Every intake season, the same conversation about Google Ads campaign types happens. An institute has an intake starting in six weeks, a modest budget, and someone in the room says: “Let’s just run Performance Max — Google’s AI will figure it out.”
Six weeks later the report looks busy. Thousands of impressions, a low cost per conversion, a graph that goes up. But the admissions team is chasing people who don’t remember filling anything in, and the class starts under-enrolled.
The campaign type wasn’t wrong because Google failed. It was wrong because the campaign type was chosen before anyone asked the only question that matters: is there already demand for this course, or do we have to create it?
This article breaks down how each campaign type actually behaves in an education account, which one to start with depending on your course and your timeline, and the measurement mistake that quietly wastes more budget than any campaign setting.
Start with the question that decides everything
Google Ads campaign types are not a quality ladder where Performance Max is the “advanced” option and Search is the beginner one. They are tools for two completely different jobs.
- Capturing demand — someone already wants what you sell and is actively looking for it. Your job is to be there at that moment.
- Creating demand — nobody is searching yet. Your job is to put the course in front of the right person and make them want it.
A “Diploma in Human Resource Management” has some existing search demand. People type it. A short parenting course on toddler tantrums has almost none — mothers don’t wake up and search for that, but they will stop scrolling for it.
Same budget, same ad account, two opposite campaign types. Choose the tool after you have diagnosed the demand, not before.
The fastest way to diagnose it: open Keyword Planner and look at the monthly search volume for your course name and its two or three closest category terms, in every language your audience uses. If the total is meaningful, you have demand to capture. If it rounds to almost nothing, no amount of bidding strategy will create students out of a market that isn’t searching — you need a demand-creation campaign and a longer runway.
A quick map of Google Ads campaign types worth your budget
Before comparing them, it helps to be precise about what each one actually is.
Search — text ads on Google results pages, triggered by keywords. Highest intent, lowest volume, most control.
Performance Max — one campaign that serves across Search, Display, YouTube, Discover, Gmail and Maps, with Google’s automation deciding placement, audience and bid. Maximum reach, minimum control, heavily dependent on your data and creative.
Demand Gen — visual, feed-based ads on YouTube, Shorts, Discover and Gmail, targeted using Google’s audience signals. Built for creating interest rather than capturing it. Closer to how you’d think about a social campaign, but on Google’s inventory.
Video (YouTube) campaigns — useful when your objective is genuinely views or reach around an event, an open day, or a lecturer introduction series. For most intakes, Demand Gen covers this better because it optimises toward action.
Display — cheap impressions, weak intent. Occasionally useful for remarketing to people who visited your course page and didn’t enquire, and rarely worth much else in an education account.
For a typical student intake, the real decision is between the first three.
What Search campaigns are genuinely good at
Search is still the most reliable of the Google Ads campaign types for student intake, and it is the one most institutes underuse.
Its advantage is intent. Someone typing “HR diploma Sri Lanka” or “part time management course Colombo” is further down the decision path than anyone you interrupt on a social feed. Those leads close faster and need less follow-up, which matters enormously when your admissions team is two people on WhatsApp.
Its limitation is volume. In a niche vertical — a specific diploma, in a specific language, in a market of a few million potential students — the total monthly search volume may only be a few hundred queries. You can dominate that entire market and still not fill a batch of sixty.
This is where budgets get wasted in the other direction. Marketers see a Search campaign hitting its ceiling, panic, and start adding broad keywords like “diploma,” “online course,” or “jobs.” Now you are paying for clicks from people looking for a completely different qualification, at a completely different price point, in a completely different city. The spend graph finally looks healthy. The enrolments don’t move.
Practical rules for Search in an education account:
- Keep keywords tight — mostly phrase and exact match. Broad match belongs in a mature account with strong conversion data, not in a six-week intake push.
- Build a separate campaign for your own institute name so branded traffic never hides inside your prospecting numbers.
- Add negative keywords before launch, not after. “Free,” “jobs,” “vacancy,” “syllabus pdf,” “past papers,” “salary” and the names of qualifications you don’t offer will save you a meaningful share of your budget in week one.
- Put the things people actually decide on into the ad copy: duration, medium, schedule, recognition and how enrolment works. An ad that filters out the wrong people is doing its job.
- Whatever Search can’t spend is your signal that you need a demand-creation campaign — not looser keywords.
The keyword problem nobody warns you about
There is a specific trap in markets where the course is taught in one language and searched for in another.
