Every few months a new Google Ads campaign type shows up in our clients' accounts with a badge that says "New" and a recommended budget that's suspiciously large. Demand Gen is the one we get asked about most right now — mostly because Google's own account reps in India have been pushing it hard in quarterly business reviews, and because it visually looks like the Instagram/YouTube feed ads that D2C founders already understand. We've now run Demand Gen across dozens of accounts inside AdsSarthi's unified dashboard — D2C fashion, F&B, skincare, and a handful of real estate developers testing it as a video-lead channel. This post is the honest version: what Demand Gen actually is, where it beats Performance Max and Search, where it quietly wastes budget, and the specific setup we use to make it work for Indian accounts.
What Google Demand Gen Actually Is
Demand Gen replaced Discovery ads in 2024 and expanded the inventory it can run on. It's a single campaign type that serves visually rich ads — image and short-form vertical video — across YouTube Shorts, the YouTube in-feed and in-stream surfaces, Gmail promotions tab, and the Google Discover feed. Unlike Search, there's no keyword targeting. Unlike Performance Max, it doesn't touch Search, Display text ads, or Maps. It's a closed-loop, creative-first channel that behaves a lot like a Meta feed placement, except it's Google's audience graph doing the targeting instead of Meta's. The reason it matters for India specifically: Discover and YouTube Shorts consumption in the Indian market is enormous, and a large share of that traffic is on data plans and screen habits that never touch a Google Search box for early-funnel browsing. Demand Gen is Google's answer to "how do we capture attention before someone searches," and for category-creating D2C products — a new skincare actives brand, a D2C mattress brand, a new-to-market snack — that pre-search attention capture is exactly the gap Search and Shopping can't fill.
Demand Gen vs Performance Max vs Search: Where It Actually Sits
We get this question from almost every new D2C account we onboard, so let's be precise about it rather than hand-wavy:
- Search captures existing intent — someone already typed the need. Highest intent, highest CPCs, smallest reachable audience.
- Performance Max is an automated umbrella across Search, Display, YouTube, Discover, Gmail and Maps, optimizing to a conversion goal with minimal manual control. It will use Demand Gen-style inventory internally, but you can't isolate creative performance or control placement mix.
- Demand Gen is manual-scoped to the visual/social surfaces only, gives you creative-level reporting, lookalike-style audience expansion, and — critically — lets you run true A/B creative tests the way you would on Meta.
In our accounts, the brands that see the best Demand Gen ROAS are the ones that already have PMax running and use Demand Gen as a controlled feeder: it builds the top-of-funnel remarketing pool and warms audiences that PMax and Search then convert more cheaply. Running Demand Gen in isolation, with no PMax or Search safety net underneath it, is where we've seen the weakest results — you're paying for attention with nowhere efficient to land it.
- D2C fashion/apparel: CPM ₹140–₹260, CTR 0.9%–1.6%
- D2C skincare/beauty: CPM ₹160–₹310, CTR 1.1%–2.0% (video creative consistently outperforms static)
- Real estate (project launch video): CPM ₹190–₹340, CTR 0.4%–0.8% — much lower engagement than D2C categories
- F&B/D2C snacking: CPM ₹110–₹200, CTR 1.3%–2.2%, the strongest category we manage on this format
Is Demand Gen Worth It for D2C Brands?
Where it wins
For D2C brands with a genuine creative production rhythm — new UGC or founder-led video every 10-14 days — Demand Gen is one of the better cost-per-acquisition channels we run for mid-funnel and warm-audience conversion. Three conditions make it work:
- You have a remarketing list worth targeting. Demand Gen's "Combined Audiences" and lookalike-style expansion perform far better when seeded with a real customer list, cart-abandoner list, or 90-day site-visitor pool via your Google tag. Cold-only Demand Gen with no first-party signal tends to underperform PMax on the same budget.
- Your video creative is genuinely short-form native — 6 to 15 seconds, vertical, hook in the first second, no repurposed 30-second TVC cut down badly. Ads that look like ads get skipped; ads that look like a Reel or a Short get watched.
- Your average order value can absorb a slightly higher blended CAC. Demand Gen tends to land 15%-25% more expensive per conversion than a well-optimized Search campaign in the same account, because it's capturing lower-intent attention. It's a volume and reach play, not a cheapest-CAC play.
Where it disappoints
We've pulled Demand Gen budget back to zero for a handful of accounts where the product category has low visual differentiation (commodity supplements, generic phone accessories) and the brand didn't have a real creative pipeline — reusing the same 2 static images for 6 weeks straight, which triggers ad fatigue on this format faster than almost any other Google surface. If you can't commit to a rotating creative calendar, Demand Gen budget is close to wasted; put it into Search or PMax instead.
Is Demand Gen Worth It for Real Estate Brands?
This is the more nuanced answer, and it's where we push back hardest on what Google reps typically pitch in India. Real estate teams get told Demand Gen is "video lead gen" — but the data across the developer and broker accounts we manage tells a different story. CTR on Demand Gen project-launch videos runs roughly half of what we see in D2C categories, and lead form completion rates from Demand Gen traffic are meaningfully lower quality than Search traffic on branded or high-intent keywords like "3BHK [locality] price" or "[project name] RERA".
