Compose

Compose is a web-based, pay-as-you-go platform for running conversion experiments and organizing a CRO program. It fits small and mid-sized businesses, agencies, and internal CRO teams that want structured planning alongside testing. Its usage-based model and broad stated limits are attractive, but teams requiring server-side experimentation need another platform.

Overview

Compose combines A/B testing with a collaborative workflow for deciding what to test next. The product is aimed at web teams rather than organizations seeking a cross-platform feature-testing system. Billing starts only when an experiment is launched, allowing teams to explore the workspace before committing to live usage.

Key features

Testing options include A/B/n, split URL, split theme, and split template experiments. Custom goals, audiences, triggers, and traffic allocation let teams tailor tests to particular conversion questions. A visual editor and audience targeting make the setup suitable for marketers and CRO specialists, while Shopify users can create no-code theme and template experiments through Shopify’s editor.

Compose’s planning layer is a meaningful differentiator for teams running a program rather than isolated tests. A shared hypothesis backlog, PIE prioritization, and an experimentation roadmap provide a common place to collect ideas, rank opportunities, and coordinate upcoming work. The product lists unlimited experiments, variants, collaborators, and projects, so those core workspace dimensions do not impose published caps.

Real-time reports and frequentist statistical calculations support ongoing interpretation after launch. GA4 and BigQuery integrations can connect experiment work with broader analytics workflows. Compose says AWS Lambda@Edge powers high-speed, flicker-free client-side testing, which is useful for storefront experiments where a visible delay could undermine results. However, server-side testing is unavailable, a decisive limitation for teams whose architecture or release process depends on backend experiments.

Pricing

The Early Adopter Plan costs nothing, with usage billed at $0.0012 per monthly tested user (MTU) at the end of each month for the prior month’s usage. It includes unlimited experiments and variants, unlimited collaborators and projects, and no minimum tested users. This is the practical starting point for small teams and agencies that want to pay in proportion to traffic rather than seats; costs rise with tested-user volume.

Compose has no traditional free trial, but users can explore the platform free until launching their first experiment. Charges begin when an experiment is initiated, and monthly usage is billed on the first day of the following month. A half-day technical session costs $400.00 USD once, paid up front, while a full-day session costs $800.00 USD once. These options suit teams buying implementation guidance rather than recurring support.

The Advanced Support Add-on uses a subscription fee with custom pricing. It adds a dedicated success manager, up to one hour of monthly meeting time, chat support, and continuous data storage. Support also includes a complimentary one-hour onboarding session, email assistance, and a help center with tutorial videos; paid advanced or technical support is available when those channels are insufficient.

Platforms

Compose supports the web. Its Shopify integration targets storefront themes and templates, and its visual editor keeps experiment creation within a browser workflow.

Who it's for

Choose Compose when a small or mid-sized business, agency, or internal CRO team needs experiment planning, client-side testing, and usage-based billing in one web tool. Shopify teams gain especially relevant no-code theme and template testing. It is a poor fit for products that need server-side tests or non-web application coverage.

Pros and cons

• Pros: unlimited experiments, variants, collaborators, and projects; structured backlog, PIE prioritization, and roadmap tools; Shopify theme and template testing without code; custom targeting and traffic controls; real-time reporting with GA4 and BigQuery integrations; billing begins only at experiment launch.
• Cons: no server-side testing; web-only platform; MTU-based charges can increase with traffic; AI hypothesis generation is marked as coming soon rather than a current capability.

Alternatives

Experimentation Platforms and A/B Testing Tools provide broader category shortlists.

Confidence by Spotify is a freemium, cross-platform option with a free tier offering 75,000 credits per month, unlimited feature flags and seats, and up to five surfaces. GrowthBook suits teams wanting open-source self-hosting, with unlimited experiments and traffic on its free self-hosted plan. Statsig offers a free developer tier with 2 million events per month, unlimited seats, and session replays. Bucketeer is free, open-source, self-hosted, and has no per-seat pricing.

Marpipe is a freemium web and extension product focused on catalog creative and feed management. Amplitude provides freemium analytics with a free 2-million-events-per-month tier. ABsmartly is a paid, broad-platform alternative with pricing by request and a 30-day professional trial. AB Tasty is a paid platform with a free trial and plans built around unlimited experiments, variations, and metrics.

Verdict

Compose is a strong choice for web-focused CRO teams that value a shared planning system, Shopify-friendly visual testing, and billing tied to tested users. Look elsewhere if server-side experiments, mobile coverage, or predictable seat-based pricing is essential.