> ## Documentation Index
> Fetch the complete documentation index at: https://docs.wisepim.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Turn impressions into clicks
> Your products rank on page one but nobody clicks them. Rewrite the snippets and copy that lose the click, then watch CTR move in the SEO report.
You are already on page one for hundreds of queries. Google shows your products tens of thousands of times a month, and almost nobody clicks. Nothing is broken and nothing needs to be built, the snippet in the search result simply is not doing its job. This is the cheapest traffic in your whole catalog, because you have already paid for the ranking.
Budget about 90 minutes for the first pass, plus a check-in four to six weeks later.
**Before you start:** Google Search Console must be connected and synced, or none of this data exists. Connect it from the app marketplace under [Integrations](/en/integrations/supported-platforms).
**What you'll get:** a short, ranked list of queries where you rank but do not convert the impression, rewritten titles, metas, and descriptions for those products, and a measurable CTR change in the SEO report.
## Find the clicks you are leaving behind
Start with the diagnosis, not the fix. The [SEO & Search](/en/analytics/seo) report already knows which of your pages rank without earning the click, so let it build your working list before you spend a single credit.
In the Analytics sidebar, open [SEO & Search](/en/analytics/seo) and stay on the performance view. It shows impressions, clicks, CTR, and average position together, which is the only place the gap between "seen" and "clicked" is visible.
Choose previous period, or year-over-year if you sell seasonal products. Without a comparison you cannot tell a genuine CTR problem from normal seasonal movement.
The **Quick wins** panel on the performance view lists high-impression, low-CTR queries, exactly the traffic you already rank for. This is your backlog, in priority order, for free.
A position between 4 and 10 with a CTR well below 3% is the classic quick-win shape. Check it against the benchmark table at the top of the report, which tells you what counts as good, average, and poor for e-commerce.
Filter for high impressions, sort by CTR ascending, and note the products behind those pages. Those products, not the queries, are what you select in the next section - every fix in this guide is applied per product.
## Rewrite what shoppers actually see
Two different things lose the click, and they need fixing in this order. The snippet in the search result decides whether anyone clicks at all; the description on the page decides whether that click was worth anything. Fix the snippet first, because it is the number you are trying to move.
Which enrichment you reach for depends on whether the product has real query history behind it:
| Pick this | When | Because |
| ----------------------------------------------- | ------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| [SEO Title & Meta Description](/en/ai/seo-meta) | The default for every product on your list | It writes only the meta fields, so your storefront copy is untouched and you are testing one change at a time. It needs no query history. |
| [Search Intent](/en/ai/search-intent) | Your highest-impression products | It reads that product's own Search Console queries from the last 90 days and rewrites the description to match them. Products with no query data are skipped. |
On the Products page, tick the products you identified, or apply a filter and select all matching results so you catch the ones on later pages too.
Click **Enrich with AI** in the toolbar, pick **SEO Title and Meta Description** from the enrichment type grid, and start the run. It writes to the meta fields only, which is why it is safe to run first and measure in isolation.
The **SEO** tab of [Data Quality settings](/en/settings/data-quality) holds the meta title and meta description character limits your products are scored against. Confirm they match what you want, otherwise good snippets will still score badly.
Same modal, **Search Intent** instead - no prompt configuration needed. Save it for products with the most impressions, because they are the ones with the query data that makes it work.
Review five to ten rewritten products against their original queries. If the copy reads like it was written for those searches, run the rest.
## Make the gain hold
A one-off rewrite decays. These steps stop you redoing the exercise every quarter: keywords make every *future* enrichment target the right terms automatically, and internal links get your ignored products crawled in the first place. Do keywords first, so anything you re-run afterwards already knows them.
Pick your project in the project switcher, open [Keywords](/en/settings/keywords), stay on the **Positive** tab, and click **Find opportunities**. It lists Search Console queries your project ranks for but does not yet target, with their clicks, impressions, and average position - proven demand, not guesses.
