owned-audience path. The baseline is your starting conversion data; the funnel improves on it. Paste it where you can see it.
   • From L3: the lead message form and the call-to-action language the channel will carry, and the tone. The offer and CTA in L5 are written in this voice.
   • From L2: the final value proposition and the proof buyers said they would need. The offer has to be a small, true taste of this value, not an unrelated bribe.
   • From L1: the sharpest segment, the observable signals, the anti-audience, and the buying context. These become the qualification bar in Step 5 and the volume math in Step 4.

2. Write one sentence: “In [channel], a [reader] at [awareness level] meets [message form], and the reason for them to respond will be [rough offer idea].” It is a hypothesis; the rest of the layer builds and tests it.

3. If L4’s baseline response was thin, or the channel was chosen without a real test, that weakness flows straight into L5: you will be building a funnel on a channel that may not convert. Note it in the L5 scope notes in outputs.md and log it in assumptions.md (Layer: L5, Status: Untested). If the channel itself is unproven, the honest move is to finish L4 Step 6 before building the machine.

Step 2. Design the offer, the reason to respond

Duration: 45-60 minutes

A lead-generation system runs on an offer: the specific reason a stranger gives you their attention, their contact details, or their time. Without one, the call to action is “buy now” or “contact us,” which only catches the rare person already at the finish line. The offer is a small, low-risk taste of the L2 value, sized to the reader’s awareness and intent in the L4 channel. A cold, low-intent channel needs a low-commitment offer that asks for little and gives something useful immediately; a high-intent channel can carry a more direct offer closer to the sale. The offer must also be true to L2: a clever bribe unrelated to your value attracts people who want the bribe, not the product.

What to do:

1. In the L5 section of outputs.md, under “Offer options,” list candidate offers and score them in the table below (it is also in outputs.md). Use a 1-5 scale. Candidates usually fall into a few shapes: a useful resource (a teardown, a benchmark, a template, a tool), an experience (an audit, a trial, a demo, an assessment), or a direct conversation (a consultation, a teardown call). The right shape depends on the channel’s intent.

   Candidate offer	Fits the channel’s awareness and intent (1-5)	A true taste of the L2 value (1-5)	Low enough risk for the reader to say yes (1-5)	Cheap enough for you to deliver at volume (1-5)	Attracts buyers, not freebie-seekers (1-5)	Total

2. Weight two columns most heavily: whether the offer is a true taste of the L2 value, and whether it attracts buyers rather than freebie-seekers. An offer that scores high on “low risk to say yes” but low on “attracts buyers” fills the funnel with people the anti-audience from L1 already warned you about. The cheap, popular lead magnet that has nothing to do with your value is the classic version of this mistake.

3. Match the offer’s commitment level to the channel’s intent from L4. In a high-intent inbound channel where the reader is already comparing solutions, a direct offer (a demo, an audit, a teardown call) can work, because asking for a small step matches where they are. In a cold or low-intent channel, lead with a no-friction resource and earn the right to ask for more later. Mismatch here is a common quiet killer: a demo request in a cold feed converts almost no one, and a lightweight download in a high-intent channel under-asks and wastes the intent.

4. Build the anti-offer list. Some offers are tempting and quietly wrong. Capture them in the table below (also in outputs.md) so they do not return in Step 6 when conversion looks low and you reach for volume.

   Tempting offer	Why it looks strong	Why it fails for this segment

   Common entries: the unrelated freebie that pulls the wrong crowd, the offer that asks for too much commitment for the channel’s intent, the offer you cannot fulfil at volume, and the discount that trains people to wait for discounts.

5. The offer’s pull is an assumption until tested. Whether this segment will actually exchange contact for this offer, and whether the people who do are buyers, is exactly what Step 6 measures. Log the chosen offer in assumptions.md (Layer: L5, Status: Untested), and mark it a leap-of-faith if the whole funnel depends on it.

