Cost Per Lead: Formula, Benchmarks, and What a Good CPL Looks Like in 2026

Introduction

You’re looking at two quotes for the same work. One agency says $85 a lead. The other says $310. Same market, same ideal customer profile, and nothing in either proposal explains the gap. So which one is expensive?

That question is why cost per lead is the most quoted and least useful number in B2B marketing. It’s cheap to calculate, easy to put in a deck, and it hides the one thing you need to know. Martal Group has run outsourced sales and lead generation for B2B companies since 2009, across 50+ verticals, and the pattern repeats: the cheapest lead on the invoice is rarely the cheapest lead in the pipeline.

Most of that $85-to-$310 gap comes down to two things: what each agency counts as a lead, and where its B2B leads come from in the first place.

Cost Per Lead, Answered in Five Lines

1. Cost per lead is the total cost of a marketing or outbound campaign divided by the number of leads that campaign produced.
2. The formula is CPL = total campaign spend ÷ number of new leads generated, so a $10,000 campaign that produces 80 leads has a CPL of $125.
3. Total spend means everything the campaign consumed: media, tooling, data, agency fees, and the loaded cost of the people running it.
4. A lead is whatever your organization has defined it to be, which is why CPL is only comparable between campaigns that use the same definition.
5. CPL says nothing about whether those leads convert, so it belongs next to your conversion rate and your cost per sales qualified lead, never on its own.

What Changed in CPL Benchmarks in 2026

Search-ad lead costs fell for the first time since before 2020. WordStream’s 2026 benchmarks report puts the all-industry average search CPL at $66.69, down from $70.11 the year before, across 13,474 US campaigns running April 2025 through March 2026.

Conversion rates rose in 87% of industries in that same dataset, which means the cheaper leads came from better closing rather than cheaper clicks (WordStream).

Paid search turned out to be the most expensive B2B lead source, not the cheapest. Across 153 B2B advertisers and $57.6M of 2025 spend, Google Ads produced leads at $524 against $202 on LinkedIn and $145 on Facebook (Metadata).

The most-cited industry benchmark is older than it looks. First Page Sage’s current cost per lead report runs on data collected between January 2022 and June 2025, so treat it as a multi-year average rather than a reading of this quarter.

In-house sales development got harder to justify on cost. The Bridge Group’s 10th biennial study found 60% of SDRs hitting quota, the lowest share in the study’s history, against a median on-target package that has barely moved in a decade.

CPL Terms That Change the Number You Get

Lead is any contact who has shown enough interest to hand over their details. It is the loosest term on this page and the reason two CPLs are rarely comparable.

MQL is a marketing qualified lead: someone who responded to outreach and matches your ideal customer profile.

SQL is a sales qualified lead: someone interested in a next step, asking for information or a meeting.

Booked is an SQL with a confirmed meeting on a calendar. An SQL is interest; booked is a commitment, and collapsing the two is the most common way a pipeline report flatters itself.

CPA is cost per action, where the action is whatever you defined it as, so it is CPL wearing a broader label.

CAC is customer acquisition cost: total sales and marketing spend divided by new customers won. It includes the sales team; CPL does not.

Blended CPL averages paid and organic together. It is the number most benchmark tables report and the number least like your own.

How This Guide Was Built, and Why

The figures below come from four primary sources: First Page Sage on industry CPL, Metadata on B2B paid media, WordStream on search advertising, and The Bridge Group on sales development economics. Each one is read against Martal Group’s own experience running B2B outbound and appointment setting since 2009.

We wrote it because most CPL articles quote a single average without saying what that average measures, which leaves buyers comparing quotes that were never comparable.

What Is Cost Per Lead?

Cost per lead (CPL) is a marketing metric that measures the average amount spent to acquire one new lead, calculated by dividing total campaign cost by the number of leads that campaign generated.

It is the bridge between a marketing budget and a sales pipeline, and it is the first number most finance teams ask for because it turns a spend into a unit price.

