If you have a limited budget and are trying to decide between Google Demand Gen and Performance Max, the short answer is this: Performance Max is the right default when you already have conversion data and want volume across every Google surface at once, while Demand Gen is the better choice when you’re short on conversion history, chasing brand awareness, or you want actual control over who sees your ads. Running both well at the same time takes a bigger budget than most of the small businesses we work with have.
We get asked about this split almost every week now, usually by an owner who read a LinkedIn post calling Performance Max “a black box that burns budget” and wants to know if Demand Gen is the fix. It isn’t a straight swap. They’re built for different jobs, and picking the wrong one for your situation is a slower, quieter way to waste money than picking the wrong bidding strategy.
What is Google Demand Gen, actually?
Demand Gen is Google’s visual, top-to-mid-funnel campaign type. It runs your ads across YouTube (including Shorts), Discover, and Gmail — a narrower set of surfaces than Performance Max, but ones where the creative is genuinely visual: image carousels, short video, and native-feeling placements inside Discover’s feed.
The part that matters most for a small business: Demand Gen gives you real audience controls. You can build custom segments from your own customer lists, target lookalike-style audiences based on people who resemble your existing customers, or hand targeting over to Google’s optimization and let it expand from there. You choose. That’s different from Performance Max, where audience signals are more of a suggestion to the algorithm than a hard rule.
What is Performance Max, and how is it different?
Performance Max is the full-funnel, heavily automated option. One campaign can serve across Search, Shopping, Display, YouTube, Gmail, Maps, and Discover simultaneously, with Google’s systems deciding in real time where your budget goes based on what’s converting. We’ve written before about what we actually turn off in Performance Max once a client’s account is live, because the default settings are tuned for scale, not for a €1,500/month budget.
The trade-off is control. You feed Performance Max your assets, your conversion goals, and some audience signals, and then you mostly watch. That works well when the algorithm has good data to learn from — meaning real, clean conversion tracking and enough volume for it to find patterns. It works badly when an account is brand new, tracking is shaky, or the business is trying to build awareness rather than harvest existing demand.
Which one should a small business actually pick?
Ask yourself three questions before you open Google Ads.
Do you have conversion history Google can learn from?
Performance Max needs data to optimize against. If your account has fewer than roughly 30 conversions in the last 30 days, or you’re launching a new product line with zero history, Performance Max tends to spend its early weeks “exploring” — which in practice means testing placements and audiences somewhat randomly while your budget pays for the lesson. Demand Gen, with its manual audience options, gives you a way to target intelligently from day one instead of waiting for the algorithm to catch up.
Is the goal a sale this month, or a customer next quarter?
If you need leads or purchases now and have the tracking to prove it, Performance Max is still the faster route to volume — it’s built to chase conversions across every available surface. If the actual goal is getting a new audience to recognize your brand before they’re ready to buy, Demand Gen’s visual, top-of-funnel placements do that job better, because you’re not asking an efficiency-obsessed algorithm to also carry an awareness objective it wasn’t built to prioritize.
How much budget are you actually working with?
This is the one owners skip. Both campaign types need a genuine testing window — in our experience closer to four to six weeks before the data is worth reading — and running two automated, semi-opaque campaign types on a combined budget under roughly €2,000/month usually just makes it impossible to tell which one is doing the work. Below that threshold, we pick one, run it properly, and revisit the decision in six weeks rather than splitting a small budget two ways from the start.
What creative do each one actually need?
This is where a lot of accounts quietly fail before targeting even matters. Performance Max wants a full asset group: multiple headlines, descriptions, square and landscape images, a logo, and ideally at least one video — if you don’t supply video, Google will auto-generate one from your images and site content, and it usually looks exactly as generic as that sounds. Feeding Performance Max a thin asset group is one of the fastest ways to get stuck serving mostly on Display at a low quality bar.
Demand Gen leans harder on genuinely native-feeling creative. Vertical video built for YouTube Shorts, image carousels that don’t look like repurposed banner ads, and copy that reads like a feed post rather than an ad headline all perform noticeably better than assets clearly recycled from a Display campaign. If your team doesn’t have the capacity to produce a handful of short vertical videos, that alone is a reason to lean toward Performance Max first and revisit Demand Gen once you can brief creative properly for it.
Common questions we get about this decision
Can Demand Gen replace my Meta ads budget instead?
Not directly. Demand Gen and Meta’s ad products reach overlapping but different audiences with different targeting logic, and neither one is a drop-in substitute for the other. Treat the Demand Gen vs. Performance Max decision as internal to your Google Ads budget, and make the Google vs. Meta split as a separate call based on where your customers actually spend attention.
Does switching from Performance Max to Demand Gen reset my learning data?
Effectively, yes. They’re separate campaign types with separate optimization histories, so moving budget from one to the other means starting a new learning period rather than transferring momentum. That’s another reason not to flip back and forth every few weeks chasing whichever one had a better week — pick a direction and give it the full testing window.
A side-by-side comparison
| Factor | Performance Max | Demand Gen |
|---|---|---|
| Placements | Search, Shopping, Display, YouTube, Gmail, Maps, Discover | YouTube (incl. Shorts), Discover, Gmail |
| Targeting control | Low — audience signals guide, don’t restrict, the algorithm | Higher — custom segments, lookalike-style audiences, or full automation, your choice |
| Best for | Scaling conversions fast when data already exists | Brand awareness, new audiences, or accounts light on conversion history |
| Data needed to work well | Meaningful, clean conversion volume | Works with less history since you can target manually |
| Typical testing window | 4–6 weeks minimum | 4–6 weeks minimum |
When running both makes sense
Once a budget clears roughly €3,000–4,000/month and the account already has Performance Max humming along on strong conversion data, adding Demand Gen as a dedicated top-of-funnel layer can work well — you let Demand Gen build awareness and warm audiences that eventually feed Performance Max’s remarketing signals downstream. We’ve paired the two this way for clients moving from a single acquisition channel toward a real funnel, using the same Target ROAS logic we use to sanity-check Performance Max bidding to decide when Demand Gen traffic is mature enough to switch on tROAS itself.
The mistake we see most often
The most common mistake isn’t picking the wrong campaign type — it’s running both from month one on a budget too small for either to get a fair test, then blaming “Google Ads” when nothing converts. If you’re weighing this decision alongside other channels entirely, our comparison of Microsoft Ads against Google Ads covers the same underlying question: match the tool to the budget and the data you actually have, not to whichever one had the most convincing case study on LinkedIn this month.
The bottom line
Start with Performance Max if you have conversion data and need volume now. Start with Demand Gen if you’re building awareness, targeting a specific new audience, or your account doesn’t have enough history for automation to work with yet. Give whichever one you pick a real six-week window before you decide it isn’t working — and don’t run both until the budget can genuinely support two learning periods at once.