Example (continuing the L0 to L4 example):

Channel is LinkedIn, where the VP of Sales champion is problem-aware to solution-aware. Chosen offer: a deal-decay teardown, a short, specific review of where a team’s warm deals go dark, delivered as a one-page benchmark plus an optional 20-minute call. It is a true taste of the value (it shows the gap the product closes), low-risk to accept, and it self-selects for teams that actually have the problem. Anti-offers: a generic “ultimate guide to sales follow-up” PDF (pulls junior reps and content-scrollers, not VPs with budget), and a hard demo request (too big an ask for a feed where the reader is not yet shopping).

Step 3. Map the conversion path

Duration: 45-60 minutes

The offer is the reason to respond; the conversion path is the sequence of steps that carries a stranger from first seeing the message to becoming a captured lead you can contact again. Most leads are lost not because the offer is weak but because a step in the path is missing, confusing, or asks for too much too soon. Map every step explicitly, because you cannot fix or measure a path you have not drawn.

What to do:

1. In the L5 section of outputs.md, under “Conversion path,” lay out the path stage by stage in the table below (it is also in outputs.md). Each row is one step the person takes, from exposure to captured lead. Name what they see, what action they take, and what could stop them at that step.

   Stage	What the person sees or does	The call to action	What could stop them here (friction)
   Exposure (sees the message in the channel)
   Engagement (clicks, reacts, reads more)
   Offer presented
   Capture (gives contact info or books time)
   Confirmation (receives the offer, becomes a known lead)

2. Minimise the friction at each step, especially before capture. Every extra field, click, or decision loses people. Ask for the least you need to make the lead workable for L6, and no more. The most common path failure is asking for too much too early: a long form, a phone number, a meeting, before the reader has any reason to trust you.

3. Decide the capture mechanism deliberately: a form on a landing page, a direct message exchange, a calendar booking, a reply, a signup. Match it to the channel. In a social feed, a native interaction (comment or DM) often converts far better than sending people to an external form, because every hop off-platform loses people. Note where the captured contact lands (your list, your CRM), because that is the start of the owned audience L4 pointed toward.

4. Mark the single most important conversion point on the path, usually the move from engagement to capture. That is the step the Step 6 test will watch hardest, and the one where small improvements move the whole funnel.

5. The friction points are assumptions about where people will drop until the live run shows you. Log the steps you are least sure about in assumptions.md (Layer: L5, Status: Untested). The bottleneck is rarely where you expect.

Step 4. Put numbers on the funnel

Duration: 45-60 minutes

This is the layer the model calls numbers-driven, and this is the step that earns the name. A conversion path without numbers cannot be judged, improved, or trusted. Here you attach a metric to every stage, estimate the conversion rate between stages, and work out the volume you need at the top to hit your goal at the bottom. The numbers also tie the whole layer back to L1 through L4: the size of the segment, the reach of the channel, and the pull of the offer all show up as figures here, and an honest funnel reveals when those upstream layers cannot support the goal.

What to do:

1. In the L5 section of outputs.md, under “Funnel math,” fill in the table below (it is also in outputs.md). Work bottom-up from the goal: how many customers or sales-ready leads you need, then back up through each conversion rate to the reach required at the top. Use the L4 baseline test for any rate you have real data on, and a clearly-marked estimate for the rest.

   Stage	Metric	Conversion rate to next stage	Target volume	Source (tested / estimated / from L1-L4)
   Reach (channel audience exposed)	impressions or sends
   Engagement	clicks, replies, reactions
   Leads captured	contacts or bookings
   Qualified leads (ready for L6)	fit the bar in Step 5

2. Sanity-check the volume against L1 and L4. If hitting your goal requires reaching more of the segment than actually exists (L1’s segment size), or more than the channel can deliver (L4’s reach), the goal is not achievable through this channel alone, and no funnel tuning will fix it. This is one of the most valuable things L5 can tell you: that the constraint is upstream, not in the funnel. Write the finding plainly.

