YouTube advertising within Demand Gen campaigns is delivering stronger results when marketers stop treating it as a monolithic channel. The current playbook centers on splitting placements into separate ad groups, applying patient audience exclusion timelines, and treating creative testing as the primary optimization lever—not an afterthought. Here’s what practitioners are finding works in mid-2026.

Split Placements to Control Spend and Find Winners

The days of running a single combined Demand Gen campaign across all placements are over. Splitting campaigns by placement type—separating Display from YouTube entirely—reveals which channel actually drives performance for your account ^[1].

Within YouTube specifically, structure ad groups by placement: Shorts, In-stream, and In-feed as separate entities. This granular approach allows you to control spend at the placement level rather than letting Google’s algorithm decide where your budget goes ^[1].

One notable finding: In-feed placements (ads appearing in YouTube search results) may be redundant if you already have strong organic YouTube presence. Some accounts perform better on Display only, others on YouTube only—the only way to know is to test them in isolation ^[1].

Audience Exclusions Require Patience: The 60-90 Day Rule

Building separate audience segments—lookalikes versus interest-based—allows direct performance comparison. Review cost per conversion by segment and exclude underperformers, but only after collecting sufficient data ^[1].

For demographic exclusions, the current guidance is to wait 60-90 days before cutting age brackets that show zero conversions. Premature exclusions based on small sample sizes can eliminate potentially valuable audiences ^[1].

One practitioner example: an account excluded the 18-34 age bracket entirely based on historical non-conversion data, but this decision came after extended observation—not a snap judgment in the first few weeks ^[1].

Creative Testing Is the Primary Optimization Lever

Creative testing has emerged as the most critical factor in Demand Gen campaign success. The current approach is to test 3-5 different creative angles per ad group—not minor variations, but distinct messaging and offer angles ^[1].

A structural best practice: keep all creatives within one angle grouped in the same ad. This allows you to evaluate angle performance rather than individual asset performance. The warning here is clear—don’t put all creatives into one ad, or you lose the ability to quickly turn off underperforming angles ^[1].

Add the “Video Play Through” column to your reporting to analyze where viewers drop off. This reveals whether the problem is your hook or your offer. Example tests might include the same message delivered by different presenters as one variable, with a completely different offer as another angle ^[1].

Testing Timelines and Creative Fatigue Benchmarks

For Display placements, allow a minimum of 2,500 impressions before making any decisions. For YouTube placements, the bar is higher—allow 30 days minimum for testing, extending to 6-8 weeks if data remains insufficient ^[1].

Creative fatigue is real and predictable. Expect to create new creatives every 4-6 weeks minimum. Building this production cadence into your workflow prevents the performance decay that comes from running the same assets too long ^[1].

Measurement: Look for Halo Effects, Not Direct Attribution

Don’t judge Demand Gen on direct 30-day spend-to-conversion math the way you would Search or Shopping campaigns. The value often appears elsewhere ^[1].

Look for the halo effect: increased branded search traffic, organic traffic lifts, or improved CPC and cost-per-conversion in Search, Shopping, or Performance Max campaigns after running Demand Gen for 90 days. This requires patience and cross-channel reporting discipline ^[1].

Demand Gen should typically function as a secondary or tertiary campaign type in your account—supporting your primary conversion drivers rather than being measured in isolation ^[1].