Lead generation: why more leads is not going to solve your revenue problem

The number that looked good was the one nobody could use

One company we worked with this spring illustrated this perfectly. The CEO's question was simple: What would doubling the number of qualified opportunities in his pipeline cost in advertising and personnel?

Before answering, as we always do, we asked what they already had.

Channel 1 was a paid listing on a software comparison site that costed about £12,000 across the year. This had produced sixteen enquiries, of which the team had qualified 12 as real. Channel 2 was organic search, costing nothing extra, which had brought 86 enquiries over the same twelve months.

"We have cracked Organic Search, and now want to add paid acquisition to the mix", the CEO said.

We asked if we doubled the number of opportunities, could his team absorb the additional 86 meetings, or if he was doubling headcount too?

He said he wouldn't need to, because "only one or two inbound queries are actually real enough for the team to work with"

And we saw his demeanour changed as he said it out loud. It dawned on him, as it does on most CEOs when they reflect on this, that the revenue problem wasn't a volume problem. And if he did try to solve it with more volume, the actual bill he'd run wouldn't just be the ad spend, but the cost of finding, hiring, ramping and managing two additional sellers. Or the cost of churning employees, if he refused to create the headcount.

Start with the one number this turns on

Take your last 40 first meetings, in date order. Mark each one yes or no on whether it became a real opportunity. Divide.

That is your meetings-to-opportunity rate. It takes an afternoon.

If your CRM stages are too messy to give it to you, leave the CRM alone. Forty meetings will tell you roughly where you are, and roughly is enough to make the next decision.

Then do the harder half, which is asking why.

Sit with the ones that died. Was it the wrong company? The wrong person inside it? Or the right person, with no problem worth a budget this year?

A rate on its own tells you the funnel leaks. It does not tell you where. And a leak you cannot locate cannot be priced against the alternatives.

Then treat what you find as a reading, not a ceiling.

One in four is what your current positioning and your current message produce, against the people you are currently reaching. It is not what the system could do.

Everything else here follows from that difference. A company that treats its conversion rate as a fact of nature will always conclude it needs more leads, because the only number it believes it can move is the one at the top.

Four routes to the same revenue number

Here is the arithmetic we ran on that call. With your own figures it takes ten minutes.

Start at the end. They wanted £500,000 of new revenue against an average contract of £30,000.

That is 17 deals.

This team wanted three times pipeline cover before they would call the plan safe. So 17 deals needed about 50 live opportunities.

One opportunity came out of roughly every four first meetings. So 50 opportunities needed 200 meetings across the year. About 17 a month.

Then price a meeting. Take last quarter's demand spend and divide it by the meetings it produced. That is the marketing cost of putting one qualified person in front of a salesperson.

Theirs came to £250.

At £250, those 200 meetings cost £50,000 across the year. A bit over £4,000 a month.

It answers the question the meeting was called to settle.

It is not the only answer to £500,000.

One step of conversion is worth £12,500 a year here. A quarter of the budget.

And unlike the budget, you do not buy it again next year.

One step means one extra meeting in twelve becomes a real opportunity. It happens when the people arriving sit closer to the people who buy.

The awkward part is that volume is the only row a team can act on by Friday. Raise a budget, loosen the targeting, ask the agency for more, report back next month.

So it wins the meeting.

The conversion row asks harder questions. Who is arriving? What problem do they have? Does your message name it? Is anyone there able to spend money fixing it?

Those are not marketing questions. They are revenue questions, and they take longer than a month.

The lower your contract value, the sooner the volume route runs out of road.

Swap the £30,000 contract for a £3,000 one. The same £500,000 now needs 167 deals, about 500 opportunities, and about 2,000 meetings.

At £250 each, that is £500,000 of marketing to earn £500,000 of revenue.

Why more leads makes the leads worse

The volume route has a second cost the arithmetic does not show.

To get materially more leads out of the same market, you widen the targeting or you lower the bar on what counts. Sometimes both.

Either way you reach further from the people who fit.

So the average arrival gets worse as the count goes up. The damage shows up two stages later, as a conversion rate nobody can account for.

You bought volume and paid for it in the number you were already failing to measure.

