SaaS lead generation

A lead can cost 12 dollars or 1,400 dollars, and the cheap one is not automatically the better buy. Customer referrals convert into real opportunities at roughly four times the rate of a LinkedIn lead form fill, which means a blended cost per lead figure hides the only thing worth knowing. The source is the unit of analysis, never the total.

Treat this as portfolio management. Ten sources exist, each with its own price, its own conversion rate into pipeline and its own lag before the first lead lands, and picking three of them on purpose beats running all ten badly. Any lead target set without a quality floor and a payback ceiling attached is a budget-burning exercise, and most quarterly plans I see ship without either.

What counts as a lead, and the six rungs between a visitor and a deal

A lead is whatever your CRM says it is, and that is where most marketing versus sales reporting fights start. The fix is a ladder with written definitions for every rung, agreed by both teams in the same room on the same afternoon, then enforced in the CRM as a required field rather than a convention.

Six rungs cover almost every SaaS motion. A visitor is an anonymous session. A hand raiser gave you an email address through a form, a trial signup or a demo request. An MQL is a hand raiser that matches firmographic fit and showed at least one intent signal. A PQL is a free account that crossed a defined usage threshold. An SAL is a lead sales has accepted and owns. An SQL is a lead that survived a discovery call with need and timing confirmed.

Two definitions cause more damage than the rest combined. The first is treating an ebook download as an MQL, which loads the queue with people who wanted a PDF and had no intention of buying anything. The second is counting a lead as accepted the moment it routes to a rep rather than when the rep works it, which quietly turns your SAL rate into a routing statistic.

Speed to first touch is the one lever on this ladder that costs nothing and is almost always broken. The lead response study Harvard Business Review published in 2011 found the odds of qualifying a lead fell by roughly 21 times between a five minute and a thirty minute response. Fifteen years later the median B2B SaaS response time is still measured in hours. Fix routing before you buy more leads.

PQLs deserve their own rung rather than a footnote. The threshold has to predict retention, not just activity, which means running the correlation against your own cohort data before you pick it. In collaboration products the strongest predictor is nearly always a second human: a teammate invited inside the first week. That pattern is what built Slack, Figma and Miro, and it is measurable in your database this afternoon.

What every lead source costs and how long it takes to produce one

Ten sources cover the realistic options for a B2B SaaS company. The table below gives the range I see in practice for each one, along with the conversion into opportunity and the lag before the first lead lands. Read the lag column first, because it is the one that wrecks quarterly plans.

Notice what the table does to the usual argument. Paid social produces the cheapest volume at the top of the range and the worst opportunity rate on the board, so the cost per opportunity is often four times that of organic capture even when the cost per lead looks half as good. Run the multiplication before the meeting, not during it.

The lag column is where plans die. A board asks for pipeline in Q1 and someone answers with a content plan, which will produce its first real lead somewhere around month five. That is not a failure of content, it is a failure of sequencing. If you need leads inside six weeks the honest answer is paid search, a marketplace listing and outbound, and nothing else on this table qualifies.

If you want the sources ranked rather than tabulated, with the reasoning for each position, read SaaS lead generation strategies, ranked. For a view of which of these actually produces the pipeline in aggregate across companies, rather than what any one team believes, we keep the data in where B2B SaaS pipeline actually comes from.

Turning a 3 million dollar ARR target into a monthly lead number

Work backwards from revenue, never forwards from a channel budget. The cascade takes ten minutes and it is the single most useful artefact a demand team produces all year, because it converts an argument about tactics into arithmetic anyone can check.

That last step is the point of the whole thing. Run the model again with hand raiser to MQL at 35 percent instead of 55, which is exactly what happens when you add a volume-heavy paid social line to hit the number. You now need 605 hand raisers a month, the blended cost per lead has crept to 185 dollars because the new source is not cheap, and monthly spend jumps from 57,750 to about 112,000 dollars. CAC goes from 7,700 to roughly 11,600 and payback stretches past ten months.

Same revenue target. Same team. A 50 percent worse CAC, caused entirely by a decision that looked like it was about volume.

The payback ceiling: the most you can afford to pay for a lead

Set this number before you set a lead target, because it is the only thing standing between an ambitious plan and an unfundable one. The formula has three inputs you already know.

Maximum allowable CAC equals ACV multiplied by gross margin, multiplied by your target payback in months, divided by twelve. At an 18,000 dollar ACV, 78 percent gross margin and a twelve month payback target, that is 18,000 times 0.78 times 12 divided by 12, or 14,040 dollars.

Then convert CAC into a lead price. Multiply your stage conversion rates end to end: 55 percent hand raiser to MQL, 60 percent MQL to SAL, 50 percent SAL to SQL and 22 percent SQL to closed won gives 3.63 percent, or one customer for every 27.5 hand raisers. If marketing carries 45 percent of total CAC and sales carries the rest, marketing’s share is 6,318 dollars per customer, which divided by 27.5 gives a ceiling of roughly 230 dollars per lead.

Every source in the table above that prices below 230 dollars is fundable at this ACV. Paid search at the top of its range is not, and neither is a field event, which is why events have to be justified on something other than lead count. That is a legitimate argument to make, it just is not a lead generation argument.

The ceiling changes the moment ACV changes, which is why we publish the lead generation playbooks by ACV band separately rather than pretending one plan fits a 3,000 dollar product and a 90,000 dollar one. At 3,000 dollars ACV and the same margin and payback assumptions, the ceiling falls to about 38 dollars a lead, which removes paid search, events, outbound and most of paid social from the board in one stroke.