And the metric earns its popularity honestly. Impressions and clicks tell you a campaign moved; CPL tells you it produced something a salesperson can call. It works across email, phone, LinkedIn and paid media, which makes it one of the few figures that lets you put wildly different B2B lead generation strategies on the same axis.

But it won’t tell you whether any of those leads were worth having. CPL counts the VP of Operations who booked a demo as one lead. It counts the student who downloaded your whitepaper for a class project as one lead too. That one fact starts most arguments with vendors.

What CPL measures, and what it quietly leaves out

It measures: total cost in, total leads out, for one campaign over one period.

It leaves out: lead quality, fit against your ICP, and conversion rate to opportunity. It also misses deal size, sales cycle length, and every dollar your sales team spends working those leads after they arrive.

So CPL is a cost metric, not a performance metric. Treating it as performance is how a marketing team ends up celebrating a number the sales team can’t use.

The Cost Per Lead Formula, and How to Calculate It Properly

The formula is one line:

CPL = total campaign spend ÷ number of new leads generated

Spend $10,000 on a campaign that produces 80 leads and your CPL is $125. Nobody gets that part wrong. The numerator is where it goes sideways.

What belongs in “total spend”

Most CPL calculations flatter you, because they only count media. Any defensible number includes everything the campaign consumed:

• Media and platform spend. Ad budget, sponsorship fees, event costs.
• Tooling and data. Sequencers, enrichment and verification, plus any dialer or CRM seats attributable to the campaign.
• Sending infrastructure. Domains, inboxes, warming, deliverability monitoring, if you’re running email at volume.
• People. Count the loaded cost of the SDRs, marketers and managers who worked on it, at the fraction of their time it actually consumed.
• Agency or vendor fees. Retainer, setup, and anything billed separately.

Leave the last two out and your CPL will look roughly half of what it is. Which is fine as a vanity number and dangerous as a planning input, because you’ll budget the next campaign against a cost that doesn’t exist.

A worked example, both ways

Say you run a quarter of outbound. Media and tooling come to $18,000. One SDR at a $80,000 on-target package spends 60% of their time on it, which adds about $12,000 for the quarter. The campaign produces 240 leads.

Media-only CPL: $18,000 ÷ 240 = $75.

Fully loaded CPL: $30,000 ÷ 240 = $125.

Same campaign, same leads, a 67% difference. And when you compare your in-house cost against an agency quote, this is usually the entire disagreement. The agency number is the loaded one.

The same formula across six channels

The arithmetic does not change from channel to channel. Only the inputs do.

• Paid social. You spend $12,000 on LinkedIn ads and collect 60 form fills, so your CPL is $200.
• Paid search. A $9,000 Google Ads quarter that produces 18 demo requests costs you $500 a lead.
• Cold email. A program running $4,500 in tooling, domains and SDR time that produces 30 replies worth following up comes to $150.
• Webinar. A $6,000 webinar with 120 registrations lands at $50 each, though a registration and a lead are not the same thing.
• Trade show. A $40,000 budget for booth, travel and staffing that yields 50 scanned conversations works out at $800 a lead.
• Content and SEO. $15,000 of writing, design and promotion over a quarter that generates 75 gated downloads gives you a $200 CPL.

Those six are arithmetic, not benchmarks. What each channel actually costs in your market sits further down, in the benchmarks section, and the spread there is wider than these round numbers suggest.

How often should you recalculate it?

Monthly for an active paid campaign, quarterly for outbound. Any shorter and you’re reading noise; any longer and you’ll miss a channel degrading. Recalculate whenever your lead definition changes, because the old number stops being comparable the moment it does.

What Counts as a Lead Is the Whole Ballgame

Two vendors quote you $85 and $310 per lead. Before you conclude one is three times more efficient, ask each of them to define a lead in one sentence. The $85 vendor usually means a form fill or a contact who opened and replied. The $310 vendor usually means a qualified conversation with a named decision-maker who matches your criteria and asked for a next step. Those are different bars, and how you qualify sales leads decides which one you’re paying for.

