The lead and pipeline economics guide

Cost per lead is the wrong unit.

Most lead-gen reporting stops at cost per lead. It is the cheapest number to produce and the most misleading one to act on. CPL tells you what you paid at the top of the funnel and nothing about what arrived at the bottom. Two campaigns can post an identical A$80 CPL while one prints money and the other quietly destroys it, because the only figures that matter live downstream: what fraction of those leads close, and what each closed customer is worth.

As a reference point, the blended cross-industry CPL in 2026 sits near US$214, up from roughly US$198 a year earlier, with a range that runs from under US$10 for mass B2C to over US$800 for specialised B2B (First Page Sage, US data, 2026; about A$325 at US$1 = A$1.52). That spread is not noise. It is driven by three structural variables: customer lifetime value, sales-cycle length, and how crowded the paid auction is in your category. A single average CPL collapses all three into a number that describes no real business.

The correct unit is lead value: the expected margin a lead generates, discounted by the probability it ever converts. Once you hold lead value instead of lead cost, the whole picture inverts. A A$10 lead that never closes is infinitely expensive. A A$200 lead that becomes a A$5,000 customer is one of the best purchases the business will make this quarter. This guide builds that picture from the unit economics up, and ends at a single operating metric that ties cost, conversion and time together.

What a lead actually costs, by vertical.

Here is the published baseline. First Page Sage compiles blended cost per lead across paid and organic channels over a multi-year window (US data, spanning January 2022 to mid-2025). No Australian publisher releases a comparable per-vertical CPL series, so we use the US benchmark for its shape, not as an Australian price list: Australian CPLs sit in a broadly similar band (roughly A$20 to A$500-plus depending on vertical and competition, per WebApex, Australia, 2026), and the relative ordering across verticals holds. The verticals most relevant to a A$5M to A$1B service business diverge by almost an order of magnitude.

The headline pattern: professional services (legal, accounting) sit near US$650 a lead, home services (HVAC) near US$90, with construction, solar and healthcare in between. The reason is not that legal marketing is wasteful. It is that a legal matter is worth multiples of an air-conditioning repair, so the auction clears at a higher price and a higher CPL is still rational. CPL scales with the value of what the lead can become. That is the entire point: a number is only expensive relative to what it buys. The same ordering holds in Australia: a personal-injury or commercial-law enquiry clears the Google auction far above a split-system service call.

Organic compounds; paid rents

Across every vertical First Page Sage measures, organic CPL runs 25 to 40 percent cheaper than paid (US data). Legal is US$516 organic versus US$784 paid; HVAC is US$69 versus US$115; construction US$174 versus US$280. Paid leads arrive the day you pay; organic leads keep arriving after the work is done, which is why the blended figure understates the true long-run economics of owned demand. The mechanism is the same in any market, Australia included.

Read the right-hand column carefully. Legal has the highest CPL of any vertical here, yet one of the lowest effective costs to acquire a customer, because legal converts at 7.4 percent against engineering's 1.2 percent. CPL ranks the verticals in almost the reverse order of CPA. Anyone benchmarking on CPL alone would draw exactly the wrong conclusion about which channel is efficient.

The lead-value formula.

Lead value is the expected contribution a single inbound lead carries, before you have any idea whether it will close. It is the product of two terms you already have: what a won customer is worth, and the probability this lead becomes one.

From there, the cost a lead truly carries is its acquisition cost spread over the close rate, which is just customer acquisition cost expressed from the lead up:

A worked example

Take an Australian home-services operator spending A$1,600 a month on paid search. The figures here are demonstrative, chosen to show the relationship.

Lead value here is concrete: A$8,500 average job × 20 percent close = A$1,700 of expected revenue per lead, against an A$80 cost. The per-lead margin dwarfs the per-lead cost, which is why this operator should be buying every lead the auction will sell at that price, not optimising the A$80 down. The principle, stated plainly: an A$10 lead that never closes costs more than an A$200 lead that becomes an A$5,000 customer. The denominator that matters is conversion, not cost.

Your maximum profitable cost per lead.

Lead value tells you what a lead is worth. The maximum profitable CPL tells you the most you can pay for one and still hit your margin target. Rearrange the economics around a target return and the ceiling falls out:

Because deal value and close rate differ by job type, the ceiling differs too. The same operator can rationally pay six times more for one lead type than another. This is why a single account-wide target CPL is a blunt instrument: it under-buys the high-value work and over-buys the low-value work.

The demonstrative figures above vary the inputs to show the shape of the relationship, not to assert measured rates for any business. The lesson holds regardless of the exact numbers: set the CPL ceiling per job type, from its own deal value and close rate. A blended ceiling is a rounding error that happens to be wrong in both directions at once.

The funnel, stage by stage.

Close rate is not a single event. It is the product of several stage-to-stage conversions, and the compounding is brutal. Published B2B funnel benchmarks (MarketJoy, 2024 to 2025) put the canonical path at roughly: lead to marketing-qualified 22 percent, MQL to sales-qualified 15 percent, SQL to opportunity 11 percent, and opportunity to closed-won 7 percent. These are international B2B averages, not Australia-specific, but the multiplicative structure is universal. Multiply them and only a fraction of one percent of raw leads ever close through that particular chain.