One business owner had already run the experiment, and wrote it up: "We don't really need more leads for the sake of having more leads. I'd much rather have fewer enquiries but from people who actually value what we do and are comfortable spending £3k+."

The volume brought people who wanted a different product at a different price. Each of those conversations cost him the hour a real one would have.

Ad platforms do it faster.

Ask a paid search practitioner why accounts drift and you get this: "Google is already pretty good at finding more of whatever you tell it is valuable."

The accounts in trouble, he says, are the ones "still telling Google that every form submission is equally valuable". Then: "people wonder why Smart Bidding keeps finding cheap leads that never buy."

The platform is not confused. It is obeying.

Tell a system to produce arrivals and it gets better at producing arrivals every quarter. Nothing in that instruction asks whether the person has a problem, whether anyone owns it, or whether they can buy this year.

The two numbers that lie to you about this

When a business goes looking for the fault, it reaches for one of two metrics. Both measure the stage before the one that matters.

The marketing qualified lead.

A marketer writing up a bad quarter reports that his head of sales told him "it's because MQLs aren't converting to SQLs". Then the detail that explains the fight: "When we set our goals, he expected a 20% conversion rate from MQL to SQL and the actual rate is way below that mark."

Two teams agreed a rate before either had measured one. They missed it. Then they argued about whose fault the miss was.

Nothing in that conversation was about a buyer.

A rate set as a target cannot be diagnosed when reality comes in underneath it. There is no finding to make, only somebody to blame. That seam is where marketing and sales stop being one system, and we have written about it in marketing and sales alignment.

The deeper fault is what the qualification is built from. Pages viewed, emails opened, a job title matched against a list.

None of that establishes that a problem exists inside that company, or that somebody there is accountable for fixing it.

A lead score built on your own telemetry predicts your own telemetry.

Cost per lead.

It has a denominator. It falls when you do well. It lets two channels sit on one line of a report.

It will also point you at the wrong channel with total confidence.

An advertiser in the US, comparing his spend against an agency's on the same account, described their leads at "roughly $350, which is fine in the solar world if we could actually get some to convert". His own came in at "a cost per lead of approximately $27".

At the stage he measured, $27 wins.

He asks whether the expensive leads convert. He never asks it of the cheap ones.

Go back to the two channels on that call. The listing cost £750 an enquiry and three quarters of its enquiries were real. Organic cost nothing, and what happened to its 86 was the thing nobody in the room could answer.

Cost per lead picks organic. It picks it because the stage where cost per lead is calculated is the only stage organic was ever measured at.

The price of a lead and the quality of a lead rose together there, which is the reverse of what the metric assumes.

The number to hold yourself against is cost per customer, by channel.

Reach everyone. Count only what converts.

There is a serious case against narrow targeting, and it deserves an answer rather than a dodge.

Byron Sharp and the Ehrenberg-Bass Institute for Marketing Science argue in How Brands Grow that brands grow chiefly by acquiring more customers, and that narrow targeting is a mistake rather than a discipline.

Alongside it sits the finding most quoted in B2B. Only about 5% of business buyers are in the market at any given moment, because firms replace things like banking, software and telecoms on cycles measured in years. That is the 95:5 rule, from John Dawes at Ehrenberg-Bass.

None of it is an argument for lead targets. It is an argument against them.

If 95 of every 100 people you reach cannot buy this quarter, reach them anyway. You want to be remembered when they can.

The damage starts when a lead target asks those 95 to behave like buyers today. They get written down as though they could buy, because the counting demanded it. That is where the junk enters.

Sharp counts customers acquired. Dawes tells you to spend on being remembered by people who will fill in no forms today.

Neither counts forms.

Reach as wide as the budget allows. Hold the counting to what converts.

Count the number you are paid for

You do not need to dismantle your reporting.

You need one conversion rate you have measured rather than agreed. An account of why it is what it is. And the nerve to make it the number the demand budget answers to.

Lead volume then becomes an input to a forecast instead of the point of the exercise.

Where this sits in our own model is the first gate, Position: who the buyer is, what the message aims at, which channels carry it. It fails in one characteristic way, which is volume poured through a definition of the buyer nobody ever settled. The gate after it is Activate, where interest hardens into conviction somebody will stake a budget on. A lead count sits between the two and measures the output of neither. Both are defined in the Demand Architecture Glossary.

