Skai Q3 Report Highlights Full-Funnel Investment and Growing AI Adoption Ahead of Holiday Shopping

Skai, the agent-native omnichannel marketing platform, today released its Q3 2026 Digital Marketing Quarterly Trends Report, revealing growing investment in Amazon DSP and expanding adoption of AI-led campaigns as advertisers prepare for holiday shopping.

Based on aggregated, anonymized Skai platform data covering nearly 1 trillion impressions, 10 billion clicks, and $10 billion in advertising spend across five quarters, the report reveals how advertisers are adapting to changing media costs and growing AI adoption across channels.

Amazon DSP investment grew 34% year over year despite Prime Day moving out of Q3. Its average CPC fell 44% to $0.99, close to Sponsored Products at $0.94, giving advertisers reason to reassess how audience building and demand capture work together.

Customer results illustrate how advertisers are putting that approach into practice. Grain Group used Skai to connect streaming TV exposure with display retargeting for Cambio Roasters, building audiences from shoppers who had seen the brand’s CTV ads. Skai brought audience activation, pacing, and performance reporting together in one platform. Across its Amazon retargeting campaigns, revenue grew 85% period over period, ROAS increased 14%, and cost per conversion fell 13%.

October’s Prime Big Deal Days provides a further benchmark for holiday planning. Across Amazon advertising, ramp-up clicks cost 28% less than event-day clicks, while event-day conversion rates were 15% higher. DSP ramp-up clicks increased 76% year over year, and the format accounted for 27% of attributed event revenue on 24% of spend.

For Cyber 5, those differences put investment timing alongside channel and format selection. Advertisers can test creative and identify responsive audiences before peak shopping, then combine those findings with conversion performance, product availability and sales data to guide investment across the full window.

The quarterly findings also show why lower media costs need to be evaluated alongside business outcomes. Across Amazon advertising, CPC fell 16% year over year, but conversion rates declined 11%, leaving conversions up just 1%. With spending on Walmart and other retail media networks each growing 37%, advertisers also face more choices about where to invest and how to assess each retailer’s contribution.

AI-led execution is taking a greater share of advertising investment. In Skai’s data, AI Max’s share of Google spend rose from 7% in January to 16% in September, while adoption increased from 38% to 66% of accounts. Automatic upgrades contributed to that expansion, making it important for advertisers to understand how changes in campaign settings translate into business results.

TikTok’s share of paid social spend increased from 13% to 20% year over year. Across paid social, CPM rose 20%, while a 28% increase in click-through rate helped CPC fall 6%. These shifts add to the challenge of deciding where to invest and how to sustain efficiency as costs and engagement change.

“Where a shopper buys tells us only part of how they chose,” said Michelle Urwin, Chief Marketing Officer at Skai. “As discovery spreads across more channels and retailers, advertisers need connected intelligence to understand how each contributes to consideration, conversion, and growth. Skai brings that intelligence together so advertisers can coordinate investment and action across the full journey.”

Access the full report at skai.io/quarterly-trends-hub.

Methodology

Analysis is based on aggregated, anonymized Skai platform data drawn from nearly 1 trillion impressions, 10 billion clicks, and $10 billion in advertising spend across five quarters spanning multiple countries and industry categories. Unless otherwise noted, year-over-year comparisons include advertisers with 15 consecutive months of spending. Q3 comparisons reflect July–September 2026 against the same period in 2025. October’s Prime Big Deal Days results supplement the quarterly analysis. Third-party research is attributed separately.