But those are not the same product. They are not even in the same category of product. And the metric name is identical.

Users in Reddit and community discussions ask a version of this constantly: how do you estimate what a lead should cost before you’ve run anything, and how do you tell whether a per-lead price is fair? And the consensus answer from people who have been burned is the same one we’d give. Settle the definition before the price. A vendor who can’t define a qualified lead in one sentence doesn’t have a definition, and you’ll be arguing about it in month three.

The five questions that make a CPL quote comparable

1. What has to be true for something to count as a lead? Job title, company size, stated interest, or just a reply. If your ideal customer profile isn’t written down, neither is your lead definition.
2. Who decides? Your team, theirs, or a shared rubric written down before launch.
3. What’s the rejection policy? If you reject a lead as unqualified, does it come off the invoice.
4. Is the lead exclusive? A lead sold to three of your competitors is worth a fraction of one sold to you alone.
5. What’s included in the price? Setup, data, domains, and tooling billed separately turn a $200 CPL into a $280 one.

Ask those five and the spread between quotes usually collapses to something explainable. Skip them and you’ll pick on price and renegotiate on definition.

Pay per lead, and why it behaves the way it does

Pay per lead is a pricing model where you’re billed a fixed amount per lead delivered, regardless of what happens next. It moves volume risk onto the vendor, which sounds like the buyer wins.

But the incentive runs the wrong way. A vendor paid per lead is paid to produce leads, so the pressure pushes toward the loosest defensible definition. That’s not dishonesty, it’s arithmetic. If you use the model, the rejection policy and the written definition are the whole protection, and both belong in the contract rather than the kickoff call.

What’s a Good Cost Per Lead? The Benchmarks, With Their Caveats

A good CPL is one your unit economics can carry. That sounds like a dodge, so here are the numbers, along with the reason they disagree with each other.

By industry. First Page Sage’s report on average cost per lead by industry covers 30 verticals on data collected between January 2022 and June 2025. Blended CPL runs from $91 in eCommerce and $92 in HVAC up to $982 in higher education. B2B SaaS sits at $237 blended, splitting into $310 paid and $164 organic. IT and managed services runs $503, manufacturing $553, financial services $653. The full vertical-by-vertical breakdown, including the paid-versus-organic split for each, sits on our CPL benchmarks by industry page.

By paid channel. Metadata’s benchmark set, drawn from 153 B2B advertisers and $57.6M of 2025 spend, puts LinkedIn at $202 per lead, Facebook at $145, Instagram at $138 and Google Ads at $524 (Metadata). Format matters nearly as much as channel: LinkedIn document ads produced leads at $142 against $362 for conversation ads, and a LinkedIn lead-gen form came in at $193 against $346 for sending the same click to a landing page.

By search advertising. WordStream’s median across 13,474 US search campaigns is $66.69, ranging from $26.84 in arts and entertainment to $131.63 for attorneys and legal services (WordStream).

Now look at those three together. A search-ad form fill costs $67, a B2B paid-media lead costs $202, a blended B2B SaaS lead costs $237, and a B2B Google Ads lead costs $524. Those four figures are not four measurements of one thing that happen to disagree. They are four different products sharing a metric name, measured by four organizations working from four definitions of the word lead.

Which is why the only benchmark that reliably means anything is your own trailing number, segmented by channel, with the definition held still.

How to use a benchmark without misleading yourself

• Check the data window. A 2026 report built on 2022–2025 data is a multi-year average, not a current read.
• Check who’s in the sample. Campaigns from thousands of small local advertisers will not tell you what an enterprise ABM lead costs.
• Check the lead definition. Ad-platform CPL counts conversions. Agency CPL counts deliverables. Your CRM counts something else again.
• Compare within a channel, never across. Your LinkedIn CPL against last quarter’s LinkedIn CPL is a real comparison, because the lead source is held constant. Your LinkedIn CPL against an industry blended average is not.

