← All articles
Benchmarks

What a healthy ROAS actually looks like in 2026, by vertical

Why a universal ROAS target is misleading, and how to construct a benchmark that reflects margin, repeat purchases, and growth stage.

June 28, 202610 min read

There is no universal healthy ROAS

A reported ROAS can only be judged against the economics of the business. Two advertisers with the same platform ROAS may have opposite outcomes because gross margin, fulfilment cost, repeat purchase rate, and lead quality are different.

Industry benchmarks are useful context, but they should never replace the account's break-even target and historical performance.

Build the benchmark in layers

Use the business break-even point as the minimum viable threshold. Then compare current performance with the account's own recent history, similar accounts in the organization, and an industry range drawn from a documented source.

  • E-commerce: account for gross margin, returns, shipping, and repeat purchase value.
  • Lead generation: replace surface ROAS with qualified-lead and closed-revenue economics.
  • SaaS: include trial-to-paid rate, payback period, churn, and customer lifetime value.
  • Local services: include answer rate, appointment rate, close rate, and job margin.
  • Finance and healthcare: evaluate lead quality and compliance constraints alongside acquisition cost.

Use ranges and state confidence

Benchmarks should show their source, sample period, currency treatment, and confidence. A range is usually more honest than a single target because account mix and attribution methods differ.

The best recommendation identifies the gap, tests the likely driver through CPM, CTR, conversion rate, and value per conversion, then recommends the action linked to the actual bottleneck.