3. Add a rough unit economics line if money is involved: cost per lead and cost per qualified lead, against what a customer is worth. A funnel that converts but costs more per customer than the customer returns is a funnel that loses money faster the better it works. You do not need precision here, just the order of magnitude, enough to catch a path that cannot pay for itself.

4. Set the one metric that matters for this stage of the system: the single number that, right now, best tells you whether lead generation is working. Early on it is usually the rate at the most important conversion point from Step 3, not a vanity number like reach. Naming it stops you from optimising impressions while leads stay flat.

5. Every rate you did not pull from the L4 baseline is an estimate, and estimates at this stage are usually optimistic. Log the funnel assumptions in assumptions.md (Layer: L5, Status: Untested), especially the conversion rate at the most important point and the assumption that the segment is large enough to feed the goal. Step 6 replaces the estimates with measured rates.

Step 5. Define a lead and the qualification bar

Duration: 30-45 minutes

Not everyone who responds is a lead worth pursuing, and not every lead is ready for L6. This step draws two lines: what counts as a captured lead at all, and what counts as qualified, meaning fit enough and warm enough to hand to follow-up. Without these lines, the funnel reports big numbers that mean nothing, and L6 and sales spend their time on people the L1 anti-audience already flagged. The qualification bar is built from L1: the observable signals that predict fit, and the anti-audience signals that predict waste.

What to do:

1. In the L5 section of outputs.md, under “Lead definition and qualification,” write a one-line definition of a captured lead (the minimum: a contact you can reach again who took the offer) and a one-line definition of a qualified lead (fit plus intent enough for L6).

2. Build the qualification bar in the table below (it is also in outputs.md), pulling the fit signals from L1’s observable signals and the disqualifiers from L1’s anti-audience. Separate fit (are they the right kind of buyer) from intent (are they warm enough to act soon), because a fit lead with no intent goes to a nurture track, not to sales.

   Signal	Source (L1 signal / anti-audience)	Fit or intent?	Threshold to pass

3. Decide what happens to leads that are captured but not qualified. They are not waste by default: some are fit-but-not-ready and belong in a slow nurture (which L6 will design), and only those that fail the fit bar (the anti-audience) are truly out. Mark the routing: qualified to L6 now, fit-but-cold to nurture, anti-audience to discard. This keeps the handoff to L6 clean and keeps good-but-early leads from being thrown away.

4. Keep the bar honest. A bar set too low to make the lead numbers look good just moves the disappointment downstream to L6 and sales, where it is more expensive. A bar set too high starves L6 of anyone to work with. The right level is the one where a lead passing the bar has a real chance of becoming a customer, judged against the L1 segment.

5. The predictive power of each qualification signal is an assumption until you see which captured leads actually progress. Log the bar in assumptions.md (Layer: L5, Status: Untested); Step 6 and later L6 data show whether the signals you chose actually predict who buys.

Step 6. Run it small and find the bottleneck

Duration: 1-3 weeks calendar time, depending on channel speed | Target: enough live volume to get real rates at each stage

Everything above is a hypothesis until traffic moves through it. Step 6 runs the system at small scale, measures the real conversion rate at every stage, compares it to the Step 4 targets, and finds the one stage where the funnel leaks worst. You are not scaling yet; you are replacing estimates with measured numbers and fixing the weakest link before you spend to fill the top. This is also where the discipline of the whole model pays off: when a stage underperforms, the question is whether the fault is in L5 (the offer, the path, the ask) or upstream (the wrong audience, a weak value proposition, a channel the buyer is not in). Resist the reflex to tinker locally before you have checked which layer the error actually lives in.

There are two halves to L5 validation: the quantitative run (the numbers) and a few conversations with people who did and did not respond (the why behind the numbers). The numbers tell you which stage leaks; the conversations tell you why, which is what you actually need to fix it.

