“Should we try Demand Gen?” is a reasonable question. I would just ask one thing before opening the campaign builder: what are we trying to find out that the current Search account isn't telling us?
Maybe you want more inquiries from people already looking for your service. Maybe you have a useful offer that needs an introduction before someone would think to search for it. Those are different jobs. Giving both campaigns the same lead target doesn't make them the same test.
Here's how I would work through that choice, including a brief you can fill in before committing budget.
Start with the buying situation
Google describes Search campaigns as a way to reach people searching for your products or services. For the search-results part of that activity, the useful starting point is the query: what does this person appear to need, and does our offer answer it?
Demand Gen uses visual ads across surfaces such as YouTube, Discover, Gmail, Maps, and the Google Display Network. It can introduce an offer when someone isn't actively searching for your brand. Google supports image-only and video-only campaigns, so a Demand Gen test does not automatically require a new video production project.
I wouldn't turn that into “Search is for sales and Demand Gen is for awareness.” Demand Gen can pursue conversions too. The distinction is the situation in which you reach someone, and what your message needs to do at that moment.
| Your actual question | A useful starting point | What I would check first |
|---|---|---|
| Can we win more relevant inquiries from people looking for this service? | A focused Search test | Search terms, offer fit, landing page, and whether more relevant demand is available. |
| Can a visual explanation get a relevant audience interested in this offer? | A bounded Demand Gen test | Audience rationale, usable creative, the next step, and a way to assess inquiry quality. |
| We have conversions, but nobody knows whether the leads are useful. | Measurement and sales feedback before expansion | What triggers the conversion and what happens after the form is submitted. |
These are planning recommendations, not rules that make one campaign type the winner. Check actual delivery settings and reporting rather than assuming a campaign name tells you every place the ads ran.
Don't use a new campaign to avoid an unresolved problem
If Search is attracting the wrong inquiries, adding another source of traffic may give you more of the same problem to sort through. I would first check whether the search, ad, offer, and landing page are describing the same thing.
A form completion could mean a sales inquiry, a support request, or someone asking for something you don't sell. Until that distinction is visible, “more conversions” isn't a particularly useful instruction. The PPC audit checklist covers the broader account review.
On the other hand, a Search campaign doesn't need to be perfect before you test something else. If the offer is clear, measurement is usable, and you have a specific audience-and-message question, a separate test can be reasonable. I would protect the budget needed for the existing work and agree on how much uncertainty the new test can afford.
Write a different creative brief, not just a different campaign name
A Search ad can respond to a need the person has just expressed. A visual ad often has to establish why the offer deserves attention in the first place.
Hypothetical example, not a client result: imagine a business selling inventory-planning software. A Search test might address people looking for inventory-planning software with a product demonstration offer. A Demand Gen test might show an illustrated example of the reorder decisions the product helps organize, then invite people to see the same demonstration.
The second version needs enough context to make sense without a search query beside it. I would start with a couple of genuinely different explanations of the problem, not a stack of color variations. That gives the creative review a useful question: which explanation attracts relevant interest?
If you start with static images, brief static images. If you need video, assign production separately and confirm who is responsible. Campaign management and video production are different scopes of work.
Also record the channels you intend to test. Google's Demand Gen channel controls are set at ad-group level, with automatic or manual selection. Ad type and audience compatibility can limit where delivery is possible. I would check the previews and then the actual channel report, rather than assuming every asset can run everywhere.
Agree on what counts before comparing results
The first comparison should be business meaning, not whichever dashboard cost per conversion looks lower. Are both campaigns reporting the same event? Does that event represent a qualified inquiry, or just a completed form?
For video, also inspect the attribution path. Google includes engaged-view conversions in its conversion reporting when the qualifying engagement and conversion-window conditions are met. Those do not require an ad click. Google's Ad event type segment helps separate click-through from engaged-view conversion credit.
That credit can be useful. It is not, by itself, proof that the ad caused a new customer to appear. I would review the split alongside your lead records and follow-up outcomes, and document the conversion windows used. Don't quietly compare click-only results from one report with broader credit from another.
A side-by-side campaign report is also not a controlled experiment. Different audiences, timing, creative, and inventory can all affect what you see. Use an initial test to make a bounded next decision. If the business question is whether the spend caused additional outcomes, it needs an appropriate incrementality test, not a stronger claim attached to the same report.
A copyable brief for the next test
Fill this in before launch. Specific answers are more useful than a long campaign plan.
- Question: We want to learn whether [offer and message] attracts [defined buyer] in [buying situation].
- Campaign choice: We're using [Search or Demand Gen] because [reason tied to that question].
- Scope: Geography, intended audience, channels, existing-customer treatment, and exclusions to check.
- Creative and destination: The versions we will test, who supplies them, and the next action the landing page asks for.
- Measured event: The exact conversion action, what triggers it, and what it does not prove.
- Quality check: Who reviews inquiries, how they classify them, and when that feedback reaches the campaign owner.
- Budget and review: Approved test spend, monitoring owner, review date, and enough follow-up time to judge the chosen outcome.
- Decision: What evidence would justify continuing, revising the message, or stopping? What would leave the result inconclusive?
I would avoid a universal “run it for this many days” rule. The review needs to account for how quickly your business gets usable evidence and how long leads take to develop. A broken destination or incorrect tracking needs attention immediately. A small sample that hasn't matured needs a different response.
The practical next step is to complete the question, measured-event, and quality-check lines first. If those are still vague, a new campaign type isn't the missing ingredient.
If you want help reviewing the account and choosing the next test, see my Google Ads freelance management services. For the separate question of how to handle channel budget controls, read the channel prioritization guide.
Platform guidance checked November 10, 2026 against the Google Ads Help pages linked above. Availability and delivery depend on campaign settings, formats, and account eligibility. The test brief and hypothetical example are practitioner recommendations, not reported client results.