Ecommerce ads

Typical $1-5M/month DTC paid budget allocation:

Meta: 40-55% of paid budget. The workhorse prospecting and retargeting channel.

TikTok: 15-25% of paid budget. Growing share each year as in-app conversion matures.

Google (Search + PMax + Shopping): 15-25% of paid budget. Bottom-funnel intent capture.

Klaviyo / Email / SMS: 3-8% of total spend, but disproportionate ROAS - often 30-50% of attributable revenue.

YouTube: 5-10% of paid budget. Higher for brand-equity DTC.

Pinterest / Snap / Reddit: 0-10% combined depending on category fit.

Influencer paid: 5-15% of paid budget when creator partnerships are part of the strategy.

Brand and category dictate variance significantly. Beauty DTC over-indexes on TikTok and influencer. Furniture and home DTC over-indexes on Pinterest. B2B-adjacent DTC over-indexes on Google and LinkedIn. Pure visual-product DTC over-indexes on Meta and Pinterest.

Three structural patterns consistently separate winning ecommerce ad strategies from generic ones in 2026.

First: catalog-feed-driven retargeting via dynamic product ads (DPA) and Google Shopping. Brands not running Advantage+ Catalog (Meta) and Google Shopping with proper feed setup are leaving meaningful retargeting performance on the table. DPA + Shopping should be 30-50% of total paid spend for most DTC. The lift over manually-managed retargeting is 2-4x ROAS.

Second: creator-style production for cold acquisition. Studio-style ads underperform creator-style by 50-80% on TikTok and 20-40% on Meta in 2026. Brands still producing primarily studio ads are structurally behind. The fix is operational - shifting creative production from agency studios to in-house creator workflows or AI-assisted production tools.

Third: server-side attribution via Triple Whale, Northbeam, Elevar, or similar. iOS 14+ degraded pixel-only attribution by 30-60%. Brands without server-side CAPI miss 20-40% of attributable conversions, which directly reduces optimizer signal quality and ROAS. Above $500k annual revenue this is no longer optional.

Fourth: integrated multi-channel attribution. Brands running each channel as an isolated silo (Meta team, TikTok team, Google team) underperform brands measuring true incremental contribution across channels. Above $5M revenue, multi-touch attribution becomes mandatory for honest channel-mix decisions.

Internal: facebook-ads-for-ecommerce for the Meta-specific playbook; ecommerce-video-ads for video-specific patterns; product-ads for catalog-driven product ad deep dive.

Pre-$100k annual revenue: Don't run paid ads. Pour budget into organic content, founder-led social, email list building, and 1-1 customer service. Paid ads at this scale rarely produce sustainable unit economics because attribution noise, creative-volume requirements, and minimum-spend thresholds all conspire against small accounts. Revisit when revenue clears $250k and you have data to optimize against.

$100k-500k annual revenue: Start with Meta retargeting and Google branded-search. Use the established demand you're already generating organically - cart-abandoners and brand-search are the warmest possible audiences and the cheapest possible conversions. Don't add cold prospecting until you've validated repeat-purchase rate and brand-search lift. Most failing accounts at this stage jump straight to cold prospecting and burn budget on unfit audiences.

$500k-5M annual revenue: Multi-channel mix becomes appropriate. Meta prospecting + retargeting + Klaviyo + Google branded should be the core stack carrying 80-90% of spend. Add TikTok when you have creative production capacity for vertical UGC - skip it if you don't, because TikTok with bad creative is just expensive Meta. Triple Whale or Elevar becomes mandatory at this stage because pixel-only attribution captures 40-60% of conversions and the optimizer needs the recovery.

$5M-50M annual revenue: Full multi-channel diversification. Meta, TikTok, Google PMax all running at scale. YouTube and Pinterest based on category fit. Influencer paid as a meaningful share of budget. Server-side attribution and multi-touch attribution are mandatory - decisions about channel mix and budget allocation at this scale cost six and seven figures of misallocated spend if the data is wrong. In-house paid media team becomes standard; agency-only setups start to underperform pure-in-house or hybrid models.

$50M+ annual revenue: Brand-equity layer becomes a larger share of budget across all channels. Direct-response weights down; share-of-voice and category-creator positioning weight up. Cross-channel attribution becomes a board-level discussion, not an ops one. Marketing mix modeling (MMM) re-enters the conversation alongside multi-touch attribution because MTA breaks down at scale and MMM gives the incremental-lift view needed for capital allocation. Most $50M+ DTC brands maintain both attribution layers in parallel.

Reels and TikTok now eat 50-60% of vertical-short ad delivery across paid social. Brands still primarily shipping square or 4:5 creative are structurally weak in delivery quality and ROAS. Migration to 9:16-native creative is the single biggest 'free' performance lift available to most ecommerce accounts in 2026, and the brands that haven't migrated are visibly behind the curve.

Advantage+ and Performance Max flattened the manual-optimization edge. Pre-2024 paid media teams earned their fees through manual audience targeting, bid management, and campaign structure. In 2026 the AI bidders outperform manual setups for most accounts past $500k revenue, which has pushed the value lever from media buying to creative production. The agencies and in-house teams that adapted to this shift are scaling; the ones still selling manual-optimization expertise are losing share.

TikTok Shop reshaped impulse-buy unit economics. Sub-$50 AOV products that fit TikTok demographics increasingly route through in-app TikTok checkout instead of Shopify checkout because the in-app flow removes the click-to-site friction that kills 40-50% of mobile-Web conversions. Brands not running TikTok Shop in fit categories are operationally disadvantaged versus competitors that adopted it early.

AI-generated creative shifted the cost curve. Tools like Shuttergen, Arcads, Captions, and Veed Fabric make it possible to produce 20-50 creative variants of an angle from a single brief at near-zero marginal cost. Brands that adopted AI-generated creative in 2024-2025 operate at 5-10x the testing velocity of brands that didn't. The gap compounds weekly and is unlikely to close - early adopters keep widening the lead.

Server-side attribution (Triple Whale, Northbeam, Hyros) became table stakes above $500k revenue. The pixel-only attribution model that worked pre-iOS-14 captures meaningfully less in 2026. Brands running without CAPI and third-party attribution layers are running with degraded optimizer signal - which means worse audience targeting, worse creative testing, and worse ROAS at every spend level.