A Sinhala-medium diploma is often searched for in English, sometimes in transliterated Sinhala typed on an English keyboard, and occasionally in Sinhala script. Three different query sets, three different competition levels, three different costs per click — and Keyword Planner will usually only show you healthy volume for one of them.
If you plan only around the English terms, you miss cheap traffic. If you plan only around volume, you will conclude the market is smaller than it is.
The same applies to how people name qualifications. Prospective students search for outcomes and job titles as often as course names: “HR job qualification,” “government job diploma,” “how to become an HR executive.” Those queries have intent — the person just doesn’t know your course exists yet. They sit in the gap between capturing and creating demand, and they are usually the cheapest volume in the account.
Check your search terms report weekly during an intake. The vocabulary your audience actually uses is almost never the vocabulary in the brochure.
Where Performance Max actually fits — and where it quietly burns money
Performance Max is not a beginner’s Search campaign. It is an automation layer that needs three things to work, and most institute accounts have none of them at launch.
1. Conversion volume to learn from. PMax optimises against your conversion data. With a handful of conversions a month, it is guessing. It needs a steady, consistent stream of conversions before its bidding becomes meaningfully better than a well-managed Search campaign.
2. Clean conversion signals. If you count every WhatsApp click as a conversion, PMax will find you people who love clicking WhatsApp buttons and never reply. The algorithm optimises for exactly what you tell it to value. Garbage signal in, expensive garbage out.
3. Strong creative assets. PMax spends much of its budget on Display and YouTube inventory. If your assets are a stretched logo and a text-heavy flyer photographed on a desk, that inventory will perform badly no matter how good the bidding is.
There is also a reporting trap that catches experienced marketers. PMax will happily serve your own brand searches — people already typing your institute’s name — and report those conversions as its own. Your cost per lead looks brilliant. In reality you paid for enrolments you were going to get for free, and the campaign that looks like your best performer is mostly taking credit.
Practical rule: treat PMax as a scaling tool for accounts that already have consistent conversion history and real creative. Exclude your brand terms so you can see what it is genuinely adding. Compare its enrolments against the period before you switched it on, not against its own dashboard. Run it as a second phase, never as the first thing you launch.
Why Demand Gen is the most underrated option for course marketing
Demand Gen runs on YouTube, Shorts, Discover and Gmail — visual, feed-based placements — but with Google’s audience data behind it rather than a social platform’s.
For education, that matters more than most people realise, because the biggest obstacle to enrolment is rarely awareness of the subject. It’s trust. A prospective student wants to know: is this real, who teaches it, what does the class actually look like, is the certificate recognised, do people finish it?
Those questions are answered by video and images, not by a text ad. And most institutes are already sitting on the assets without recognising them:
- Convocation photos and graduation footage — the single most persuasive proof that people finish and get certified
- A recorded class session showing the lecturer actually teaching, in the language the course is taught in
- Lecturer credentials delivered on camera in thirty seconds
- Short student testimonials shot on a phone, especially from students who work full time
- Screen recordings of the online class environment, which quietly answers “will I be able to follow this from home?”
Demand Gen also reaches the audience that would never search: the working adult who has been vaguely thinking about “doing something part time” for two years, and the parent who didn’t know a structured course on their exact problem existed.
Practical rule: use Demand Gen to create demand and feed the top of the funnel, then let Search catch the people it warmed up. If your branded search volume rises while Demand Gen runs, it is working — even if its own conversion column looks unimpressive on last-click.
The mistake that costs more than picking the wrong Google Ads campaign type
Most wasted budget in education accounts does not come from picking the wrong campaign type. It comes from measuring the wrong thing.
If your conversion action is “clicked the WhatsApp link,” you are optimising for curiosity. Every algorithm in the account — Search, PMax, Demand Gen alike — will chase the cheapest curiosity it can find, and it will find plenty.
Fix the signal before you fix the campaign:
- Define a real conversion. A completed enquiry form, or a WhatsApp conversation that reaches a qualifying reply — not the click that starts it.
- Import offline conversions where you can, so Google learns from people who actually enrolled rather than people who said hello and disappeared.
- Track cost per enrolment, not cost per lead, and compare it against your course fee. That single number tells you which campaign is profitable and which is just cheap.
- Give admissions a way to mark lead quality, even a simple spreadsheet column with three values: junk, real enquiry, enrolled. Without it you are optimising blind.
- Watch the gap between clicks on your contact link and conversations that actually get a reply. That ratio is the most honest performance metric in an education account, and it is usually the one nobody reports.
A mediocre campaign type with a clean conversion signal will beat a sophisticated one trained on noise, every single time.
