Landing page for a Demand Gen campaign: converting a click that was never searching

There's an invisible border inside Google Ads, and plenty of advertisers cross it without changing a word of their landing page. On one side, the search network: someone types a query, your ad answers it, the click arrives loaded with explicit intent. On the other, Demand Gen — the format that replaced Discovery campaigns — whose ads appear in YouTube feeds, Shorts, Discover and Gmail. There, nobody asked for anything: the person was scrolling a feed, your visual stopped them for two seconds, they clicked out of curiosity. That click doesn't convert with the same page. This guide extends our Google Ads landing page guide and our article on landing pages for Performance Max to the specific case of discovery traffic.

Demand Gen: a click that comes from no query at all

The format is quickly summarised: you supply visuals, videos, headlines, descriptions, a URL and a goal; Google runs those creatives across its discovery surfaces — YouTube feeds, Shorts, Discover, the Gmail Promotions tab — to audiences you describe and the algorithm expands. It's a visual, native format that deliberately resembles the content it sits among. Targeting options and interface details change regularly; the part that doesn't move governs your whole page design: none of those placements is a results page. No query upstream, so no intent to confirm, so nothing to “satisfy” in the Search sense.

Online advertising is perceived by default as an interruption: Chang-Hoan Cho and Hongsik John Cheon, in “Why do people avoid advertising on the Internet?”, published in 2004 in the Journal of Advertising (study on Google Scholar), show that ad avoidance rests on three pillars, the main one being perceived goal impediment: the more an ad gets in the way of what the person was doing, the more actively they push it aside. So the visitor arriving from a feed doesn't merely have weak intent — they consented to an interruption and expect something in return, immediately. If the page asks for something before giving anything, they go back to their feed.

Why a page built for Search converts badly in Demand Gen

The temptation is obvious: the Search page already exists, it converts, so Demand Gen gets plugged into it. It fails for a structural reason. A good Search page is efficient: it confirms the query in the headline, goes straight to the form, assumes the need is already formulated and that the visitor is only choosing a provider. It rests on an assumption — “you know what you want” — that discovery traffic never satisfies. Facing someone who tapped a visual in a feed, that efficiency turns into rudeness: you're asking for their phone number while they still don't know what this is about. The result isn't a poor conversion rate, it's a bounce rate.

The conclusion fits in one sentence: don't plug Demand Gen into your Search page — duplicate it and rewrite the top of it. Here's what to change.

The hero must continue the creative — the image as much as the sentence

Message match is the baseline rule of any advertising landing page: the ad's promise must reappear in the page headline. In Demand Gen it comes with a requirement people forget on Search — continuity has to be visual. The visitor doesn't remember a sentence, they remember an image: a photo, a face, a dominant colour. If your creative shows an open notebook on a wooden table and the page opens on an abstract blue illustration, the thread snaps, and with it the little context the visitor had in mind. So the hero must reuse the same visual or a direct variation, the same person if the creative features one, the same graphic register, and the promise in the same words. It's the principle governing the page of a YouTube Ads campaign, pushed one notch further: here the visitor didn't even watch a video before clicking.

A second reason demands that care. Native formats blend into the content surrounding them, to the point that their advertising nature isn't always identified. Bartosz Wojdynski and Nathaniel Evans, in “Going Native: Effects of Disclosure Position and Language on the Recognition and Evaluation of Online Native Advertising”, published in 2016 in the Journal of Advertising (study on Google Scholar), showed that a large share of participants failed to recognise as advertising a piece of content that was in fact labelled as such, with the position and wording of the disclosure clearly shifting that recognition rate. In other words: part of your clicks come from people who thought they were clicking editorial content. A page that switches abruptly into sales mode creates a rupture for them. The hero must honour the creative's editorial promise before showing its commercial colours.