Your Cost Per Lead and Your Cost Per SQL Are Different Numbers

Leads narrow as they move down the funnel, and the multiplier between the top and the bottom is bigger than most teams model. So this is the point where CPL stops being a budgeting metric and starts being a liability.

Take a real one. On a 14-month outbound engagement Martal Group ran for MAX USA Corp, an industrial tools manufacturer entering the US market, the campaign produced 1,596 leads. Of those, 1,364 cleared the client’s criteria as MQLs, an 85% rate that reflects how tightly the list was engineered rather than how loose the bar was. Those became 203 SQLs. And 203 SQLs became 107 booked meetings.

Run the arithmetic on any CPL you like. At $100 a lead, that program’s cost per SQL is roughly $786 and its cost per booked meeting is roughly $1,492. The headline number and the number that matters differ by a factor of fifteen.

So this is the calculation to run before you sign anything. Take the quoted CPL, apply your own historical lead-to-SQL rate, then your SQL-to-meeting rate, and compare the result against your average deal value. A $300 CPL on a program that converts well can be cheaper per meeting than an $85 CPL on one that doesn’t, and the invoice will never tell you which you bought.

Work out your own cost per SQL in three lines

Three inputs and two divisions get you the number that matters.

1. Start with the CPL, quoted or measured. Total spend divided by leads.
2. Divide by your lead-to-SQL rate. If 100 leads produce 13 SQLs, divide by 0.13, and a $150 CPL becomes $1,154 per SQL.
3. Divide again by your SQL-to-meeting rate. If 203 SQLs produce 107 meetings, that rate is 0.53, and $1,154 becomes $2,177 per booked meeting.

Run it once on your own historical rates and once on the vendor’s claimed ones. Two quotes that looked three times apart on price often land within a few percent of each other here, and occasionally they swap places. Martal’s ROI calculator runs the same chain against the cost of an in-house team if you want both sides chained out together.

Why the in-house comparison usually surprises people

And you can run the same math on an internal team. The Bridge Group’s 2025 SDR research, covering 351 B2B companies, reports a median SDR on-target package of $80,000, an average ramp of three months, average tenure of 1.9 years, and a global median quota of nine fully qualified opportunities a month, with 60% of reps hitting quota.

Those inputs give you a floor, not a forecast. Salary alone against achieved quota lands north of $1,200 per qualified opportunity before you’ve bought a single tool, seat, data subscription, or hour of management time, and before you’ve absorbed three months of ramp on a rep who stays under two years. Add tooling and overhead and most teams land somewhere between $1,500 and $2,500 per qualified opportunity.

So that is the number an agency quote should be measured against, rather than the $75 media-only figure.

CPL vs CPA vs CAC: Which One Belongs in Your Board Deck

All three measure cost per something, and the something is different in each.

• CPL is cost per lead. Marketing spend divided by leads. It stops at the point of interest.
• CPA is cost per action, where the action is defined by the campaign: a demo request, a trial signup, a download. In lead-gen contexts CPA and CPL are often the same figure with different labels, which is why the two get used interchangeably and shouldn’t be.
• CAC is customer acquisition cost. Total sales and marketing spend divided by new customers acquired. It includes the sales team, the tooling, and every lead that went nowhere.

CAC is always the biggest of the three and always the most honest, because it’s the only one that requires someone to have bought something. A $200 CPL can sit underneath a $2,000 CAC or a $20,000 CAC depending entirely on what happens after the lead arrives, and that spread is invisible in the CPL.

So split them by audience. CPL belongs in a channel report, where you’re comparing this month’s email program against last month’s. And CAC belongs in a board deck, where the question is whether the business acquires customers for less than they’re worth. Putting CPL in the board deck is how a marketing team ends up defending a number that has nothing to do with revenue.