The widths above show survivors as a share of the starting cohort. The single largest drop in most B2B funnels is MQL to SQL, the handoff between marketing and sales, where intent is re-tested and most leads fail. That is also the cheapest stage to fix, because it is a definition and routing problem more than a demand problem.

The compounding maths is the argument for measuring every stage, not just the ends. A 22 percent lift at a single mid-funnel stage flows through every downstream multiplication. Moving MQL-to-SQL from 15 to 18 percent raises closed-won survivors by a fifth without touching spend, lead quality or headcount. Stage conversion is the highest-leverage, lowest-cost place to add profit in the entire pipeline.

The five-minute window.

The largest single lever on stage conversion is not lead quality. It is speed. The foundational study here is Oldroyd, McElheran and Elkington, "The Short Life of Online Sales Leads," Harvard Business Review (2011), which analysed 1.25 million inbound leads across 2,241 US companies. It is US data, but it measures buyer behaviour in the minutes after an enquiry, which does not change at the equator: the same response-time effect shows up in Australian inbound pipelines. It remains the most-cited empirical work on lead response, and its findings are stark.

Qualification odds do not decline gently. They collapse on a roughly logarithmic clock: the first minutes carry almost all the value, and a lead left for a day is, statistically, a different and far worse asset than the same lead worked immediately.

The operational implication is precise. The expected value of a lead is a function of response time, not just source. An A$200 lead worked in five minutes can outperform an A$40 lead worked tomorrow, because the close-rate term in the lead-value formula is itself time-dependent. Speed is not a service nicety; it is a multiplier on the only variable that converts cost into margin. The cheapest improvement most pipelines can make is to route and contact inbound leads inside the first five minutes, before any spend, targeting or creative is touched.

Win rate by source.

Sources do not convert equally, and the gaps are enormous. Inbound, website-generated leads close at roughly 31 percent, more than double the overall average. Referrals sit near 25 percent. Organic leads close around 15 percent. Pure outbound lists convert at about 1.7 percent, and cold email near 1 percent (Digital Bloom / SalesHive aggregations, US and international data, 2025). A lead from the best source is worth twenty times a lead from the worst, before you account for cost.

And yet most companies still do not track win rate by source at all. We deliberately avoid a precise figure for how many, because the percentages that circulate do not trace back to a primary survey we would stake a claim on, but the pattern is not in dispute: the default reporting view is a single blended win rate. That is the single most expensive measurement gap in lead generation. Operators blend a 31 percent source and a 1.7 percent source into one "average" and allocate budget against a number that describes neither. Win rate by source is the join that turns CPL into CPA and lets you reallocate toward the sources that actually close.

Pipeline velocity ties it all together.

Everything so far feeds one operating metric. Pipeline velocity measures how much revenue your pipeline produces per day, combining volume, value, conversion and time into a single rate you can manage.

A worked figure: 50 open opportunities × A$20,000 average deal × 22 percent win rate = A$220,000 of expected pipeline value; divided by a 67-day cycle, that is A$3,284 of pipeline produced per day. Halve the cycle and velocity doubles, with no change in spend or win rate. This is why cycle compression is so valuable, and so often ignored.

Industry benchmarks (US data below, deal values in USD) show the four levers trading off against each other. Construction wins on velocity despite the lowest win rate and longest cycle, purely on deal size. SaaS competes through speed and conversion on smaller deals. There is no single right shape; there is only knowing which lever is yours to pull.

Cycle length is getting worse, which makes velocity harder to hold

The overall B2B sales cycle has stretched to roughly 6.5 months, up from 4.9 in 2019, with the average deal now involving 6.8 stakeholders against 5.4 in 2020, and compliance reviews adding two to four weeks (Gradient Works, US and international B2B data, 2025). A lengthening denominator silently erodes velocity even when volume and win rate hold. Measuring cycle length, and attacking it deliberately, is now part of defending profit, not just accelerating it.

The operator's playbook.

The economics reduce to a handful of moves, ordered by leverage and inversely by cost.

1. Stop reporting CPL alone. Pair every CPL with its close rate and deal value, and compute effective CPA per job type. CPL without conversion is a number that ranks your channels backward.
2. Set the CPL ceiling per job type. Use Max CPL = (deal value × close rate) / target return. Buy aggressively below the ceiling for high-value work; a single blended target leaves money on both sides of the table.
3. Contact inbound leads inside five minutes. The Oldroyd HBR data makes this the highest-return, lowest-cost change available. Route and respond before you touch spend or creative.
4. Track win rate by source. Most operators do not. The join between source and outcome is what turns lead cost into customer cost and tells you where to reallocate.
5. Fix the MQL-to-SQL handoff. It is usually the largest leak and the cheapest to repair, because it is a definition and routing problem. A few points here compound through every downstream stage.
6. Manage to pipeline velocity, not activity. One number, four levers. Attack the cycle length as deliberately as you chase volume; halving the cycle doubles output with no extra spend.

None of this requires more spend. It requires holding the right unit of account (lead value, not lead cost), measuring the stages most operators leave dark (source-level win rate, stage conversion, response time), and managing the one metric that ties them together (pipeline velocity). That is the whole of lead and pipeline economics: convert sales effort into durable margin faster than you did last quarter.