Ask for little: entry-level offer and a two-step form

On traffic with no intent, conversion is first a matter of proportionality: what you ask must stay smaller than what the visitor just received. A guide for an email, a simulation for a postcode: the trade feels fair. A thirty-minute sales call for three seconds of curiosity: it feels absurd. Concretely:

• A low-commitment entry offer — guide, template, assessment, calculator, trial, event registration — rather than a purchase or a quote. The sale happens later, by email and retargeting, to people who now know you.
• A form cut to the bone: the email alone is usually enough. Every extra field is paid for in cash on cold traffic, and the phone number is the most off-putting of all.
• Or a two-step form if you must qualify: a first non-identifying step (a choice, a situation), contact details only afterwards — see our article on the multi-step form.
• A single call to action, repeated, phrased as a benefit (“Get the guide”) rather than an effort (“Fill in the form”).
• No wall before the content: no pop-up on arrival, no banner covering the first screen.

If even a short form feels too demanding for this traffic, we compare both approaches in our article on the Google Ads lead form versus a landing page. In most cases the dedicated page still wins: it's the only place where you can build the trust that's missing.

Move social proof up: reassure before you ask

On a Search page, testimonials and client logos traditionally come after the offer: the visitor knows what they want, and social proof breaks the tie. In Demand Gen it changes function — it no longer arbitrates, it authorises. It answers the only question someone discovering you asks: “are these people real?”. So it has to move much higher: in compact form on the first screen (a rating, a customer count, three logos, one named testimonial), then in expanded form right before the form. Our guide to social proof on a landing page details the formats that work — one verifiable, attributed proof being worth more than ten anonymous testimonials.

Mobile, mobile network, a few seconds: speed as an entry condition

YouTube feeds, Shorts, Discover, Gmail: these contexts are overwhelmingly mobile, often browsed off wifi. Your page won't be judged on a 27-inch screen wired to fibre, but on a phone with patchy mobile data, by someone with a feed to get back to. Fiona Nah, in “A study on tolerable waiting time: how long are Web users willing to wait?”, published in 2004 in Behaviour & Information Technology (study on Google Scholar), already established that tolerable waiting time for information retrieval sits around two seconds, with visual feedback during loading extending it noticeably. Two decades later, on traffic with no intent, that tolerance hasn't grown: the cost of abandoning is zero, and the feed is one thumb swipe away.

The consequences are well known: a page designed mobile first rather than merely “responsive”, a first screen readable without scrolling, a call to action reachable by thumb, compressed images, no non-essential third-party scripts, and a load time measured in real conditions rather than locally.

Tracking and attribution: conversions arrive later

Demand Gen's last trap isn't on the page, it's in how you read the results. On Search the loop is short: query, click, conversion, often in the same session. On discovery traffic it stretches — the visitor sees a creative, clicks, reads, leaves, comes back three days later through a brand search or a retargeting ad, and converts then. Judge the campaign on same-day conversions only and you'll cut a channel that was feeding the rest of the account. Four precautions apply.

1. Isolate this traffic with dedicated UTMs so it doesn't blend into Search in your reports — the full method is in our UTM parameters guide.
2. Check the conversion window declared in Google Ads: too short a window mechanically truncates conversions on a long-cycle channel and skews bidding.
3. Read post-click and view-through conversions separately: the latter are attributed far more generously and should never be naively added to the former.
4. Measure cleanly on the page side, with a conversion matching a real action, fired once, imported into Google Ads — see our guide to conversion tracking with GA4.

Add an overall reading to that: compare final cost per lead, email sequence and retargeting included, rather than the first-visit conversion rate. It's the only metric that does justice to a channel whose job is to create demand that didn't exist.

Demand Gen, YouTube Ads and Performance Max share the same underlying mechanics: traffic that asked for nothing, and a page that must build everything in a few seconds, on mobile, before asking for anything. LanderKit templates are built for that role — fast, mobile-first Next.js landing pages structured around a single goal with reassurance sections already in place: email capture with the Ebook / Infoproduct template, event registration with the Webinar template. Duplicate, match the hero to your creative, shorten the form.