PDP
by
Product page decisions for ecommerce teams: find which element is costing the sale on each page, then recommend the edits to images, copy, reviews, badges and bundles.
Product pages get built once at launch and then only touched when someone notices they look dated. The ones that convert badly usually do so for a specific, findable reason: a missing size reference, a first image that shows the product without context, three reviews when the category needs twelve, a returns line buried below the fold. Redesigning the whole page changes all of those at once and teaches you nothing. This skill finds which element is costing the sale on each page and recommends the edits, by element.
What It Does
- Compares each page against the pages that convert best in the same category.
- Reads where visitors stop scrolling and where they leave.
- Finds the question the page fails to answer, from search and support volume.
- Checks review count and recency against what the category needs to convert.
- Identifies the bundle the basket data already supports.
- Recommends the edits per element, ordered by what each is worth.
Use Cases
- Fix the one element costing a page conversions rather than redesigning it.
- Find the sizing question the page never answers.
- Prioritize review collection on the pages where it actually converts.
- Add a bundle the basket data already supports.
Example Session
This product page underperforms. What is wrong with it?
Four issues. The first two are worth more than the other two combined.
- Element | Issue | Evidence | Worth | Fix
- First image | No scale reference | 62% leave before scrolling | $31,000 per quarter | Add worn shot first
- Sizing | No fit guidance | 28% of support tickets on this SKU | $24,000 per quarter | Add fit note above fold
- Reviews | 4 reviews, category needs 12 | Converts 1.9% vs 3.1% peers | $9,000 per quarter | Request from buyers
- Returns policy | Below the fold | 11% scroll that far | $4,000 per quarter | Move to buy box
Sixty-two percent of visitors leave before scrolling, which makes the first image the only element most people see and the one with no sense of scale. That plus the fit question, which generates more than a quarter of this SKU's support tickets, is $55,000 a quarter. The review count matters and will take a month of requests to fix, so start it now and ship the two image changes this week.
Why This Beats Prompting It Yourself
A Page Redesign vs The Element Costing The Sale: The Hidden Risk
A generic prompt reviews the page as a whole and suggests improving the images, tightening the copy and adding social proof. That is a redesign brief, and after it ships you will know the page converts better without knowing which of the five changes did it. The next page gets the same undifferentiated treatment.
When You Still Need The Question The Page Never Answers
Support tickets and site search are a record of what customers could not find on the page, and almost no page review reads them. Twenty-eight percent of this SKU's tickets being about fit is the strongest signal on the page, and it is invisible if you are only looking at the page. The fix is one sentence above the fold.
Known Limitations
Scroll and exit data need session recording or event tracking, and without it the skill works from conversion comparisons and support volume and says which evidence is missing. Category peer comparison needs enough pages in the same category to be meaningful. It recommends the edits and does not change your storefront or generate imagery.
Install It
- Download the ZIP.
It is free and there is no account to create. - Unzip it into your agent's skills folder.
Claude Code reads ~/.claude/skills/, which is hidden by default: the command in the folder block opens it. Other agents scan their own directory, so drop the same folder there instead. - Ask your agent to use it.
Restart the agent if it was already running, then it picks the skill up with no config.
skills folder
Copy
~/.claude/skills/pdp/
SKILL.md
references/element-benchmarks.md
pages/
# one file per page, with the ranked edits
# macOS: create the folder and open it in Finder
mkdir -p ~/.claude/skills && open ~/.claude/skills
# Windows: paste in the Explorer address bar
%USERPROFILE%\.claude\skills\Frequently asked questions
How is this different from a page audit?
An audit lists everything that could be better. This sizes each element against what it is costing in revenue and orders them, which usually means two changes matter and the rest are cosmetic. It also reads support and search volume, which page audits do not.
Which agents does the skill run in?
Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision.
What does it need connected to work?
Page-level traffic and conversion, order data, support tickets tagged to products, and site search queries. Scroll depth is optional and improves the image recommendations noticeably.
Why read support tickets for a page problem?
Because a support ticket is a customer telling you what the page failed to explain, after they cared enough to ask. Most never ask, they just leave. A question appearing in a quarter of a SKU's tickets is almost always costing more sales than it generates tickets.
Does it write the copy?
It recommends the copy for short elements like a fit note or a returns line, and specifies what a new image needs to show rather than producing it. The image recommendations are the ones most likely to be worth the most and need a photographer.
What is included with the skill?
The skill itself, the element benchmark reference, the per-page recommendation format, and the category thresholds. It is free and security scanned. We re-publish the ZIP when the benchmarks change, so download it again if your rankings start shifting.
A product page edit is a margin decision at scale.
Polar ties page changes to net revenue and contribution margin across 45+ sources, so images, copy and badges are chosen on what the SKU earns.
<script type="application/ld+json">{"@context":"https://schema.org","@type":"FAQPage","mainEntity":[{"@type":"Question","name":"How is this different from a page audit?","acceptedAnswer":{"@type":"Answer","text":"An audit lists everything that could be better. This sizes each element against what it is costing in revenue and orders them, which usually means two changes matter and the rest are cosmetic. It also reads support and search volume, which page audits do not."}},{"@type":"Question","name":"Which agents does the skill run in?","acceptedAnswer":{"@type":"Answer","text":"Any agent that supports the open SKILL.md format: Claude Code, Cursor, Codex CLI, GitHub Copilot, Gemini CLI, Manus, Grok and others load it unmodified. The format is portable, the location is not. Each agent scans its own skills directory, so you drop the same folder into whichever one yours uses. There is no config file to edit and no API key to provision."}},{"@type":"Question","name":"What does it need connected to work?","acceptedAnswer":{"@type":"Answer","text":"Page-level traffic and conversion, order data, support tickets tagged to products, and site search queries. Scroll depth is optional and improves the image recommendations noticeably."}},{"@type":"Question","name":"Why read support tickets for a page problem?","acceptedAnswer":{"@type":"Answer","text":"Because a support ticket is a customer telling you what the page failed to explain, after they cared enough to ask. Most never ask, they just leave. A question appearing in a quarter of a SKU's tickets is almost always costing more sales than it generates tickets."}},{"@type":"Question","name":"Does it write the copy?","acceptedAnswer":{"@type":"Answer","text":"It recommends the copy for short elements like a fit note or a returns line, and specifies what a new image needs to show rather than producing it. The image recommendations are the ones most likely to be worth the most and need a photographer."}},{"@type":"Question","name":"What is included with the skill?","acceptedAnswer":{"@type":"Answer","text":"The skill itself, the element benchmark reference, the per-page recommendation format, and the category thresholds. It is free and security scanned. We re-publish the ZIP when the benchmarks change, so download it again if your rankings start shifting."}}]}</script>