6a. Run the funnel at small scale

Duration: the bulk of the time

Put real but limited volume through the path: a small budget, a batch of outbound, a few weeks of organic posting with the offer, whatever the L4 channel allows cheaply. Instrument every stage so you can see the drop from one to the next. Keep running until each stage has enough volume that the rates are not noise.

What to measure:

• The conversion rate at every stage of the Step 3 path, not just the final lead count. The whole point is to see where people drop, not only how many arrive.
• The rates against the Step 4 targets. Mark each stage as at, above, or below target.
• The quality of the leads against the Step 5 bar: what share of captured leads actually qualify. A high capture rate that produces few qualified leads points at the offer attracting the wrong people.
• The one metric that matters from Step 4, tracked over the run.

Find the bottleneck: the single stage with the largest gap between its rate and its target, or the largest unexpected drop. That stage is where you work first. Fixing the worst stage moves the whole funnel more than polishing a stage that already works.

6b. Ask the people who did and did not respond

Duration: light, a handful of short conversations

Numbers locate the leak; people explain it. Talk to a few who converted and, more valuably, a few who engaged but did not. Recruit from the L1 sharpest segment, the same way every layer has.

What to ask those who responded:

• What made you take the offer? What nearly stopped you?
• Was the offer what you expected when it arrived?

What to ask those who engaged but did not convert:

• You looked but did not take it up. What held you back?
• Was anything unclear, or did it ask for more than you wanted to give at that point?
• What would have made it an easy yes?

You are listening for whether the leak is a fixable L5 problem (too much friction, a confusing step, an offer mismatched to the channel’s intent) or an upstream signal (they are not really the segment, the value did not land, they are not in a buying frame in this channel). The difference decides whether you fix the form or go back a layer.

6c. Capture immediately after the run and conversations

Do this while the data and quotes are fresh. Add entries to the L5 section of captures.md using this structure:

Funnel run [number] | Channel + offer:
Dates / volume:

Stage rates measured (each stage, rate to next, vs target):

The bottleneck stage (largest gap or drop):

Lead quality (share of captured that qualified):

One metric that matters (value over the run):

---

Response conversation [number] | Converted / did-not-convert:
Reader (role, segment fit):
Why they did or did not respond (verbatim):
Friction or mismatch named (verbatim):
Is the leak an L5 fix or an upstream signal?:

Surprises (anything you did not expect):
Any L5 assumptions this confirmed or challenged:

The measured stage rates and the verbatim reasons for not converting are the most valuable output. The rates say where the funnel leaks; the quotes say whether to fix the funnel or go back a layer. Capture both exactly.

6d. Synthesise and fix the bottleneck

Do this once each stage has real volume. Go through your captures in captures.md and update the Funnel synthesis fields in the L5 section of outputs.md, covering:

Measured rates vs targets. Replace the Step 4 estimates with the rates the run produced. Where did reality beat or miss the estimate, and by how much? The corrected funnel is the real one; the estimated one was scaffolding.

The bottleneck and its likely cause. Name the worst stage and what the conversations suggest is behind it. Be specific about whether it is an L5 problem (friction, offer, ask) or an upstream one (audience, value proposition, channel). This single judgement is the most important output of the layer.

Lead quality. What share of captured leads qualified, and did the offer attract buyers or freebie-seekers? If capture is high but quality is low, the fix is usually the offer (Step 2) or the channel targeting (L4), not the form.

Is the constraint upstream. If the funnel leaks worst at a stage that no local fix improves, and the conversations point to people who are not really the segment or do not feel the value, the constraint is above L5. Say so plainly and go back. This is the model working as intended: the loud error in L5 revealing a quiet fault in L1, L2, or L4.

The fix you are making, and the re-test. State the one change you are making to the bottleneck stage and what you expect it to do. Then re-run that stage. One change at a time, so you know what moved the number.

