Image

Surveys, Combined Analysis, and a Smarter Sunny QA

5 November 2026

 

We've rolled out updates to help you understand why tests win, read overlapping tests together, and catch setup mistakes before launch. Here's what's new:

Surveys: Hear Why a Variant Wins Attach an on-site survey to any test and read the answers per variant, right alongside your metrics. Build yours with ready-made question types, preview it on desktop and phone, and preview it on your own site. Learn More

 

Combined Analysis: Read Overlapping Tests as One Running 2–3 experiments on the same page? Combine them and Mida reads every combination as a multivariate test, so you see which mix actually wins. Mida suggests combinations when there's enough traffic to read them. Learn More

Sunny QA: Catch Setup Mistakes Before Launch Sunny now checks each setup step as you go, previews every variant, and applies the fixes you approve in one click. It also warns when your page targeting would drop visitors arriving with UTM parameters. After you publish, it can re-check a live test every 3, 7 or 30 days. The results page's Debug button is now called Run QA.

 

Ask the Agent from your dashboard — Your home screen now opens with a chat box for the Agent: type a question or pick a suggestion like "Create an A/B test" and it gets to work. The redesigned dashboard also puts your running experiments up front, only flags "Winner ready" once a test has at least 7 days and 25 conversions in every variant, and links straight to what's new in Mida.

Contributor role and launch approvals — Contributors can build tests and request approval to launch. There's a new Approvals tab, and Agent-proposed changes come with reasons and a risk rating.

Team activity feed and audit log — Owners and admins can see who changed what, and when.

Goal segments — Split any result by whether visitors completed a goal after entering the test, in one click.

Measured metrics — Measure a secondary metric as times per visitor or value per visitor, and count only each visitor's first or last purchase for revenue.

Pageview (Multiple URLs) goal — One goal that covers a list of URLs.

Smarter winner calls — There's a stricter bar for calling a winner before day 14, a new "Strong winner" label, and tests with 3 or more variants now judge each variant against the original.

Bounce rate, clarified — Bounce rate is now measured from when a visitor enters the test, over a 30-minute window, so it reflects your test rather than the whole visit.

 

Fixes That Protect Your Data

Chance to Beat Original was wrong on very large tests, tests with low conversion rates and tests with zero conversions. Fixed.

Frequentist confidence could show inverted in the experiments list. Fixed.

Purchases now count toward revenue goals used as secondary metrics.

Goal segments now catch every event name a goal can fire under.

GA4 audience sync is more reliable.

Install checks now confirm the key your tag actually runs with.

EU accounts no longer occasionally see screenshots from the US region.

 
Start New Experiment

Happy experimenting! 

From, the Mida Team ❤️

FacebookXInstagram LinkedIn

The email was sent to you@example.com. To no longer receive these emails, unsubscribe here.