Rocket Science #01 infographic
To be added. The same infographic launches this edition on social.
Meta was cheaper per good-fit registrant on both offers
Meta brought in ICP-fit registrants for less than LinkedIn on both offers we tested: £49 vs £150 per ICP-fit webinar registrant, and £90 vs £170 for our breakfast event.
The common view is that Meta drives junk leads for B2B. With every signup checked for ICP fit the moment it lands, the data says otherwise. A strong offer, plus creative that calls out the right people, lets Meta do the targeting.
That doesn’t make LinkedIn the loser. It’s still most of our budget, and it does work this test can’t measure. The argument is about roles: use Meta to fill our own audience with good-fit people cheaply, and let LinkedIn do what it does best.
Why we ran it: rising LinkedIn costs and Meta’s algorithm updates
It started with two observations. LinkedIn was getting more expensive across the board, and Meta’s latest algorithm updates had made it more interesting for B2B. To test it properly, we needed to measure lead quality from the moment someone signs up.
We connected Clay to HubSpot so that every new contact is checked for ICP fit the moment they sign up, whatever channel they came from. A contact counts as a good fit when both of these are true:
The right person
A marketing leader, another senior leader or a founder.
The right business
The kind of company we do business with.
This does two jobs:
- A quick quality read on every channel. We don’t wait for pipeline to find out whether a channel is working. We can compare good-fit signups per pound within days.
- Better data for Meta. We feed the good-fit signal back to Meta as a conversion event. The more conversion data Meta gets, the better its delivery gets.
Why a good-fit signup and not a deal?
If we optimised further down the funnel, Meta wouldn’t get enough conversions to learn from. ICP fit at signup is the middle ground: specific enough to mean quality, frequent enough to train on.
Test 1: our October webinar
“Handed Next Year’s Revenue Target”
Meta delivered good-fit registrants at £49 each, against £150 on LinkedIn. This test leans towards Meta: we deliberately put more budget there this month, alongside our usual channels.
The setup was a Meta instant lead form, plus LinkedIn ads and organic channels driving to the webinar page on our website.
| Paid channel | Meta (instant lead form) | LinkedIn (website form) |
|---|---|---|
| Spend | £890 | £300 |
| Cost per signup | £15.08 | £37.50 |
| Good-fit rate | 31% | 25% |
| Cost per good-fit registrant | £49.44 | £150.00 |
Meta was cheaper per signup and had a higher fit rate. Signups cost 2.5x more on LinkedIn, and fewer of them were a good fit.
Our read on LinkedIn’s fit rate
LinkedIn’s 25% fit rate is unusually low for us. Across all new contacts from LinkedIn in the last 40 days, the fit rate was 58%. We’d put this test’s figure down to the small number of LinkedIn signups, and to some coming from the right companies but with job titles outside our ICP. That’s why the offer matters as much as the targeting.
Test 2: our September content breakfast
“The CMO’s Playbook for Turning Content into Pipeline”
For the breakfast event, Meta cost £90 per good-fit registrant and LinkedIn £170. This is the fairer test. Spend was closer, and breakfasts have always been LinkedIn’s territory for us. LinkedIn still had more budget.
Both channels drove traffic to the same event page and registration form on our website. Meta ran from 27 August to 29 September 2026.
| Paid channel | Meta | |
|---|---|---|
| Spend | £627 | £1,700 |
| Cost per signup | £19.59 | £170.00 |
| Good-fit rate | 22% | 100% |
| Cost per good-fit registrant | £89.57 | £170.00 |
The two channels did different jobs. Meta brought in volume cheaply, and fit had to be filtered afterwards. LinkedIn cost 8 to 9x more per signup, but every signup was a good fit.
Organic did a lot of the work
44% of the breakfast’s good-fit registrants came from organic LinkedIn, direct visits to the event page and search, with no ad cost. That’s the halo of the brand LinkedIn helps build.
What it means: each channel has its own job
Meta won on cost per good-fit registrant both times. The bigger finding is that Meta has a real role in direct-response lead gen for a B2B company.
