Demand Gen campaigns: where they stop and PMax starts

What demand gen campaigns are

Search this term and the results split down the middle. Five of them describe a Google Ads campaign type. Two describe demand generation as a marketing strategy. One of them is a competing ad platform that never mentions the campaign type at all.

The campaign type is one line long. Demand Gen campaigns capture engagement and action across YouTube including Shorts, Discover, Gmail, Maps and the Google Display Network.

The strategy behind the name is a different thing. Demand generation identifies consumer needs, promotes the product and generates potential leads.

Both are correct about their own subject. A reader who arrives with a Google Ads account open needs the first, and the second is the subject of B2B demand generation.

The split matters more than a naming quibble. A budget threshold from the campaign type page is an operating number, and a funnel diagram from a strategy page is a way of thinking. Reading one as the other produces a plan with no threshold in it.

Google itself publishes on both sides of the line without saying so. Its help page describes the product, and its marketing page for the same product describes an outcome and adds that results will vary by advertiser.

Google publishes the boundary itself

The question people ask beside this term is where Demand Gen stops and Performance Max begins. Google answers it in the FAQ of its own help page, in one sentence, on a page most comparisons do not quote.

The budget rule, on the page that publishes it

1. 1The recommendation is a multiple of the target, not a flat minimum, so it moves with the campaign it is set for.
2. 2The same section names the state an underfunded campaign lands in, which is where a stalled Demand Gen campaign usually shows up.

Google Ads Help answers it in its own FAQ. Performance Max finds more converting customers across all of Google’s channels. Demand Gen gives more control over creative testing, audience targeting and placements.

That is a difference in what you hand over. Performance Max takes the placement decision and the creative assembly. Demand Gen leaves both with you and takes the bidding.

The handover reads the same from the other side, on the Performance Max help page. It says the campaign combines your assets into images and videos, bids with Smart Bidding, and funnels your budget automatically toward the highest performing inventory and placements.

In funnel terms it is blunter. An operator who runs both calls Performance Max a bottom of funnel campaign that presents itself as full funnel, and Demand Gen a top of funnel campaign.

The reading is the same either way. If you have creative worth testing and a placement opinion, the control earns its place. If you have neither, the automation is doing work you were not going to do.

What the auction does with either of them is unchanged, and how Google Ads works sits underneath both.

The surface list is the other half of the answer, and it is longer than most comparisons show. Demand Gen reaches YouTube feeds including Shorts, Discover, Gmail, Maps and the Display Network, which is a wider spread than a video campaign and a narrower one than Performance Max.

That spread is why the placement control matters. Gmail and the Display Network behave differently from a YouTube feed, and a campaign left on all surfaces is buying five inventories with one bid.

The Display Network sits on that list for a reason that changes the campaign type. Search Engine Journal reported in May 2026 that Google is retiring standalone Display campaigns and folding the inventory here.

A migration tool arrives in June 2026. New standalone Display campaigns close later in the year, and the rest move automatically through 2027.

A Display budget you run today is a Demand Gen budget on a schedule.

Set the four fields that decide the campaign

A Demand Gen campaign is configured on one screen, and four of its fields carry the whole decision. The rest of the setup follows from them.

A Demand Gen campaign settings screenWhich fields carry the decision and which follow

The field names and the options are Google's. No performance column appears here, because the numbers in those columns belong to an account, not to a campaign type.

Two of those fields have published numbers behind them. Google’s Lookalike segment documentation names narrow, balanced and broad as 2.5, 5 and 10 percent of the target location, and says the old 100-user minimum seed list no longer applies.

Read the rest of that page before you plan around those percentages. Google says the tiers will no longer function as a targeting constraint.

From March 2026 the reach level and the seed list act as signals, and the system may serve outside the threshold you picked. Search Engine Land reported the change in February.

The three percentages still describe the pool you nominate. They no longer describe who sees the ad.

That is the only page carrying a number for the audience field, so it is the one to open if you go looking. For custom segments the FAQ asks for 10 to 15 keywords carrying interest or purchase intent.

The creative field has published floors too: images at least 300 by 300 pixels in four aspect ratios, and video no shorter than five seconds. Those are the specs your assets have to clear before any of the rest matters.

1. Set the objective first

   The bid strategy list changes with it. Demand Gen supports Maximize Clicks, conversions and value based bidding, and the three learn against different things.

2. Write the target CPA before the budget

   The budget rule below is a multiple of it, so the order matters. A target chosen after the budget is a target chosen by the budget.

3. Apply the multiple

   Google says at least ten times the target CPA as a daily budget. A practitioner running these campaigns says twenty, or a hundred dollars a day, whichever is higher.

4. Choose the placements deliberately

   Placement control is the reason to pick this campaign type. Leaving all surfaces on hands back the advantage you selected it for.

5. Check conversion tracking before launch

   Google lists it as required. Without it the bidding has nothing to learn against and the budget rule has no target to multiply.

