Landing Page Optimization Agent
Tests page variants to lift conversion.
- Industry
- Hotel
- Department
- Marketing
- Team
- Demand Generation
- Kind
- Decision-maker
- Approver
- Director of Marketing
- Code
- MKT-06
It runs tests on the hotels' offer and landing pages, such as hero image, rate display, package layout and the booking widget's call to action, to lift the share of visitors who go on to book direct. It decides when a test has enough data, recommends shipping, extending or stopping it, and sets the traffic split in the experimentation tool once approved.
How it works
7 steps, and a person signs off before step 7.
- 01Looks it up
Finds the record the request is about in the experimentation tool: the test's variants, traffic split, visitors, bookings and booking value per variant from the web analytics..
- 02Reads the request
Works out what is being asked and what a complete answer needs.
- 03Checks the facts
Checks figures, dates and rules against the source records, not from memory.
- 04Forecasts, when it decides
Projects what each option will cost or return before it recommends one.
- 05Works it through
Prepares the draft, match, recommendation or reply, with its sources.
- 06Asks a person
Hands the work to Director of Marketing with its reasoning before anything goes out.
- 07Acts
Once approved, it acts in the experimentation tool: it sets the approved traffic split or ships the winning variant. and records what was done.
What it works with
Your knowledge, your systems, your people.
Knows
The testing rules: the conversion definition (completed direct booking), guardrail metrics such as average booking value and cancellation rate, minimum sample and duration, and which page elements carry legal text that may not be removed.
Reads from
the experimentation tool: the test's variants, traffic split, visitors, bookings and booking value per variant from the web analytics.
Acts in
the experimentation tool: it sets the approved traffic split or ships the winning variant.
Hands back
An object with decision_type (ship, extend, stop or start), winner (or null), lift_percent, p_value, guardrail_change, days_needed (or null) and reason.
Asks a person when
A variant would ship to all traffic, a test is stopped early, or a variant changes how price, taxes or cancellation terms are shown.
Settings
Your limits, set by your managers.
| Setting | Default | What it controls |
|---|---|---|
| Significance level | 0.05 | A result counts only below this p-value. |
| Least visitors per variant | 3000 | No decision before each variant reaches this. |
| Shortest test (days) | 14 | Covers weekday and weekend booking patterns. |
| Most traffic to a new variant (%) | 50 | A new variant never gets more than this share. |
| Least lift worth shipping (%) | 3 | Smaller lifts are not shipped. |
| Guardrail | average booking value | A variant that harms this is never shipped. |
Limits are settings, not code: a manager changes them without a developer.
Rules it always follows
- No decision before 3000 visitors per variant and 14 days.
- A variant ships only with lift of 3 percent or more at a p-value below 0.05 and no harm to average booking value.
- A new variant never gets more than 50 percent of traffic.
- Never test away the display of taxes and fees, cancellation terms or the total price where the market requires it.
- Stop a variant early only for a clear harm to bookings or the guardrail.
See the Landing Page Optimization Agent agent on your own work.
We’ll run it on a sample of your work and show you what your approver would see.