# Adding a Request Demo CTA Beside Start Free A/B Test

Page: https://alidemirbas.com.tr/en/lab/ab-test-library/tests/adding-a-request-demo-cta-beside-start-free-ab-test

## Hypothesis

If a secondary “Request a demo” link sits beside the free-start CTA, visitors who need a sales conversation may find a next step that fits them in the hero. Even if some free signups move to the demo route, the qualified opportunity rate across both paths could go up.

## When to run this test

Hero visitors heading to the contact page, demo questions arriving through support, or weak signups from enterprise traffic can mean that people who want a sales conversation find no path with a single CTA. With a second link in place, enterprise visitors should take the demo path while small teams keep signing up on their own. First make sure sales can answer demo requests at the same pace for the whole test; if response times shift midway, you cannot tie a change in opportunities to the CTA.

## Primary metric

**Qualified opportunity rate**: Count deduplicated qualified opportunities per assigned visitor using the same qualification rule and sales follow-up period.

## Other metrics

- **Free signup rate**: Measure whether the demo option reduces self-serve signups among assigned visitors.
- **Valid demo request rate**: Count submitted, valid demo requests rather than clicks alone.
- **Opportunity close rate**: Compare opportunities won after the same sales follow-up period.
- **Hero total action rate**: Count visitors who complete a free signup or demo request, deduplicating those who do both.

## What to check before and during the test

- **Second path**: Add one clearly secondary “Request a demo” link beside the existing free-start CTA.
- **Action mix**: Read each path's volume alongside total qualified opportunities.
- **Hierarchy**: Keep free signup primary and use one fixed location for the demo option.
- **Copy**: Keep both labels unchanged; “Talk to sales” would be a separate copy test.
- **Audience**: Review small-team and enterprise traffic separately.

## What to watch when reading the result

- Judge the change by the qualified opportunity rate across both paths; a rise in valid demo requests is not a success on its own. Check whether the free signup rate fell and compare the opportunity close rate over the same sales follow-up period; if the demo path pulls people who were ready for self-serve into a slower sales process, do not ship it.
- This hero, these labels and one fixed demo location define the scope of the result. Keep the hero headline unchanged during the test so any difference traces back to the demo link, and review small-team and enterprise traffic separately. Also make sure the two buttons are not crammed on top of each other on mobile.
- Compared with signups, qualified opportunities are rarer and take longer to appear. Do not settle an unclear difference before every cohort has finished the follow-up window. Plan the run time up front: divide the sample per variant from the [sample size calculator](https://alidemirbas.com.tr/en/calculators/sample-size-calculator), based on your current qualified opportunity rate, by daily visitors per arm, and add the follow-up window.

## Frequently Asked Questions

### Do more demo requests mean the test succeeded?

Not alone. A demo route may shift visitors away from self-service. Use total qualified opportunities per assigned visitor across both routes. Check free signups and close rates over equal sales follow-up periods.

### How should demo requests and opportunities be counted?

Count valid submitted requests rather than clicks. Deduplicate qualified opportunities per assigned visitor using one qualification rule and follow-up window. Count people using both routes once in the overall action rate.

### What should the sales team keep consistent?

Apply the same service standard and sales process. Shared team workload can affect both groups, and partial rollout may not represent full-rollout demand. Plan capacity separately and track response times and opportunities over equal windows.