Open a keyword's row editor and give your best terms **High** priority so AI weights them more. Scope each one to a locale and, if it only belongs in snippets, to the **SEO meta** target field. A keyword applied everywhere pollutes markets and fields it was never meant for. The page autosaves, there is no Save button to hunt for.
Same page, other tab: terms you never want written back in. Tone and banned wording belong there too - [Teach WISEPIM your house style](/en/guides/teach-wisepim-your-house-style) covers that side of the list in depth.
The link equity view in the SEO report flags products with no inbound links. Run [Internal Linking](/en/ai/internal-linking) on them from the same **Enrich with AI** modal - it needs an existing description to work, so products without one are skipped. Its link-target pool is a snapshot taken at run time, so re-run it after you launch a new category.
Your winning SEO prompt belongs in [Enrich Prompts settings](/en/settings/enrich-defaults) as the project default, so every product added from now on gets the same treatment without anyone remembering to do it.
## How to tell it worked
CTR moves before revenue does, and neither moves this week. Check the SEO report first, and only then look for money.
| Signal | Where to look | When it moves |
| ---------------------------------------------------- | ---------------------------------------------------------------------------------- | -------------------------------------------------------------------------------------------------- |
| CTR on the reworked queries | Performance view in [SEO & Search](/en/analytics/seo), against the previous period | 4–6 weeks after the rewrite - the window [Search Intent](/en/ai/search-intent) itself recommends |
| A query crossing 3% CTR at the same average position | Same view, against the report's benchmark table | The report treats above 3% as good e-commerce CTR, under 1% as poor |
| Average position holds steady | Same view, same window | Higher CTR at an unchanged position is the clean win; a position drop means something else changed |
| The Quick wins list gets shorter | Quick wins panel | As reworked queries cross the CTR benchmark |
| Orphan products drop | Link equity view in the SEO report | After the Internal Linking run and the next recrawl |
| Net lift on the enriched products | [Enrichment Impact](/en/analytics/enrichment-impact) | 3–4 weeks of post-enrichment sales at the earliest |
Enrichment Impact compares 30 days before against 30 days after each enrichment event, so a check the week after a run tells you nothing. Rows marked low confidence need another 2–4 weeks of sales before they mean anything.
## When it doesn't work
Every number on this page comes from Google Search Console. If the report has no data, the integration is not connected or has not synced yet. Reconnect it under [Integrations](/en/integrations/supported-platforms) and give it a sync before you conclude you have no impressions.
Products that are very new, very niche, or never ranked have no queries in Search Console, so there is no intent signal to act on and they are skipped. Point Search Intent at products with existing organic presence, and use [SEO Title & Meta Description](/en/ai/seo-meta) for the rest - it has no such requirement.
That is the exact symptom this guide addresses: Google is showing your pages more and shoppers are choosing someone else. Rising impressions with flat clicks is a title and meta problem, not a ranking problem, so re-run the snippet rewrite rather than chasing new content. Falling impressions are the other story - that is a ranking drop.
A transactional rewrite on a product whose traffic is informational can move the copy away from what ranked it. Restore the original copy and re-run, or write a description prompt in [Prompt Lab](/en/essentials/prompt-lab) that finds the middle ground and apply it through the Descriptions enrichment instead of [Search Intent](/en/ai/search-intent).
It aims for a handful of contextual links per product and needs anchor points to place them, so a one- or two-sentence description gives it nowhere to go. Enrich those descriptions first, then re-run [Internal Linking](/en/ai/internal-linking). If the links it did add point at unrelated pages, check the product's category assignment - category context drives the matching.
Older, longer meta copy stays in place until you regenerate it. Re-run [SEO Title & Meta Description](/en/ai/seo-meta) on the affected products, and check the limits on the **SEO** tab of [Data Quality settings](/en/settings/data-quality) match what you want scored.
## Related
The report behind every number in this guide: performance, coverage, cannibalization, and links.
Rewrite descriptions from the real queries that found each product.
Make sure the rewritten copy sounds like your brand, not a template.
Keep the quick-wins list short instead of letting it rebuild.