ANALYTICS & REPORTING

Landing page performance audit

Use when you have page-level data and need to know which pages to fix, keep, or retire.

landing-page-performance-audit.md

Download .md

You are auditing landing page performance from analytics data.

Page data (URL, sessions, entrances, bounce or engagement rate, conversions, conversion rate, revenue if available):
{{PAGE_DATA}}

Site goal: {{PRIMARY_GOAL}}
Period: {{PERIOD}}

Classify every page into one of: Fix (traffic, poor conversion), Feed (converts well, starved of traffic), Keep (working, leave alone), Merge (overlaps another page listed), Retire (no traffic, no conversions, no strategic role).

Output a table per class: URL | Sessions | Conversion rate | Value if applicable | Reason for classification | Next action | Effort.

Then:
- Rank the Fix pages by estimated conversions gained if their conversion rate reached the site median, and show that arithmetic.
- For Merge, name the specific pairs and which URL survives.
- For Retire, state what would be lost, including any links or rankings you cannot see from this data.

Rules:
- Do not classify a page with fewer than {{MIN_SESSIONS}} sessions into Fix or Retire. Put it in a separate "too little data" list.
- The median must be calculated from the data given, not assumed. State it.
- Do not recommend a redesign. Actions must be specific to a page element or the traffic feeding it.

Fill in before running

Replace each placeholder with your own detail. The more specific you are, the less the model invents.

{{PAGE_DATA}}
{{PRIMARY_GOAL}}
{{PERIOD}}
{{MIN_SESSIONS}}

Getting a better result

Include a traffic source split if you have one; a page can convert badly only because of one channel.
Check the Retire list against your backlinks before acting on it.
The Feed list is usually the fastest win and the one people forget to look at.

Questions about this prompt

When do I use this rather than reading the top pages report?

When you need a verdict per page rather than a ranking. The report tells you what gets traffic. This sorts every URL into Fix, Feed, Keep, Merge or Retire, which is a decision you can hand to someone. Reach for it when the page count is past what you would sift through by hand.

What do I need in front of me?

Page level rows with sessions, entrances, engagement or bounce, conversions and conversion rate, revenue where you have it, the site goal, the period, and a minimum session count below which nothing gets classified Fix or Retire. A traffic source split helps, since a page can convert badly only through one channel.

What comes back?

A table per class with reason, next action and effort, then Fix pages ranked by conversions gained if they reached the site median, with the arithmetic shown and the median calculated from your data. Read the Feed list first. Pages converting well on starved traffic are the cheapest win and the one everyone overlooks.

What is the mistake that costs me here?

Acting on the Retire list without checking backlinks and rankings, which are not in the data you pasted and which the prompt says it cannot see. Deleting a page carrying referring domains is not cheaply reversible. Treat the too little data list as pages to collect on rather than pages to ignore.