Three sources publish three budget rules

One number decides whether the campaign can learn, and the three sources that publish it run from ten times the target to twenty. The quantity is the same daily budget in each case, and each version is consistent inside its own reasoning.

The platform sits at the bottom of the range, and the two operators sit above it. A lower threshold lets more advertisers start; a higher one protects the person who has to explain a stalled campaign.

Neither interest makes any of the three wrong. It does mean a campaign funded at the lowest rule is running closer to the edge than that page suggests.

One more number is not advice at all. From 1 April 2026 the Google Ads API refuses a Demand Gen campaign budgeted under five dollars a day, returning a BUDGET_BELOW_DAILY_MINIMUM error.

Search Engine Land reported the enforcement when Google announced it on the Ads Developers Blog.

That five dollars answers a different question from the rules above: it is the least the platform will accept, not the least the campaign needs. The multiples are the second number, and at a fifty dollar target CPA they sit a hundred times higher.

A fourth rule turns up in a different shape. Start at twenty to thirty dollars a day, a flat figure that does not move with what a conversion is worth to you.

That is the distinction the chart above is built on. A multiple of target CPA scales with the business; a flat daily figure asks the same of a fifty dollar conversion and a five hundred dollar one.

The same operator adds two more thresholds from experience: two to four weeks before judging performance, and fifty conversions in thirty days before target ROAS bidding is worth using.

Those two thresholds are connected, and the connection is what the budget rule is protecting. Conversions arrive at roughly the daily budget divided by the target CPA, so the budget decides whether fifty of them fit inside thirty days.

That is the case for a multiple stated without either party’s interest in it. At ten times the target the threshold arrives in five days. At one times the target it takes fifty, and the window is thirty.

Funding your campaign at the level that reaches the threshold inside its window is what our Demand Gen management sets first.

Four demand gen campaign mistakes to check first

Each of these is common enough to have a name, and each one costs money in a way that only shows up weeks later, long after the decision itself.

The last one deserves its own figure. Google publishes three headline numbers for this campaign type, and Dennis Moons takes each one apart on the same page he publishes his benchmarks.

Three published claims, each with what it is measured against

1. 1The first claim names no baseline, so it cannot be checked against anything.
2. 2The second compares the campaign type against a product Google has retired.
3. 3The third describes larger budgets and longer runtimes, which improve any campaign.

The figures may all be right. None of the three is evidence a reader can check, and a budget defended with one of them is defended with somebody else's baseline.

A second page publishes CTR and CPC planning ranges you could plan against and gives no sample size either. Two pages in a row means the gap is the convention here, not one publisher’s oversight.

One version of that figure does hold up. Run the campaign against your own previous period, on your own account, and the uplift you measure is about your creative and your placements.

That test costs the learning period and the budget the rule above sets, which is the real price of finding out whether the 26 percent applies to you.

That table is one account and it says so. What it shows is worth carrying anyway: the two campaign types arrived at the same cost per acquisition through opposite economics, a cheap click with a low conversion rate against an expensive click with a high one.

It also shows why the view through column matters. The 180 view through conversions have no counterpart on the Search row, and a comparison that drops them is comparing one campaign with part of another.

Read it on the window it belongs to

A top of funnel campaign judged on a bottom of funnel window looks like a failure every time. The measurement problem here is a matching problem, not a tooling one.

Last click attribution gives this campaign type almost nothing, because the click it buys is rarely the click before the order. Reading it that way produces a decision to pause it, every time, whatever it did.

Report assisted conversions and the view through window beside the last click number. If your account cannot show them, that is the gap to close before the campaign is judged.

Compare the campaign against the period before it launched, not against a search campaign running beside it. Those two are buying different moments and the comparison flatters the wrong one.

Check the backdrop before you read the drop. Optmyzr tracked 21,425 Google Ads accounts across five quarters to Q1 2026 and counted 5.65 billion fewer impressions served, an 11 percent fall in auction volume.

Click-through rate rose over the same stretch, from 1.83 to 2.22 percent. A campaign losing impressions while its click rate holds is reading the platform, not itself.

Where the click is priced, and why a campaign type does not change that, is set out in how Google Ads works. What Performance Max does with the same auction is the subject of Performance Max.

The conversion count matters twice over. It gates target ROAS bidding at fifty in thirty days, and it is also what the bidding learns from, so a campaign short of it is being judged on a model that has not been trained.

That fifty is not only a practitioner rule. Google’s own Demand Gen FAQ recommends moving from Maximize Conversions to target CPA or target ROAS after the campaign has received at least 50 conversions.

The same page tells you what to do next. Set a target ROAS around 20 percent below your last 30-day figure, then move bids in steps of 10 percent either way.

Report the daily conversion rate beside the cost figures for the first month. It tells you whether the campaign is on track for the threshold long before the performance number means anything.