The quality floor, and what happens the month somebody removes it

A quality floor is a written rule that a lead does not count unless it meets a fit test. Company size band, industry, geography, and where relevant a technology or spend signal. It goes in the CRM as a hard gate, not a scoring model, because scoring models get tuned until they produce the number the tuner needs.

Here is the failure I have watched at three companies now. Pipeline is short at week nine of the quarter, someone opens the paid social account and lifts the audience from a tight 300 account list to anyone with the word marketing in their title, and lead volume jumps 40 percent inside eight days. The lead target gets hit. The meeting is calm. Eleven weeks later the opportunity number is down and nobody links the two events because the reporting lag is longer than the memory of the meeting.

That quote is a composite of three conversations rather than one person’s words, and it is the most common shape of failure in this category. The fix is dull and effective: report cost per opportunity next to cost per lead on the same row, every week, and let nobody publish one without the other.

Set the floor at the ICP definition your best customers actually match, which you derive by exporting closed won accounts from the last eighteen months and looking at the distribution rather than at the logos you wish were in there. Then accept the consequence. A real floor cuts reported lead volume by somewhere between 20 and 45 percent the month you apply it, and someone senior will ask why leads fell. Have the cost per opportunity chart ready for that meeting.

Why the published lead generation benchmarks disagree with each other

They disagree because they measure different populations and almost none of them publish sample sizes. This matters because these numbers end up in board decks as though they were comparable, and they are not.

First Page Sage publishes SaaS customer acquisition cost by channel and puts organic search near 205 dollars against paid search near 341 dollars. Those figures circulate widely, and the firm does not disclose sample size, company stage or ACV band alongside them. Benchmarkit’s annual B2B SaaS performance study reports a median CAC payback around 16 months from a survey base of roughly a thousand companies. ChartMogul measured median free trial to paid conversion at 8.9 percent across about 200 products. HubSpot’s marketing surveys are self-reported by marketers, which reliably flatters response rates and conversion figures.

Now put two of those side by side. A 205 dollar customer acquisition cost and a 16 month payback period can only coexist if a customer generates about 13 dollars of monthly gross profit, which implies an annual contract value near 200 dollars. Almost nobody reading this sells a 200 dollar product. The two numbers describe different populations and different definitions of CAC, and stacking them in one slide produces a plan nobody can hit.

Three sources, not nine

Pick three. One that produces leads inside two weeks, one that compounds over quarters, and one that converts at a high rate even though it will never scale. Paid search, organic capture and referral is the default combination for most companies between 1 and 10 million ARR, and it works because the three fail in uncorrelated ways.

The discipline is in the fourth. Do not add it until the second source clears its payback ceiling on its own numbers, not on blended ones. Teams break this rule constantly, usually because a new channel is more interesting than a functioning one, and the result is five half-funded programmes that each miss the threshold where they would have started working. A LinkedIn programme at 4,000 dollars a month is an expensive way to learn nothing.

Sequencing by ACV helps. Under 10,000 dollars ACV, lead with organic capture and free tools, add a marketplace listing, and keep outbound out of it because the economics do not clear. Between 10,000 and 40,000, the combination above is right, with a G2 or Capterra listing as the fourth once you have the reviews to win the click. Above 40,000, outbound lead generation for SaaS becomes the anchor and everything else supports it, because the payback ceiling is high enough to fund a rep.

Referral is the source everyone agrees is best and nobody staffs. It converts at 20 to 35 percent into opportunity, it costs almost nothing, and it requires one person owning a repeatable ask at a defined moment in the customer lifecycle. Most companies leave it to chance, then describe the result as word of mouth. The mechanics are in referral lead generation for SaaS, and the ask point matters more than the incentive.

If you want the full build sequence rather than the portfolio logic, the B2B SaaS lead generation playbook runs the first ninety days channel by channel, and the channel by channel guide scores every source on four axes including the preconditions each one needs before it can work at all.

What breaks first, and what this costs you

Routing breaks first. A lead arrives at 4pm on a Friday, sits in a queue over the weekend, gets a first email on Tuesday morning, and the buyer has already booked a call with a competitor who answered in nine minutes. No amount of channel optimisation recovers that, and it is the cheapest thing on this page to fix.

Attribution breaks second, and it breaks permanently. Once you run four sources, the same buyer touches three of them before converting, and your platform reporting will confidently credit whichever one carried the last click. Add a single self-reported field to the demo form asking how the buyer first heard about you. Self-reported data and platform data will disagree by a wide margin, and the honest answer is to report both and explain the gap rather than to pick the flattering one.

The third thing that breaks is the team’s tolerance for the lag. Organic capture takes four to eight months to produce its first meaningful lead, which is longer than the patience of most leadership teams and longer than the tenure of a lot of demand gen managers. Going in without a written agreement on when you will judge the channel means you will kill it at month five, two months before it would have worked.

The honest cost of all this: the modelling and instrumentation above takes a competent demand gen lead roughly three weeks of real work, and it makes the reported lead number go down. If your organisation rewards the lead number, doing this properly will look like failure for one quarter before it looks like sense.

What to do in the next two weeks

Build the cascade first. Take your new ARR target, your ACV and your trailing twelve month close rate, and produce the monthly hand raiser number. Do it in a spreadsheet you can show people, because the argument you are about to have is about inputs, not about channels.

Then calculate the payback ceiling and compare it against what you currently pay per lead by source. Most teams find at least one line that has never cleared it, and killing that line funds the two that do. If your internal numbers are too thin to support the exercise, start from the ranges in the table above and replace each cell as your own data arrives.

Last, decide which three sources you are actually running this year and write down the date you will judge each one. If you are considering outsourcing any part of this, read hiring a SaaS lead generation agency before you take the first call, because the retainer conversation goes very differently once you arrive with a payback ceiling in hand.