What Demand Gen does work for in real estate:
- Project launch awareness in the first 2-3 weeks of a new inventory release, when you need broad reach in a target micro-market fast, before Search volume for the project name even exists yet.
- Retargeting site visitors who didn't fill a form — showing a walkthrough video or amenities reel to warm traffic converts noticeably better than cold Demand Gen.
- NRI and out-of-city buyer segments who are browsing YouTube and Discover heavily but aren't actively searching yet — Demand Gen catches intent 2-4 weeks before it shows up as Search volume in our attribution data.
What it does not replace: Search campaigns on high-intent locality and project keywords, and Performance Max feeding your CRM lead form. If a real estate team asks us "should we move budget from Search into Demand Gen," our answer is almost always no — layer it on top of Search, don't substitute it.
Budget Thresholds: When to Actually Turn This On
Google's minimum recommended daily budget guidance for Demand Gen is deliberately vague because the platform wants volume. From what we've seen work in Indian accounts, the practical floor is different by category:
- D2C brands: don't start Demand Gen below ₹1,500-₹2,000/day. Below that, the algorithm doesn't get enough conversion signal in a learning period to exit the "Limited" delivery status, and you'll burn budget on impressions with no optimization happening.
- Real estate: ₹2,500-₹4,000/day minimum for a project-launch push, run for a fixed 3-4 week window rather than always-on — real estate demand is lumpy and campaign-window-bound, not evergreen the way a D2C product is.
- Agencies managing multiple small accounts: if a client's total Google budget is under ₹25,000/month, we generally advise against Demand Gen entirely — the spend gets split too thin across Search, PMax and Demand Gen to let any one channel exit learning. Consolidate into Search + PMax first, add Demand Gen once the account crosses that threshold.
You can see live budget pacing and channel-split recommendations for your own accounts inside the AdsSarthi dashboard once your Google Ads account is connected — we flag when an account is spreading itself too thin across channel types before any one of them has enough signal to optimize.
Creative Requirements That Actually Move Performance
Demand Gen is unforgiving of lazy creative in a way Search and even PMax aren't, because it's competing directly against organic Shorts and Reels content in the same feed. The specifics that consistently move the needle across our accounts:
- Vertical 9:16 as the primary asset, not a horizontal video with black bars — Shorts placement penalizes non-native aspect ratios in delivery quality.
- Hook within the first 1-2 seconds — a price point, a bold claim, or a visual pattern interrupt. Anything that reads as a "commercial open" (logo intro, slow pan) gets skipped before the algorithm can even register engagement.
- Vernacular captions matter more here than on Search. Adding Hindi or regional-language burned-in captions to video creative — even for an English voiceover — measurably improves watch-through in tier-2/tier-3 targeted campaigns. This is one of the reasons we built a 13-language vernacular creative generator into AdsSarthi — most D2C teams simply don't have bandwidth to localize video captions per campaign, and it's one of the highest-leverage, lowest-effort wins on this specific channel.
- Rotate creative every 12-16 days minimum. We track frequency and fatigue signals closely on Demand Gen because it decays faster than Search creative — a static image asset that worked well in week one is often flat by week three.
Setting It Up: A Practical Walkthrough
Step 1: Audience seeding before launch
Before you enable Demand Gen, make sure your Google tag is capturing at least 90 days of site visitor data and, ideally, a customer match list uploaded from your CRM or Shopify customer export. Launching Demand Gen cold, with no first-party audience signal, is the single most common reason we see accounts write it off as "not working" within two weeks — the algorithm needs a seed to build from.
Step 2: Asset group structure
Build at least 3-4 distinct asset groups if you're testing across multiple products or offers, each with a minimum of 3 video assets and 3 image assets. Thin asset groups — one video, two images — rarely exit the learning phase cleanly.
Step 3: Conversion goal alignment
Point Demand Gen at the same primary conversion action as your PMax and Search campaigns (purchase or qualified lead), not a soft engagement goal like "video views," unless your explicit objective for that specific campaign is brand awareness. We've seen accounts accidentally optimize toward view-through metrics that look great on the surface but produce almost no downstream revenue.
Step 4: Give it a real learning window
Don't touch budgets or creative for the first 7-10 days. Demand Gen's learning period is genuinely sensitive to interruption — every time you edit an asset group or shift budget materially, the learning phase resets. This is one of the most common manual-management mistakes we catch during account audits, and it's part of why pacing automation matters — you can start with our free AI audit to see whether your existing Demand Gen campaigns are stuck in a repeated learning-reset loop.
The Honest Verdict
Demand Gen is worth turning on for most established D2C brands in India once they're spending above roughly ₹1,500/day on Google and have a real remarketing pool and a creative production rhythm to feed it. It is not a Search replacement, and treating it as one — especially for real estate lead gen — is where we see the most wasted budget. For real estate teams specifically, it earns its place as a project-launch awareness and retargeting layer, not a primary lead channel; keep Search carrying the high-intent load. If you're unsure whether your account has the underlying signal and creative cadence to make Demand Gen worthwhile, that's exactly the kind of channel-mix question our Growth and Agency plans are built to answer with real account data rather than a generic recommendation from a quarterly Google review.