Create a search baseline before rewriting the whole catalogue

A search baseline should describe the data you actually have. Approximate weekly impressions and clicks are useful, but they cannot reveal every query, conversion or customer segment.

Write the scope beside the number

Record the source, date range, property and any filters. If the numbers are an estimate supplied by the business, label them as such. Do not present them as an independently verified analytics export.

For example, approximately 120 clicks from 5,000 impressions gives a ratio of about 2.4%. That arithmetic does not establish a purchase conversion rate or explain why any particular page was clicked.

Separate page visibility from business outcomes

Search impressions and clicks answer a different question from completed orders. A customer may visit from search, return through a message and order later. Without the relevant measurement, do not assign the sale to the last page you edited.

Google describes Search Console as a tool for monitoring search performance and diagnosing aspects of a site's presence in Search. Use the reports available to the authorised property rather than inventing missing detail. Search Console introduction

Save a small, useful comparison set

Choose the pages you intend to change and retain their available baseline. Include query context where you have it, but keep the absence of data explicit. A screenshot of top pages is not a full query export.

Note promotions, stock changes and seasonality that may affect interpretation. The comparison is stronger when you know what else changed.

Set a narrow first question

Ask whether a specific page now attracts more relevant visits or answers the customer's question more clearly. Avoid changing titles, navigation, offers and the entire catalogue at once if you want to learn which problem was solved.

A small site can still make sensible improvements without perfect analytics. Fix demonstrably incorrect content while keeping performance claims modest.

Review without declaring victory too early

Choose an observation period appropriate to the traffic and change. Do not invent a universal number of days that proves success. Look for consistent evidence and investigate surprising movements.

Use the growth and selling hub to connect the search work to the actual buying journey. A baseline is valuable because it limits what you can honestly conclude, not because it supplies a flattering chart.

Illustrative case to check

Your only evidence is a weekly estimate of 5,000 impressions and 120 clicks plus a list of top pages. Keep those as the baseline and label the missing query and order data. You can still correct a misleading title, but you cannot claim that the edit increased sales by a particular percentage without the relevant measurement.