Guide · Channel allocation
Don’t pick a winner on cost per lead or on a young cohort. Compare cost per opportunity on cohorts old enough to have matured, then decide what job each channel does.
Short answer
Neither wins by default; they do different jobs. Google captures buyers already searching for a solution, and LinkedIn reaches specific roles and companies before they search. Compare them on cost per opportunity for leads old enough to have progressed, not on cost per lead, which usually flatters Google.
Most comparisons stop at cost per click or cost per lead, and on those measures LinkedIn nearly always loses. The question a head of marketing actually has to answer is different: for the next unit of budget, which channel produces more qualified pipeline, and how sure can we be with the volumes we have?
Answering that needs three things most comparisons skip: identical stage definitions for both channels, enough time for each channel’s leads to progress, and an honest view of how few opportunities a month B2B accounts really produce.
Channel job cards
Write one card per channel and agree it with sales. A channel judged against the wrong job will always look like it is failing.
Google defines impression share as the impressions your ads received divided by the total they were eligible for (Google Ads Help). LinkedIn needs at least 300 member accounts to run an ad set (LinkedIn Help), but an audience that small will saturate quickly. For how Zephra runs each channel, see Google Ads and LinkedIn Ads.
Search leads are cheaper because intent is already there and clicks cost less. They also include people who searched a broad term, clicked once and filled the form without being a fit. LinkedIn leads cost more, but the targeting filters out many wrong-role and wrong-company submitters before they ever see the ad.
So the two channels drop out at different stages. Google often keeps more leads early and loses more at qualification. LinkedIn loses fewer at qualification and sometimes loses more later, when the buyer wasn’t ready. Only a comparison at a stage close to revenue captures both effects.
Cost per lead also punishes LinkedIn for changes that improve quality, such as adding qualifying questions to Lead Gen Forms. Move the comparison to cost per qualified lead at the least, and to cost per opportunity when you have enough history.
There is no reliable universal figure. Work it out instead: affordable cost per opportunity = average pipeline value per opportunity × win rate × the share of won revenue you are willing to spend on media. At US$40,000 per opportunity, a 25% win rate and 12%, that is US$40,000 × 0.25 × 0.12 = US$1,200.
Framework
Four steps that stop a young cohort from deciding your budget.
Qualified, meeting and opportunity mean the same thing whichever channel the lead came from, and the same people set them. See the funnel guide for a shared stage model.
Report opportunities against the month the lead arrived, not the month the opportunity was created. Otherwise this month’s spend gets credit for last quarter’s leads.
Wait until each cohort is at least as old as that channel’s median days from lead to opportunity. Calculate the median per channel; they often differ.
Move budget only if the gap holds for two or more mature cohorts and each channel has enough opportunities to read. With single-digit counts, use judgement and move in small steps.
| Cohort month | Channel | Spend | Leads | Qualified | Median days to opportunity | Cohort age today | Mature? | Opportunities | Cost per opportunity |
|---|---|---|---|---|---|---|---|---|---|
| Month | Google / LinkedIn | Yes / No | Spend ÷ opportunities |
Worked numbers
A US mid-market B2B software company spends US$15,000 on each channel in June. Here is the June cohort read twice.
| Channel | Leads | Cost per lead | Qualified | Cost per qualified lead | Opportunities by day 60 | Cost per opportunity, day 60 | Opportunities by day 120 | Cost per opportunity, day 120 |
|---|---|---|---|---|---|---|---|---|
| Google Search | 150 | US$100 | 50 | US$300 | 12 | US$1,250 | 15 | US$1,000 |
| 60 | US$250 | 30 | US$500 | 6 | US$2,500 | 12 | US$1,250 |
An Indian IT services firm selling to US mid-market buyers spends ₹8,00,000 on each channel. Google yields 20 opportunities (₹40,000 each) and LinkedIn 10 (₹80,000 each).
But LinkedIn’s opportunities average ₹36,00,000 against Google’s ₹12,00,000. Pipeline per rupee spent: Google ₹2,40,00,000 ÷ ₹8,00,000 = 30; LinkedIn ₹3,60,00,000 ÷ ₹8,00,000 = 45.
Formula: pipeline per unit of spend = opportunities × average opportunity value ÷ spend. Check win rates by channel before acting, since larger deals often close less often.
Short answer
Fund Google up to the demand it can capture efficiently for your core terms, then fund LinkedIn to reach your target roles and accounts at a frequency the audience can sustain. Shift budget only when mature cohorts show a persistent gap in cost per opportunity, not after one month.
Zephra manages budget allocation within and across Google, Meta and LinkedIn using lead quality and business outcomes from CRM data or a spreadsheet, with spend boundaries and approvals your team sets. See the product overview.
LinkedIn’s precision is also its ceiling. A tightly defined audience of a few tens of thousands of members can only absorb so much spend before the same people see the same ads repeatedly. The signs, in the order they usually appear:
When two of those appear together, adding budget is the wrong move. Refresh creative, widen roles or company size with care, or move the marginal budget back to Google. Changes to small audiences take longer to read, so leave room between them.
Meta usually produces the cheapest leads of the three and the widest spread of quality, because targeting by job and company is limited. It can work for B2B offers with a broad buyer, such as small-business software or training, and for retargeting people who already visited from Google or LinkedIn. Judge it with the same mature-cohort method. See Meta Ads for lead generation.
To size a LinkedIn test before running this comparison, work backwards from the qualified leads you need with the LinkedIn Ads cost calculator; the Conversions API guide covers sending those stages back to LinkedIn.
Questions, answered
There is no reliable universal figure; it depends on deal size, audience and your qualification rules. Work backwards instead: the cost per opportunity you can afford equals pipeline value per opportunity times win rate times the share of won revenue you will spend on media. Then compare LinkedIn and Google on the same stage definitions.
Lag. LinkedIn often reaches people before they are actively buying, so its leads can take longer to become opportunities. A cohort read before it matures understates LinkedIn. Compare once each cohort is older than the channel’s median days to opportunity.
Use it, but not alone. Last-click source tends to undercount LinkedIn, because people who first saw a LinkedIn ad often come back through a brand search. Add a self-reported “how did you hear about us” field and read the two together.
Only if the buyers are reachable as a defined audience and a single deal is worth enough to pay for expensive leads. If search demand for what you sell exists and isn’t yet captured, Google usually comes first.
Sometimes, for offers with a broad buyer and for retargeting. Expect cheaper leads with a wider spread of quality, and judge Meta on the same mature cost per opportunity as Google and LinkedIn.
Look monthly for saturation and tracking problems, and shift budget quarterly based on cohorts that have matured. Monthly reallocations on young cohorts mostly chase noise.
Your account
The free ad account assessment reviews Google, LinkedIn and Meta for wasted spend, structure, constraints and measurement gaps. No CRM needed, no campaign changes.