Where CPL sits among the cost metrics

Seven figures track the same spend at seven different depths. Each step down the list costs more per unit and means more.

• Cost per click (CPC). You pay for the visit. CPL is CPC divided by your conversion rate, so CPC moves CPL without ever appearing in it.
• Cost per thousand impressions (CPM). You pay for reach. It sits two steps above CPL and says nothing about whether anyone acted.
• Cost per lead (CPL). You pay for a contact who showed interest. Everything downstream of that is still unknown.
• Cost per action (CPA). You pay for a defined action. In lead generation it is frequently CPL under a different label.
• Cost per SQL. This is the first figure in the chain that needs a salesperson to agree the lead is real.
• Cost per booked meeting. This is the first figure tied to something on a calendar.
• Customer acquisition cost (CAC). The only one that requires somebody to have bought.

So optimize a channel with the metrics near the top of that list. Defend a budget with the ones near the bottom. Both sit inside a wider set of lead generation KPIs.

What Agencies Charge Per Lead, and How to Read the Quote

B2B lead generation gets priced four ways, and the model shapes the vendor’s behavior as much as the number does. You’ll see all four across the top lead generation companies, often inside the same proposal.

• Monthly retainer. You pay one fixed fee for a defined scope and capacity. It aligns the vendor to the program rather than to volume, and you carry the performance risk.
• Pay per lead. You pay a fixed price for each delivered lead, so volume risk sits with the vendor. The incentive to define a lead loosely sits there too.
• Pay per appointment. You pay for each booked meeting. It is the tightest alignment of the four and the most expensive per unit, because the vendor absorbs the full funnel.
• Hybrid. You pay a base fee plus a performance component. Most engagements land here, and it is usually the most defensible split, because neither side carries all the risk.

So convert every quote into an expected cost per SQL before you compare them. A $5,000 retainer producing 12 SQLs a month is $417 a lead in the only currency that counts, regardless of how many top-of-funnel contacts appear in the report. And ask what sits outside the monthly number. Setup fees, sending domains, data subscriptions and call tooling get quoted separately more often than not, and that is where most of the sticker shock in community threads comes from.

Check whether the quoted cost per lead includes phone. Cost per call leads from cold calling tend to run higher than email-sourced leads because a conversation costs more to produce than a reply. It also tends to qualify faster, which is the trade you’re actually making.

How to Lower Your Cost Per Lead Without Buying Worse Leads

Any CPL can be lowered by loosening the definition, which is relabeling dressed up as optimization. So diagnose the cause first. A high CPL has a small number of usual sources, and the fix that works for one of them does nothing for the others.

Cause 1: The lead definition drifted

A CPL that looks fine while sales rejects most of what arrives is a definition problem wearing a cost problem’s clothes. The number is honest; the thing it counts isn’t.

• Write the qualification bar down in one sentence and hold every channel to it.
• Recalculate CPL from the date the definition last changed, not from the start of the quarter.
• Report lead count and rejected-lead count side by side, so drift shows up before the invoice does.

Cause 2: Your targeting is too broad

Contacting more accounts than your criteria can support raises volume and lowers fit at the same time, which reads as efficiency right up until the conversion rate arrives.

• Build an engineered lead list instead of exporting one. Accounts get matched against your criteria, scored by fit and buying signal, and kept current while the campaign runs.
• Narrow to one segment, prove the conversion rate, then widen. Widening first hides which segment was carrying the result.
• Cut the accounts that clear your filters but have never bought anything like your product.

Cause 3: Your data is dirty

Bounced emails and dead direct dials are paid-for volume that produces nothing, and they degrade sending reputation on the way through, so the damage compounds.

• Start with cleaning your email list before launch rather than after the first bounce report.
• Verify direct dials as well as emails. A disconnected number burns a rep’s time, which costs more than the record did.
• Re-verify on a cycle. Contact data decays continuously, so a one-time clean has a short shelf life.

