Why Our Q3 Data Favors Live Raids Over Programmatic Ads

You probably have two slides in your quarterly deck: programmatic CPMs rising, and conversions flat. Those slides are why you are reading this. We ran a controlled test across September, isolating live streamer raids against matched programmatic banner buys, and the numbers are not close.

What we tested, and how

We matched budgets, audiences, and timing to remove excuses. Two $100,000 buys, same geo, same creative assets adapted for format, same one-week flight windows during peak autumn streaming hours, 6:00 p.m. to 11:00 p.m. ET. Tracking used UTM tags, unique promo codes in chat overlays, and a brand lift survey panel fielded pre and post exposure to measure purchase intent.

• Live raids: 150 coordinated, community-driven raids across mid- and top-tier streamers, organic chat activation, in-stream overlays, and a time-limited offer. Total reach: 2.4 million unique viewers. Average exposure length per viewer: 24 minutes.
• Programmatic banners: Standard display and native buys across gaming sites and social feeds, served during the same flight windows. Total served impressions: 16.7 million. Average exposure length per impression: 1.7 seconds.

We measured direct conversions, tracked promo redemptions, and ran brand lift surveys to capture purchase intent changes.

The cold, hard numbers

Raw outcomes tell the story.

• Cost per purchase

  • Programmatic banners: 100 purchases, cost per purchase $1,000. Clicks: 20,000. Click-through rate: 0.12 percent. Click-to-purchase conversion: 0.5 percent.
  • Live raids: 6,400 purchases, cost per purchase $15.63. Landing visits from raids: 80,000. Click-to-purchase conversion: 8 percent.

• Brand lift, measured as change in purchase intent in our survey panel

  • Programmatic: +1.4 percentage points.
  • Live raids: +8.6 percentage points.

• Engagement

  • Average watch time per viewer during a raid: 24 minutes, average chat messages per minute: 420, repeat exposure per viewer across the month: 4.2 times.
  • Banners: viewable seconds, essentially no conversation, negligible repeat attention during peak streams.

• Attribution signal

  • Promo code redemptions and UTM-tagged conversions accounted for 78 percent of raid-driven purchases, versus 62 percent of programmatic purchases traceable through UTMs.

These are not small multiples. Live raids delivered 64x more purchases for the same budget, and a roughly 6x greater lift in purchase intent.

Why live raids generate compounding ROI

Three mechanisms explain the math.

1. Attention, not impressions. People who watch a streamer are present for minutes, not milliseconds, and they are actively engaged. That attention converts. In our tests, a single 20+ minute exposure produced a purchase intent lift similar to 6 to 8 banner impressions.

2. Social proof in real time. Chat activity and in-stream endorsements made the brand feel recommended by peers. The immediate question-and-answer in chat removed friction, accelerating clicks and conversions.

3. Frequency with trust. The average viewer saw our brand 4.2 times across raids. Each additional exposure added about 1.6 percentage points of purchase intent up to the third exposure, creating a compounding effect you cannot buy in a single programmatic flight.

What this means for your marketing portfolio

If programmatic is delivering CPM-driven scale but no downstream revenue, raids offer a measurable, demand-driving complement. Here is the simple math for a marketing director: reallocating even 20 percent of a programmatic budget into coordinated live raids in our test increased incremental purchases by roughly 3,200 for the same spend, with stronger brand lift and far lower cost per acquisition.

We do not claim raids replace all programmatic activity. Banners still play roles in upper funnel reach and retargeting. However, when your KPI is purchase intent and measurable conversion, community-driven live activations create outsized, compounding ROI that programmatic display cannot replicate.

Caveats, and how we made the data clean

We controlled creatives, timing, and audience targeting. We used unique promo codes and UTMs to avoid double counting. Brand lift was fielded with a 1,200-person panel matched to our demo. We adjusted for viewability, bot traffic, and non-human impressions in the programmatic buys. The delta remained robust across sensitivity checks.

Final thought, and an invitation

Numbers like these change how you think about budget allocation. If you want the raw dataset, the breakdown by streamer tier, or a 30-minute walkthrough tailored to your brand and KPIs, reply here and we will get you the spreadsheet and a live session. I am happy to map what these results mean for your next quarter, without any marketing fluff.