After synthesis, mark the relevant rows in assumptions.md as Validated or Invalidated, and note what the evidence showed. If the funnel was invalidated by an upstream cause, the fix is in that layer, not in L5 tactics. Do not scale a leaking funnel.

6e. If you genuinely cannot run a live test

If you cannot put real volume through the path in the time available, reduce the unknowns as much as possible and mark the gap clearly.

• Run the smallest possible manual version. Even ten hand-done outbound touches or one week of posting the offer produces directional rates that beat pure estimates.
• Pressure-test the path with people in the segment. Walk a few segment-fit people through the path and watch where they hesitate; usability problems show up even without volume.
• Benchmark the rates. Compare your estimated conversion rates against published benchmarks for the channel and offer type, marking clearly that benchmarks are not your numbers. If your plan only works at rates well above benchmark, the plan is fragile.

Document each in the L5 section of captures.md, noting it is secondary. Log the absence of a real run as an assumption in assumptions.md. A funnel never run live is a diagram; L6 needs to know it is building on projected rather than measured rates.

Step 7. Write the final lead-generation system

Duration: 30-45 minutes

You now have an offer, a mapped path, measured rates, a qualification bar, and a fixed (or diagnosed) bottleneck. Lock the system.

A complete L5 output has six parts. Write a line or two for each:

1. The offer: what it is, why it fits the channel’s awareness and intent, and how it is a true taste of the L2 value.
2. The conversion path: the stages from exposure to captured lead, with the capture mechanism and where the lead lands.
3. The funnel math: the measured conversion rates, the target volumes, the one metric that matters, and the unit-economics sanity check.
4. The lead definition and qualification bar: what a captured lead is, what a qualified lead is, and the routing for fit-but-cold and anti-audience leads.
5. The bottleneck and the fix: the weakest stage, its likely cause (L5 or upstream), and the change you made or are making.
6. The upstream finding, if any: whether the run revealed a constraint in L1 through L4 that L5 cannot solve, so it is visible and not silently absorbed.

Keep it as structured fields. Format does not matter. Measured numbers do: this is the layer where “it should convert” is not good enough.

What to do:

1. Write the final system directly into the matching fields in the L5 section of outputs.md.
2. Read it once as a skeptic who suspects you are about to scale a funnel that only works on paper. Can you point to measured rates and a known bottleneck, not projections? If not, run Step 6 before you commit budget.
3. Run the diagnostic from the top of this page one more time. If all five questions now have clear written answers, L5 is done.

What you’ve built

After completing the steps above, the L5 section of outputs.md should contain:

This is not a deliverable for anyone else. It is a constraint on L6.

Assumption sweep

Before moving on, scan the L5 section of outputs.md for any field you filled in from reasoning rather than evidence. Common ones at L5:

• The conversion rates (did the live run produce them, or are they still estimates you hope hold at scale?)
• The offer’s pull (did the segment actually exchange contact for it, or did you assume they would?)
• The lead quality (did captured leads actually qualify, or did you assume the offer attracts buyers?)
• The qualification signals (did the signals predict who progresses, or are they your best guess?)
• That the segment is large enough and the channel deep enough to feed the goal (did the math confirm it, or did you skip the sanity-check?)

Each unconfirmed field is an assumption. Log it in assumptions.md now if you have not already. The conversion rate at the most important point and the offer’s pull are usually the highest-impact assumptions in the layer; if untested, mark them leap-of-faith and run Step 6 before scaling.

What this layer hands off to L6

Before moving on, confirm the L5 section of outputs.md is complete. L6 opens by reading it. Specifically, L6 needs:

• The qualified leads and the qualification bar, because follow-up acts on exactly these people. L6 needs to know who it is nurturing and why they passed the bar.
• The fit-but-cold leads and their routing, because the slow-nurture track L6 designs is built for them. The leads that are not ready now are L6’s main long-game.
• The offer and what the lead has already received, because L6 continues a conversation that L5 started; it must not restart from zero or re-pitch what they already took.
• The measured funnel rates and the bottleneck, because L6’s job is partly to lift the stages after capture, and it needs to know where L5 left off and what is already weak.
• The lead’s source channel and awareness level, because a lead from a cold channel needs warmer, slower follow-up than one from a high-intent channel who is nearly ready.