Cause 4: The conversion rate is the real problem

CPL is cost per click divided by conversion rate, so a landing page or a call script that converts better lowers CPL without touching spend. WordStream’s finding that conversion rates rose in 87% of industries while CPL fell is that relationship in action.

• Match the landing page to the promise in the ad or the email, word for word where you can.
• Strip form fields down to what qualification genuinely needs.
• Test one variable at a time and hold it long enough to read.

Cause 5: You’re on the wrong format or destination

On paid social the creative format and the place you send the click move cost per lead more than most teams expect, and the expensive options are often the popular ones.

• Metadata’s 2025 data puts LinkedIn document ads at $142 per lead against $362 for conversation ads, and document ads are the least used format in that set.
• The same dataset shows a LinkedIn lead-gen form producing a lead for $193 against $346 for sending the same click to a landing page.
• So test format and destination before you touch targeting. Both are cheap to change. And both move the number fast.

Cause 6: You’re over-weighting retargeting

Warm audiences are assumed to convert cheaper. But on LinkedIn that assumption is wrong. Retargeting there costs $234 per lead against $194 for prospecting (Metadata), while Facebook runs the expected way at $120 against $166.

• Check your own prospecting-versus-retargeting split per channel rather than applying one rule across all of them.
• Cap retargeting where the data doesn’t back it. Move the difference into prospecting.

Cause 7: Your channels are running separately

By the third touch, a prospect who has seen your email, your LinkedIn profile and your call recognizes the name. Three channels firing independently at the same person pay three times for one impression of familiarity.

• Sequence email, phone and LinkedIn as one coordinated campaign rather than three parallel ones.
• One reply should redirect the next touch everywhere.

Cause 8: Your mix is too paid-heavy, or you’re replacing leads instead of working them

Organic CPL is lower than paid in every industry in the First Page Sage dataset. But organic is slower to build. And it doesn’t scale on demand, so this is a portfolio decision rather than a switch.

• Shift mix toward organic where the timeline allows, and keep paid for the months you need volume on command.
• Nurture the leads you already paid for. A lead that isn’t ready now is cheaper to convert in four months than a new lead is to acquire, so a lot of CPL problems are really follow-up problems.

When Chasing a Low CPL Is Exactly the Right Call

CPL does have a job, and the case for it is a real one worth stating plainly.

Inside a single channel, with the lead definition fixed and the audience held constant, CPL is a genuinely good optimization metric. It’s fast, it’s sensitive, and it moves before anything downstream does. If you’re testing two LinkedIn ad formats this week, waiting for cost-per-SQL data would mean waiting a quarter for a sample too small to read. CPL gives you an answer on Friday.

There’s a second case. Early-stage programs often don’t have a conversion rate yet. You can’t compute cost per SQL from three SQLs, and pretending you can produces a number that swings forty percent on a single deal. In that situation CPL plus a hard qualification bar is the right instrument, and cost per SQL is something you graduate to once volume makes it stable.

So the problem was never CPL itself. Trouble starts when it gets compared across channels, across vendors, or across definitions, and the result is treated as though it meant something. Keep the comparison inside one lane and the metric does honest work.

The Number to Manage, and What to Do This Week

Cost per lead is a useful unit price and a poor buying signal. It compares cleanly inside one channel with one definition. And it falls apart the moment either of those changes, which is precisely when people reach for it.

So do three things. Write down your organization’s definition of a lead, an MQL and an SQL, in one sentence each, and make every vendor quote against it. Recalculate your own CPL with people and tooling included, so you know what you’re actually comparing against. Then work out your cost per SQL and your cost per booked meeting, and manage those.

If you’d rather have someone run that math against your funnel before you commit a budget, Book a consultation. We’ll model what your leads cost today, what they’d cost through a managed program, and where the difference comes from.