The company on that call did not need more leads.

It needed to know what happened to 86 of them.

The arithmetic is the easy half

Everything above is a spreadsheet. You can build it in an afternoon and most of you should, before you spend another pound.

The hard half is what you do with what it tells you.

Your conversion rate will tell you that something is wrong. It will not tell you that your positioning is the problem, that the segment you built the company on has moved, or that the message you have defended for two years is aimed at somebody who does not sign.

Those findings are expensive to reach alone, because you are the person who made the decisions they overturn.

When you are close to something, and you do not know what good looks like, you are not the best person to deal with it. It is why doctors are told not to treat their own families. It is not a question of skill. An award-winning cardiologist is still the wrong person to diagnose his own child.

You need a degree of objectivity to be honest about the finding and about the fix.

Run this

For whoever owns the funnel.

1. Get the rate. Last 40 first meetings, yes or no on whether each became a real opportunity, divide.
2. Get the reason. Read the ones that died. Wrong company, wrong person, or right person with no problem worth budget this year.
3. Write your revenue target and divide by average contract value. That is your deal count, rounded up.
4. Multiply by your pipeline cover for the opportunities you need live.
5. Multiply by the meetings each opportunity takes. One in four means a multiplier of four.
6. Price a meeting. Last quarter's demand spend divided by the meetings it produced. Marketing cost only.
7. Run it again with the conversion rate one step better. The difference is what a quality improvement is worth to you per year.
8. Audit your two largest channels on cost per customer. A channel where you cannot complete that row is unmeasured, not cheap.
9. Before approving any lead target, say which of the four inputs it is the cheapest way to move. If nobody can, it is a wish with a number on it.

Thresholds worth having. Under about 30 opportunities a year in a channel, your conversion rates are anecdotes: plan in ranges, not points. A meetings-to-opportunity rate below one in five is usually a targeting problem rather than a selling problem. And if cost per lead is falling while cost per customer stays flat or rises, your leads are getting cheaper because they are getting worse.

Frequently asked questions

Why aren't my leads converting?

Most often because they were never people who could buy. Check the definition first. If a lead in your business is a form submission, a download or a job title on a list, you have selected for people who engage with your marketing rather than companies with a problem and a budget. Fix the definition before the follow-up process.

Is more leads ever the right answer?

Yes, in one situation. Everything downstream converts at a rate you have measured, you can say why it is what it is, and the arithmetic still leaves you short of meetings. Without the measured rate the answer is no, because you cannot tell whether volume is the cheapest input to move.

What is the difference between lead generation and demand generation?

Lead generation captures contact details from people already looking. Demand generation creates the conditions in which people start looking, then captures them. The distinction matters most for measurement: demand work pays back over quarters, and judged on a monthly lead count it will look like failure while it is working.

Why don't MQLs convert to SQLs?

Because most scoring models are built from what a company can see about itself. Pages viewed, emails opened, a job title matched against a list. None of that establishes that a problem exists inside the company which somebody is accountable for fixing, so the leads a model promotes were never selected for their ability to buy.

What should marketing be measured on instead of MQLs?

Two numbers. The conversion rate from the stage marketing owns to the stage sales owns, measured rather than agreed, and the cost of a qualified meeting. Neither improves by loosening a definition, so neither is easy to game, and both put marketing and sales on one number instead of two.

How do I know if my cost per lead is too low?

Compare it against cost per customer across at least two channels. Where the channel with the lowest cost per lead is not also the channel with the lowest cost per customer, cost per lead is misleading you about where to spend. A cheap lead that never converts costs the same salesperson hour as an expensive one that does.

How many leads do I need to hit my revenue target?

Work backwards. Revenue target divided by average contract value gives deals; multiplied by pipeline cover gives opportunities; multiplied by the meetings each opportunity takes gives meetings. On a £500,000 target at a £30,000 contract value, three times cover and one opportunity per four meetings, that is 200 meetings a year.

Why do sales and marketing disagree about lead quality?