The quality of your L5 leads sets the ceiling on what L6 can do. Follow-up cannot turn the wrong people into customers; it can only carry the right ones further. If L6 keeps stalling, check whether the leads it receives actually clear a meaningful bar, and whether the loud numbers in L5 are masking a quiet fault upstream. The recurring lesson of this layer holds at the handoff too: a problem that shows up in follow-up often originates before it.

Common failure modes

There is no offer, just a call to action. “Contact us” or “book a demo” with nothing given in return catches only the already-ready and loses everyone else. Design an offer that gives the reader a real, low-risk reason to respond, sized to the channel’s intent.

You tinkered with L5 while the fault was upstream. The numbers screamed here, so you rewrote the ad and changed the form ten times while the real problem was a vague audience, a consensus value proposition, or a channel the buyer is not in. When a stage will not improve with local fixes, check the layer above before optimising the button. This is the single most common and most expensive L5 mistake.

The offer attracts freebie-seekers, not buyers. A popular lead magnet unrelated to your value fills the funnel with people who want the magnet and never the product. The capture numbers look great and the qualified numbers do not. Tie the offer to the L2 value and weight “attracts buyers” over “easy yes.”

The funnel has no numbers. You launched on “we will see what happens,” so when it underperforms you cannot tell which stage is broken. Put a metric and a rate on every stage; a funnel you cannot measure is a funnel you cannot fix.

Too much friction before capture. A long form, a phone number, a meeting request before the reader trusts you at all. Each extra ask sheds people. Ask for the minimum that makes the lead workable, and earn the bigger asks later.

A captured contact is treated as a qualified lead. Everyone who downloads anything is called a lead and shipped to sales, who drown in people who will never buy. Define the qualification bar from L1 signals and route fit-but-cold leads to nurture, not to a closer.

You scaled a leaking funnel. The small run showed a bad rate at a key stage, and you poured budget into the top anyway, multiplying the leak. Fix the bottleneck at small scale first; scale amplifies whatever the funnel already does, including losing.

The goal needs more of the segment than exists. The math only works if you reach more buyers than L1’s segment contains or the L4 channel can deliver. No funnel tuning fixes a goal the upstream layers cannot support. Surface it and either widen the channel plan or reset the goal.

“Leads are flat, so marketing is failing.” Said with the funnel treated as the whole of marketing. Lead generation is one layer, and its numbers are only meaningful if the four above it hold. Flat leads are a symptom; the cause is often higher up.

Sources

• $100M Leads by Alex Hormozi. The core source for L5. On lead magnets as a true taste of the paid value, the ways to generate leads, and why the offer (not the ad) is the lever. Behind Steps 2 and 3.
• $100M Offers by Alex Hormozi. The value equation behind offer design: dream outcome, perceived likelihood, time, and effort. Background for scoring the offer in Step 2 and for matching commitment to the channel’s intent.
• DotCom Secrets by Russell Brunson. On the conversion path as a deliberate sequence of steps and the value ladder from free offer to sale, with friction removed at each step. Behind the path map in Step 3.
• Lean Analytics by Alistair Croll and Benjamin Yoskovitz. The numbers-driven discipline: a metric for every stage, the one metric that matters right now, benchmarks, and reading a funnel honestly. The foundation for Step 4 and the measurement in Step 6.
• Predictable Revenue by Aaron Ross. On building a repeatable lead-generation system rather than ad hoc effort, separating a captured contact from a qualified lead, and the clean handoff to the next stage. Behind the system framing and the qualification bar in Step 5.