This test only looked at one thing: how well each channel turns a strong offer into good-fit signups. On that measure Meta is valuable, and it needs two things to work:
- Creative that does the targeting. The ad has to call out who it’s for.
- An offer the audience wants. Both the webinar and the breakfast were strong offers, and the offer matters a lot.
LinkedIn does work this test doesn’t measure:
- Organic lift, including the organic registrants in both tests.
- Brand recognition that makes people more likely to click our Meta ads. Some of Meta’s performance probably rests on people already knowing Rocket SaaS.
- Keeping us in front of the right people over time.
- Best at
- Putting the right message in front of the right people and keeping us in their feed
- What it needs
- Budget, which is getting more expensive across the board
- Role in our mix
- Brand, reach and the lift that makes every other channel work better
Our audience comes back. Once someone subscribes to the newsletter and joins one webinar, they’re likely to join another or come to a breakfast. If Meta brings good-fit people into that audience cheaply, LinkedIn no longer has to carry every acquisition and can focus on what it does best. LinkedIn is still where most of our monthly budget goes, and its costs keep rising.
Why it matters commercially
Webinars and breakfasts are two of our most effective funnels, so every extra good-fit attendee is worth real pipeline.
of all our deals are influenced by webinars
in pipeline from each breakfast event, which converts at a good rate
So the growth lever is clear: get more good-fit people attending, as cost-effectively as possible. This test suggests Meta can be a big part of that.
How we measured it
These are two tests over about six weeks, so treat them as an early read that we’ll keep testing. Here’s how the numbers were built.
Who’s counted
Registrants whose contact was created in 2026, the period when Clay enrichment has been running. Older contacts have no enrichment, so they’d count as not-fit by default. We also removed an internal test contact.
Source
HubSpot original traffic source (first touch).
Spend
Platform spend for each campaign, in GBP.
Breakfast window
The breakfast registration form is shared across events, so we counted every submission from 19 August to 2 October 2026.
A stricter cut
A few of the breakfast’s paid registrants were created months before the campaign, so the ad may not be what brought them in. Counting only contacts created during the campaign, Meta comes out at £104.50 and LinkedIn at £188.89 per good-fit registrant. Meta stays ahead.
Watch the summary
Rocket Science #01 summary
A short talking head summary. YouTube embed to go here once it’s filmed.
FAQs
Does Meta work for B2B lead generation?
In our tests it did. Meta brought in ICP-fit registrants for less than LinkedIn on both offers: £49 vs £150 per good-fit webinar registrant, and £90 vs £170 for our breakfast event. It worked because both offers were strong and the creative called out who each event was for, which let Meta do the targeting.
How do you decide whether a new lead is a good fit?
We connected Clay to HubSpot, so every new contact is checked for ICP fit the moment they sign up, whatever channel they came from. A contact counts as a good fit when they’re the right person (a marketing leader, another senior leader or a founder) at the right business (the kind of company we do business with).
Why optimise Meta for good-fit signups instead of deals?
Optimising further down the funnel wouldn’t give Meta enough conversions to learn from. A good-fit signup is the middle ground: specific enough to mean quality, frequent enough to train on. We feed that signal back to Meta as a conversion event.
Should we move budget from LinkedIn to Meta?
Not wholesale. Meta was cheaper per good-fit registrant, but LinkedIn does work this test can’t measure: organic lift, brand recognition that probably helps our Meta ads perform, and keeping us in front of the right people over time. We use Meta to fill our own audience cheaply and let LinkedIn focus on reach and brand. LinkedIn is still where most of our budget goes.
Was it a fair comparison?
The webinar leaned towards Meta, because we put more budget there that month. The breakfast is the fairer test: spend was closer, it’s a format we’ve always run on LinkedIn, and LinkedIn still had more budget. Meta came out ahead on both, and stayed ahead when we counted only contacts created during the breakfast campaign.
What does Meta need to work for a B2B offer?
Two things: creative that calls out who the offer is for, and an offer the audience wants. Feeding good-fit signups back as a conversion event helps too, because the more conversion data Meta gets, the better its delivery gets.