Because they count different objects using one word. Marketing counts qualified leads against a scoring model; sales counts opportunities worth working. Both counts are accurate and they describe different things, so no amount of arguing inside the numbers will settle it. It resolves when both teams are measured on the conversion rate between them.

Key frameworks

The Leads Fallacy. The false belief that more leads automatically produce more revenue. It fails because the methods that raise volume lower average fit, and because funnel arithmetic multiplies rather than adds, so a volume gain at the top can be cancelled by a conversion loss below it.

The CPL Inversion. The pattern where campaigns with a higher cost per lead deliver a lower cost per customer, because the leads convert. A £200 lead that becomes a customer beats a £50 lead that does not.

Full-Funnel Math. The calculation that shows where value is created and destroyed by modelling the whole path from spend to revenue. It reveals that conversion improvements often outperform volume increases.

Position (PACED gate 1). The architecture phase, holding strategy and execution together: the ideal customer profile defined to the level of title, company stage, industry and trigger event, the message mapped to a problem the buyer is accountable for, the channels where that buyer congregates, and the live execution carrying the message through them. Pricing and packaging belong here, because price is a positioning decision before it is a finance one. The gate clears when an ICP-matched prospect raises a qualified hand. Gate metric: ICP-fit of pipeline. Achievable benchmark: 0.85, directional and set deliberately per company. Primary failure mode: volume into broken positioning.

Activate (PACED gate 2). The conversion of passive interest into validated conviction, and the most motion-sensitive phase in the model, because the object being activated changes. In a product-led motion the object is an individual user reaching the in-product action that correlates with retention. In a sales-led motion it is a buying committee aligning around one business case until the group, not the enthusiast, is convinced and the economic buyer has validated the decision. The gate clears when conviction is validated, not assumed. Gate metric: trial-to-activation, or committee conviction. Achievable benchmark: 0.65. Primary failure mode: the Ghost Champion, where months are invested in an enthusiastic mid-level advocate while the economic buyer is never activated.

Sources

1. Business owner's post on enquiry quality and enquiry volume, August 2026. https://www.reddit.com/r/smallbusiness/comments/1vliwmw/marketing_woes_and_attracting_timewasters/
2. Paid search practitioner on conversion inputs and automated bidding, July 2026. https://www.reddit.com/r/PPC/comments/1v88khi/your_google_ads_account_might_not_be_the_problem/
3. Marketer on an MQL to SQL target missed in Q1, March 2024. https://www.reddit.com/r/marketing/comments/1bjqy8s/marketing_responsible_for_sqls/
4. Advertiser comparing their own cost per lead against an agency's on the same account, July 2026. https://www.reddit.com/r/marketing/comments/1v83isp/are_our_third_party_ads_truly_being_hampered_by/
5. Dawes, J. (2021). The 95:5 rule, Ehrenberg-Bass Institute for Marketing Science, published through LinkedIn's B2B Institute. https://business.linkedin.com/marketing-solutions/b2b-institute/b2b-research/trends/95-5-rule · Weinberg, P. and Lombardo, J., "The 95:5 rule is the new 60:40 rule", Marketing Week. https://www.marketingweek.com/peter-weinberg-jon-lombardo-95-5-rule/
6. Sharp, B. How Brands Grow: What Marketers Don't Know, Ehrenberg-Bass Institute for Marketing Science, Oxford University Press.
7. American Medical Association, Code of Medical Ethics Opinion 1.2.1, "Treating Self or Family": physicians generally should not treat themselves or members of their immediate family. https://code-medical-ethics.ama-assn.org/ethics-opinions/treating-self-or-family
8. PacedRevenue, "The Revenue Debt You Cannot See" (v1.0, 29 May 2026; author revision 21 August 2026): the PACED gate definitions and per-gate failure modes. https://pacedrevenue.com/whitepapers/revenue-debt-you-cannot-see/

Hassaan Ahmad, Managing Partner and Chief Revenue Officer. Published 1 September 2026.

Related reading: How to define your ICP · Marketing and sales alignment · Full funnel optimization

Start with the free read, or go straight to the scored one. The PACED diagnostic is free and names the gate most likely to be binding. The PACED Review scores every gate of your engine against your real numbers: £2,499, and it stands alone whether or not you go further.