Tilly vs Intelligems

Intelligems and Tilly both run A/B tests for Shopify stores, on different things. Intelligems is built around the numbers that decide margin: it tests prices, discounts, shipping rates and thresholds, checkout and post-purchase offers, as well as content, and reports revenue, profit and cost of goods on every experiment. Tilly's experiments agent tests the words and creative around the sale: email subject lines, content and send times in Klaviyo, ad angles in Meta, and copy and section order on a preview copy of your theme.

Choose Intelligems if your biggest open question is what to charge, what to offer or where to set free shipping, and you have the traffic to test it. Choose Tilly if your tests are about what to say and to whom, and you want the test ideas and variants written for you from what your customers worry about.

Intelligems' figures come from its Shopify App Store listing and pricing page, each cell linked and dated. Its pricing scales with sitewide order volume, so the figures shown are the published starting prices.

On its Shopify listing, Intelligems' Smart Content plan starts at $69 a month, Smart Merchandise at $112.10, and Unlimited at $349, all scaling with sitewide order volume, with a 14-day free trial.

Tilly charges per test rather than by order volume: 10 credits for an A/B test including writing the variants, from a Store plan of $49 a month for 500 credits after 100 free credits. A CRO audit is 20 credits and each written fix 10.

At low order volume the two cost a similar amount to start. As orders grow, Intelligems' price grows with them; Tilly's does not, but it also does not test prices or shipping at all.

This page is Tilly's, so Intelligems' strengths go first. On pricing experiments, it does what Tilly does not attempt. source, checked 25 September 2026

• Price, discount, shipping, checkout and post-purchase offer tests. Tilly does not test prices or shipping.
• Profit, not just conversion: revenue, profit and cost of goods side by side on every experiment.
• Personalisation by traffic source, device, Klaviyo segment or behaviour, so a winning variant can be rolled out to the audience it won for.
• A strong public record on Shopify: a 4.8 rating from 195 reviews when we checked.

Tilly writes the tests. Every other agent feeds the experiments backlog: the CRO agent finds a page that does not answer an objection, the ads agent sees one angle beating another, the email agent sees one email carrying a flow. Each idea arrives as a hypothesis tied to a persona and a worry, with the variants already written.

It tests beyond the storefront. Subject lines, content, offers and send times in Klaviyo, and creative, angle and hook in Meta, are tested the same way as page copy, from one place.

It calls winners carefully. A minimum sample before any call, a maximum run time, one primary metric chosen up front, and a plain-language result. If a test does not reach enough data, Tilly says so rather than picking a winner by chance.

Results teach the other agents. A finished test is written into your brand profile, so the next email, ad or page fix reads it.

It starts with free diagnosis. The store audit is free and the CRO audit is 20 credits, so a store can see what Tilly would change, and why, before any test is set up or any traffic is split.

These rarely replace each other. A store can keep Intelligems for price and shipping tests and use Tilly for email, ad and copy tests; they touch different parts of the business.

If you are moving page tests from Intelligems to Tilly, finish or stop running experiments first so no shopper sees a half-ended variant, and note the winners you rolled out, because Tilly will not know about them unless you tell it.

Run Tilly's free audit, then a CRO audit on your key product page. The fixes it writes arrive on a preview copy of your theme; the bigger ones go to the experiments agent as tests with the hypothesis already written.

Keep price tests with a tool built for them. Tilly will not change prices, and it should not be the reason you stop learning what your customers will pay.

Give each tool the tests it is best at for a quarter, then compare what you learned. A store that knows its best price and its best product-page copy is in a better place than one that ran twice as many tests on either alone.