The question comes back the moment your ad budget gets serious: do we put this on Meta or on Google? That does not depend on which platform is better, but on one question per vacancy: does search volume already exist for the profile you need to fill? Below you will find how the two channels differ mechanically, when each one wins, and a decision framework you can apply to your own vacancy portfolio.
The mechanical difference: creating demand versus capturing it
Meta is interruption. You buy the attention of people who are not looking for anything at that moment. You decide who sees the ad — location, age, behaviour, or a lookalike of your own candidate database — and interrupt their feed with an offer. You create the demand yourself.
Google Search is the opposite. Someone types ‘welder job Eindhoven’ and you buy a position in a moment that already exists. You do not create demand, you harvest it.
Meta scales with budget as long as your audience is large enough: more money means more reach. Google Search is capped by search volume. You cannot buy more impressions than there are searches. An agency that misses this raises its Search budget and only watches cost per click go up.
When Google Search wins
Search wins when there is a demonstrable stream of searches that you only need to capture.
- Roles with genuine search volume: warehouse staff, drivers, administrative roles, call centre, cleaning, and almost every entry-level job.
- People searching right now: a contract ending, a recent move, a job just lost. They are further along in their decision than anyone you interrupt on Meta.
- Your own brand name: candidates typing in your agency name. Without a brand campaign, a job board outranks your own site.
- Vacancies with a hard deadline, where you have no weeks to warm up an audience.
The downside: the same terms are bid on by the large job boards and your competitors. Clicks cost more than on Meta, but convert better, because the candidate was already searching.
When Meta wins
Meta wins the moment search volume is missing, which is the case for most scarce profiles. A CNC machinist, service engineer or experienced production worker on shifts already has a job and will not type anything into Google this month. There is no demand to harvest, so you have to create it.
- Passive candidates who are open to something better but are not looking.
- Volume requests: twenty production workers never come out of regional search volume.
- Propositions you have to show: shift allowances, travel compensation, the machines, the team. Image and video do that; a text ad does not.
- New regions or a new client site where your name means nothing yet.
- Building a talent pool for recurring requests instead of starting from zero for every vacancy.
Volume role or one hard-to-fill vacancy
A second axis often gets skipped: how often do you place this profile? Meta's algorithm optimises on conversion data. If a campaign gets a handful of applications per week, the targeting sharpens itself. Run a campaign for a single senior role you fill once a quarter and that learning never starts: you pay for reach while the algorithm keeps guessing.
Rule of thumb: recurring volume profiles belong on Meta, while one-off, high-value vacancies are better served by Search, LinkedIn and targeted sourcing — channels that can reach one person without statistics being involved. In technical recruitment a combination usually works; how to build one is covered in the article on campaigns for technical staff.
Google is more than Search
The comparison is almost always made as if Google were only Search. Display, YouTube and Demand Gen are in fact interruption channels, exactly like Meta. They do not compete with your search campaign but with your Meta campaign, usually with weaker targeting signals for recruitment. YouTube is the exception worth attention: video works well for employer image and is easy to steer regionally.
Google for Jobs: the channel most agencies leave on the table
Above the ads, Google shows a job widget. You do not get in there with budget, but with structured data. Put JobPosting markup on your own vacancy pages — at minimum job title, location, datePosted, validThrough, hiringOrganization, employmentType and a real description — and Google shows your vacancy for free.
When this fails the cause is almost always technical: vacancies sit in an ATS iframe, on a vendor subdomain, or only on job boards. Google then indexes the job board's version. Three things to check: are the vacancies indexable on your own domain, is the markup valid in the Rich Results Test, and are expired vacancies marked as expired? That last one is structurally forgotten.
Budget: do not blindly split a small budget in two
Both channels need a minimum before they learn anything. Meta needs conversions per ad set to exit the learning phase; Google's bid strategies need clicks and conversions. A budget you spread thin hits that minimum on neither. Yet ‘everything on one channel’ is rarely the answer either. The way out is not fifty-fifty, it is sequential.
Worked example. Say: a budget of €1,500 per month for one profile. Assumptions for Meta: CPM €9, CTR 1.2%, landing page conversion 6%. Assumptions for Search: 250 relevant searches per month in your region, you capture 60% of them, CPC €1.80, conversion 10%.
- Meta with €750: €750 ÷ €9 = 83,333 impressions. Of those, 1.2% = 1,000 clicks, of those 6% = 60 applications: €12.50 per application.
- Search with the same amount: 60% of 250 searches = 150 clicks × €1.80 = €270 spent, not €750. Of those, 10% = 15 applications: €18 per application.
- That leaves €480 of Search budget you cannot spend without broadening into search terms that have nothing to do with the vacancy.
The lesson is not that Meta is cheaper — those numbers are assumptions, not benchmarks. The lesson is the sequence: fill Search until the search volume runs out, then put the rest on Meta. Search is capped, Meta is not. Do it the other way around and you pay Meta for reach while cheap, high-intent clicks sit untouched. How to calculate those costs is covered in the article on cost per candidate with Meta Ads.
Creative and measurement differ fundamentally
Assets
For Meta you produce image and video: vertical, hook within two seconds, three to five variants per audience and replacement as soon as frequency climbs. The message has to work without context, because the viewer was not looking for anything. For Search you produce text and manage search terms: headlines, sitelinks and above all a tight negative keyword list, or you end up paying for ‘work from home no experience’.
What both share: they send traffic to the same place. An ad cannot rescue a weak destination, so the vacancy landing page is what you invest in first.
Attribution
Google reports on last click and therefore takes credit for applications Meta generated: someone sees your ad, searches your agency name three days later and applies. Meta counts view-through conversions and claims conversions that would have happened anyway. Both are partly right and both mislead you at the same time.
Practically: steer on the total, not on the platform dashboards. Look monthly at qualified candidates and cost per placement across all channels combined, switch one channel off for two weeks and watch what happens to that total, and put a ‘how did you hear about us?’ question in your application form.
A decision framework for your vacancy portfolio
Go through your open vacancies and answer these five questions for each one.
- Is there search volume? Check Keyword Planner for job title plus region. If yes: Search first, until the volume runs out.
- How often do I place this profile? Several times a month: Meta gets enough data. A few times a year: sourcing, Search and LinkedIn.
- Does this audience search at all? People on permanent contracts working shifts do not. Interruption is then your only route.
- Is my name known in this region? If not, Meta builds the awareness that makes your Search campaign cheaper afterwards.
- Are my landing page, structured job data and follow-up within the hour in place? If not, fix that first. No channel compensates for it.
Conclusion
Meta or Google is not a matter of belief but a calculation per vacancy. If the demand already exists you harvest it on Search. If it does not, you create it on Meta. Set up Google for Jobs regardless, because it costs no media budget. And agencies on a modest budget are better off with a clear sequence than with a neat split.
Want to know what that sequence looks like for your portfolio? Schedule a call and we will go through your open vacancies: per profile we determine which channel does the work and what budget is realistic for it.
