# First Search Impressions

A Creators Toolbox version with original steps, prompts and an illustrative example, inspired by [Shubham Kshatriya's source workflow](https://x.com/shubham_habib/status/2123280693304381542).

Turn a new site's early Search Console impressions into a small page-improvement experiment.

1. Choose a useful observation window. A single day on a new site can swing sharply with a handful of queries. Start with 7 or 28 days when available and record when the site launched. Keep country, device and search-type filters consistent. Search Console data can lag and small samples do not establish a stable trend.

2. Export queries and pages, then look for repeated, relevant searches. Positions around 8 to 20 can suggest pages worth examining, but average position is an aggregate rather than a promise that every searcher saw the same result. Ignore unrelated impressions unless they reveal confusing page copy.

3. For each promising query, inspect which URL appears. If the wrong page ranks, compare the pages' purposes, headings and internal links. Do not immediately create another page targeting the same words. First decide whether the existing page should answer the question or whether a distinct task needs its own destination.

4. Read the visible results for the query and note missing information. Identify the user's next question and answer it with verified details. For a free subtitle tool, that might mean supported file types, export limits and an example file. Avoid copying competitors' wording or padding the page to a target length.

5. Add two or three relevant links from existing pages that naturally discuss the task. Use descriptive anchors and real links. A link from a video-editing guide to the subtitle tool is useful; linking every page to it with the same keyword is not a substitute for relevance. Check that the target returns the expected page and is indexable.

6. Change a small set of pages and record the release date. Compare the next seven-day window with the preceding one using the same filters, while remembering the data is observational. Watch query relevance, impressions, position and completed actions. A low click rate on a tiny sample is a reason to inspect the result, not proof that the title is broken.

7. Keep a simple review log: query, observed page, intended page, evidence gap, change and next check date. If impressions grow but conversions remain absent, inspect the offer and journey. If the pages are not indexed, return to technical diagnosis. Do not promise a deadline for rankings on an 18-day-old site.

Illustrative log: "free subtitle line checker" appears 35 times over 14 days and points to the general tools hub. The dedicated checker lacks a hub link and does not explain supported formats. Add a relevant hub link and a tested SRT example, then compare another 14 days. The numbers and query are an invented example, not a performance claim.

See [Google's AI search guidance](https://developers.google.com/search/docs/appearance/ai-features) and [Search Console metrics](https://support.google.com/webmasters/answer/7576553) for the underlying measurement limits.

Written by Creators Toolbox. External tool access and costs vary.