Timing changes the answer too
How far you are from the intake date should influence your mix of Google Ads campaign types as much as the course itself.
More than eight weeks out. You have room to create demand. This is when Demand Gen earns its place, warming an audience that will search for you later. Starting it two weeks before the intake and judging it on immediate enrolments is the most common way institutes conclude “video doesn’t work for us.”
Four to eight weeks out. Search should be running at full weight, with remarketing to everyone who visited the course page. This is the highest-value window and the worst time to be experimenting with a new campaign type.
Under three weeks. Stop testing. Put budget behind the highest-intent traffic you have — brand, course-name searches and remarketing to warm visitors — and let admissions do the rest. Launching PMax in the final fortnight gives the algorithm no time to learn and takes budget away from what already converts.
A starting structure that works on a real budget
If you’re choosing between Google Ads campaign types without months of clean data, this sequence is hard to beat.
Phase 1 — Capture what already exists. A tight Search campaign on course-name and category keywords, plus a separate brand campaign so you can read your branded traffic honestly. Send traffic to a page that answers fee, duration, schedule, recognition and how to enrol before it asks for anything.
Phase 2 — Create demand. Once Search is saturated and can’t spend more without loosening targeting, add Demand Gen with your best real footage — classes, lecturers, graduation. Expect a longer path to conversion and judge it on assisted enrolments and branded search lift, not last-click.
Phase 3 — Scale. Only when conversions are consistent and the signal is clean, test Performance Max with brand terms excluded, and measure its incremental enrolments against what Phases 1 and 2 were already producing on their own.
On a small monthly budget, a reasonable starting split is roughly half to Search and remarketing, a third to Demand Gen, and the remainder held back for the final two weeks before the intake closes, when intent peaks and you want room to bid.
The “obvious” choice wastes budget because it skips straight to Phase 3 in an account that hasn’t earned it.
Where you send the traffic matters as much as your Google Ads campaign types
It’s worth saying plainly: no Google Ads campaign types compensate for a weak destination.
If your ads point at a Facebook page, the visitor has to scroll through unrelated posts to find the course. A homepage makes them hunt for it. If they point at a WhatsApp number with no prior information, you’ve moved your entire sales process onto a person answering the same five questions manually, sixty times a day.
A single course page that states the fee, the duration, the schedule, the medium, the awarding body and the next start date — then offers one clear way to enquire — will lift conversion rate more than most bidding changes ever will. It also improves ad relevance and lowers your cost per click, so the benefit compounds.
What about Meta ads?
For many institutes, Meta and Google aren’t competing options — they do different jobs, and the honest answer is that most intakes need both.
Meta is strong at reaching people who aren’t looking, at a low cost, with visual proof. Google Search is strong at catching people at the exact moment of intent. Where Google’s Demand Gen earns its place alongside Meta is the YouTube inventory and the fact that the same account, audience data and conversion signals sit next to your Search campaigns, which makes the handover from interest to enquiry easier to see.
If your budget only supports one platform this intake, choose based on the same diagnosis as before: existing search demand means start with Google Search; no search demand means start where you can create it.
Before you launch, check these five things
- Do you know whether people are searching for this course, or only for the outcome it leads to?
- Is your conversion action a real enquiry, or just a click on a contact link?
- Are your brand searches separated so you can measure what prospecting actually contributes?
- Do you have video and photo assets good enough for feed placements — and if not, what can you film this week?
- Can admissions tell you which leads enrolled, so the account learns from enrolments rather than noise?
If you can answer all five, the campaign type almost picks itself.
Frequently asked questions
Can I run Search and Performance Max at the same time?
You can, but expect PMax to absorb some of the traffic your Search campaign would have won, including brand searches. If you run both, exclude brand terms from PMax and watch total enrolments rather than each campaign’s own numbers.
How much budget does an intake campaign need?
Less than people assume for Search, more than they assume for demand creation. The useful benchmark isn’t a daily budget, it’s your cost per enrolment against your course fee. If a seat is worth far more than what you spend to fill it, the constraint is your available inventory of intent, not your budget.
Why do my leads never reply after enquiring?
Usually one of three reasons: the ad promised something the course doesn’t deliver, the conversion action was too easy so it collected curiosity rather than intent, or the follow-up came hours later. Response speed on WhatsApp does more for conversion rate than most account settings.
Is Demand Gen worth it for a small institute?
Yes, provided you have real footage and a runway of several weeks. Without either, the budget is better spent on Search and remarketing.
Running an intake soon and unsure which structure fits your budget? Share your course, your intake date and what you’re planning to spend, and we’ll map out the phase you should actually start from.