CPL 2026 calculator - cost per lead in B2B and B2C
Consider CPL by niche with benchmarks. Find out what CPL you can afford - the “ceiling” formula through LTV and lead-to-paid. With shareable URL.
CPL formula - what to substitute and where
CPL = spend ÷ leads. Divide the channel budget by the number of leads for the same period. A lead is an application with a contact - phone or email, by which you can get through. Not a purchase, not a click, not an impression. More about myself CPL term and why is it separate from CPA, and about lead generation as a process - separate articles.
The main mistake of beginners is to count CPL for all channels at once. I have on CPL projects in Yandex Direct and VK Ads differ by 3–4 times for the same same landing page. The average for two channels shows “normal”, but in fact one drags down the economy, the second drowns it. You count CPL - you count for each channel separately, in one period, according to the same definition of lead.
The second is the definition of a lead. If in one table you have in the “leads” email subscription and pilot application for 500,000 ₽ - the number means nothing. I believe that one project can have 2-3 levels of leads: feed subscriber, MQL, SQL. Each with its own CPL and its own benchmark.
CPL benchmarks for Russian niches 2026 - why B2B is 8 times more expensive than e-com
The table below is my working niche range for 2026: a guideline for comparisons, not an industry standard. Numbers answer the question “where?” I hit this range”, but do not answer the question “is it profitable “Do I”—these are different planes.
| Niche | Top 25% (below) | Median CPL | Weak (higher) |
|---|---|---|---|
B2B SaaS (demo-request) CPL = demo-request / trial | 2,800 ₽ | 6,000 ₽ | 14,000 ₽ |
B2B services (agencies, integrators) CPL = contact request | 2,000 ₽ | 5,000 ₽ | 12,000 ₽ |
Local services (beauty, auto, repair) CPL = call/form | 300 ₽ | 800 ₽ | 2,000 ₽ |
Education (courses, EdTech) CPL = webinar registration | 550 ₽ | 1,400 ₽ | 3,500 ₽ |
Medicine (clinics, telemedicine) CPL = application for admission | 450 ₽ | 1,200 ₽ | 3,000 ₽ |
Fintech (banking, insurance) CPL = questionnaire/application | 900 ₽ | 2,500 ₽ | 6,000 ₽ |
E-com (subscription/waitlist/lead magnet) CPL = email subscription | 80 ₽ | 200 ₽ | 500 ₽ |
Gaming (pre-registration, beta access) CPL = email / TG subscription | 65 ₽ | 160 ₽ | 400 ₽ |
B2B SaaS median CPL = 5500 ₽, e-com waitlist = 220 ₽. 25 times difference and it is not random. Four reasons:
- Qualification. Demo-request requires email from corporate domain, position, company size. Email subscription – one tick.
- Competition in the auction. In B2B SaaS at the same the request “CRM for an agency” hits 30 advertisers with budgets in the millions. There are three enthusiasts in the “pre-order indie game” niche.
- Audience. The decision on a B2B purchase is made by 5–7 people, you need to get to the right one. In e-com, the decision is made by one.
- Target action. “Sign up for a demo” is commitment to spend an hour of time. “Leave an email” - no.
Comparing CPL B2B SaaS with CPL e-com is comparing cost FAANG interviews with the cost of a TikTok like. The action is different the price is different, there is nothing to argue.
“Good CPL” is determined by unit economics, not industry
The main point that I repeat to clients is: not “CPL should be low”, and “CPL should fit into LTV × margin × lead-to-paid CR.” Benchmark by niche tells where you end up relative to others. Unit economics says whether you can even work on this CPL.
CPL ceiling formula:
max CPL = LTV × margin% × (lead-to-paid CR%)
Example. LTV of the client is 60,000 ₽ per year. Gross margin 40%. Conversion from a lead in the payer (lead-to-paid CR) 12%. Then:
max CPL = 60 000 × 0,40 × 0,12 = 2 880 ₽
If your actual CPL is 4,500 ₽, the channel is unprofitable. On paper the median for B2B SaaS = 5,500 ₽, you are “better than the market”, but you lose money from each paid lead. This is the main discrepancy between the “benchmark” and "its economy." You can calculate both numbers here: LTV/CAC calculator and unit economics calculator.
The reverse also works. If LTV is high (for example, enterprise SaaS with a check of 1.2 million per year), the CPL ceiling can be 15,000–25,000 ₽, and “above the market” is normal. Those who don’t count the ceiling are cutting expenses where it needs to be increased.
5 levers to reduce CPL
When the actual CPL breaks through the ceiling, there are five areas of work. I have arranged them in order of return on effort, from practice 2024–2026 years.
- Look-alike instead of broad interests. Loading audience of paying clients, the office is looking for similar ones. On my projects LAL 1–3% gives CPL 1.8–2.5 times lower than cold targeting by interests. Read more about look-alike.
- Landing page with a conversion of 4-6% instead of 1.5%. This is the second leverage by force. Without reworking the offer and structure, no bet will save. I discussed the boarding checklist in the article landing page 2026.
- Lead-magnet instead of “contact us”. Checklist, guide, calculator, benchmark report. Conversion rate is 3-4 times higher than the form “Leave a request.” What is lead-magnet and which ones will work in 2026.
- Retargeting for “form dropouts”. Segment: reached The forms were started to be filled out, but were not sent. CPL for this audience is 3–4 times lower than cold. Read more about retargeting.
- Bundle of retargeting + LAL. On LAL you catch volume, on in retargeting you squeeze out the warm ones. Removes the ceiling of one tank audience, at which CPL begins to grow after the first 30–40% of the buyout.
CPL vs CPC vs CPA - three similar metrics, three different meanings
Even agencies confuse metrics. A short table - what each one measures, and when which is useful?
- CPC — cost of one click. Measures quality creativity and the state of the auction. Useful at the start of a campaign and for creative tests. More details: CPC term.
- CPL — cost of application with contact. Measures the bunch “creative + landing page + offer.” Useful for lead-gen funnels where the deal is It closes not on the website, but through the salesperson.
- CPA — cost of an arbitrary target action: application, registration, adding to cart, purchase. CPL is private CPA case. More details: CPA term.
Rule: the deeper the metric along the funnel (CPC → CPL → CPA purchase) - the the closer to the economy and the longer the statistics accumulate. CPC is possible optimize for 50 clicks, CPL for 30–50 leads, CPA purchase for 15–20 sales. At small volumes, the lower metrics lie.
B2B CPL - why all market figures are lies
A separate painful conversation. In B2B application with contact ≠ qualified lead. Between them sits another filter - a salesperson or scoring that weeds out “pilot students”, “marketers” competitors”, “not our ICP”. What is called a lead in the office is MQL. What will really go into the deal is SQL.
On my B2B projects, the MQL → SQL conversion is around 25–35%. Then there is an actual CPL for SQL that is 3-4 times higher than the CPL for MQL from advertising office. When an agency shows “CPL 2,000 ₽ in B2B SaaS” is almost certainly considered an MQL. According to SQL it will be 6,000–8,000 ₽, and this is already close to the market median.
What to do. On reports, separate: “CPL by account” and “CPL by SQL”. Second is the only one that can be compared with LTV and considered a ceiling. Learn more about the B2B funnel design - lead-generation B2B 2026.
And one last thing. If no one considers SQL in a B2B project, you have no performance marketing, you have a “traffic department”. These are different functions, with different responsibilities and different results.