Skip to content
TrafficBlog TrafficBlog Est. 2019
Vol. 7 · Issue 41 — Tuesday, 9 AM ET refresh Lisbon · Austin · Berlin · ISSN 2789-0144
SEO与数字营销

SEO Metrics Intake Pack: What to Lock Before You Edit a Page

For an SEO executor: what the 9 SEO metric groups count, what to freeze in one comparable export, and when to repair capture or eligibility before editing copy.

TTrafficBlogAuthor

Brand promotion: this article introduces Guangsuan's SEO measurement guidance. It is a vendor introduction, not an independent review.

If one chart rises and another falls the day before you plan edits, hold the rewrite. These nine groups of metrics answer different questions. A shift in one figure only proves something about that stage. For method definitions, see Guangsuan's 9 SEO metrics guide.

Freeze the comparison frame first

Put every number on the same footing: same property, same search type, one finished period, same pages, same country and same device. Note language, brand split and landing pages on each file. Leave the unfinished latest day out. Keep absent rows marked absent; do not fill them with another tool benchmark.

This discipline matters because Console and site analytics do not record the same event. Console records shows and picks inside search. GA4 records what could be observed after arrival. Gaps in time zone, coverage, permission choice, tag firing and canonical assignment can pull the totals apart, so do not force them to match.

Nine groups in working language

1. Search shows and picks versus site visits

A Console pick is not a site session. When Console looks normal but analytics is near empty, inspect tags, permission handling, streams and filters first. Then list which landing pages, countries and devices moved before discussing season, stock or copy.

2. Query cover and mean position

Mean position is a show-weighted summary for the filters you chose, not one live rank every searcher saw. Outside keyword totals rest on their own lists, markets and dates. Group demand by product family, use case and buying question, keep the country, device, language and brand split fixed, then open the live result for a slipped query and ask whether it favours a product, guide, video or local answer.

3. Pick share inside narrow slices only

Compare only matching purpose, position band, device, market and page. New low-position shows can drag a blended share down while total picks still rise. Teaching arithmetic only with invented figures: 20 selections out of 1,000 shows equals 2%, which explains division and sets no pass mark. State what the page answers and describe only genuine material. Do not add an unverified certificate, test claim or performance figure to win picks.

4. Contacts and ratios with both sides named

A chat tap, finished form, download, basket move and paid order are separate outcomes. Test that the live event fires, keep trials, repeats, spam and business-confirmed contacts on separate lines, and name the ratio you use. Sessions holding the chosen event divided by sessions differs from events divided by sessions when one visit holds several events. As a rule, query lines cannot be tied one by one to a single order.

5. Bounce under GA4 rules

In GA4, bounce is the share of visits without engagement, the reverse of engagement share. It is not the older one-page idea. Confirm the actual property setting; default engagement conditions include time beyond a short threshold, a chosen event, or at least two views. One answered question on a single page can still satisfy the reader, while repeated view hits from a setup fault can inflate engagement.

6. Focus time and depth against the task

Focus time records time with the page in front. Longer reading can mean careful study or confusion. Finding a detail on one page is not worse than needing four pages. Read time with later pages and completed tasks together, then add the absent specification, unit, condition or next step rather than adding sections by quota.

7. Eligibility to appear

Start from a short list of canonical addresses meant for indexing and work important entries first. Login areas, parameter copies and material kept out by choice stay outside that list. Group the indexing report by reason and open samples. Asking for indexing starts a review and never guarantees entry. A crawled-but-not-indexed note alone does not prove thin copy, and a correctly merged duplicate is not by itself a removal job. Use canonical only for duplicate or highly similar material.

8. Outside links versus later movement

Total links and referring sites are different counts. Coverage, refresh timing and de-duplication vary by crawler, so one absent row does not prove a live link vanished. The Links report cannot stand in for a live indexing check of each source page. Where outside help is used, match source, target, anchor, attribute, availability and check time against the delivery list. Going live, entering the index, and later ranking better stay as three separate steps.

9. Vendor scores inside one scale only

Scores from outside vendors use different models and cannot be swapped. Compare a close peer set inside the same tool and window while still reading real pages, real links and search outcomes. Do not tune only for the outside number while neglecting technical access, page material and contact handling.

Repair sequence before copy edits

  1. Capture first: Console normal but analytics near empty means restore recording before judging demand.
  2. Eligibility next: key canonical outside the index means inspect access, signals, material and discovery path, repair only the block proven on that path, then log a recheck.
  3. Presentation after: shows up with blended share down and picks up means inspect query and position mix and compare narrow slices before rewriting titles.
  4. Contact path last: picks up with business-confirmed contacts flat means inspect purpose, market, mobile form and qualification, then walk the product path.

Log dates, filters, what changed, owner and open items. For whole-site scope across technical repair, content supply and links, see Guangsuan Google SEO service.

Country and device split you can repeat

Apply the device filter first, then the country filter, on identical property and dates. Open Pages to locate the address that moved, then open its query list with shows, picks and position together. Tag new, continuing and off-task queries rather than sorting by share alone, because a row with very few shows can swing on one pick. For one important query, read the title and summary actually shown for that country and device and note whether the page type matches what results favour. If the query asks for model, delivery, price or compatibility but the ranking address only explains ideas, add product or comparison material and an enquiry entry before changing wording.

Handoff list before asking for help

Bring facts another person can repeat: site address, target country and language stated separately, affected addresses, one finished export with filters, what changed and when, who can approve code or copy, and how your team qualifies an enquiry. Note tags and permission handling where analytics is near empty while Console holds data. Guangsuan joins technical repair, page work and outside resources in one effort, with scope set by the live service page and agreement.

FAQ

Do flat shows require a rewrite? No. Flat shows mean the page is seldom offered for the query. Confirm eligibility, entry links and whether the page type fits what results favour first.

Does a falling blended share force a title change? Not alone. Check whether new low-position queries widened the mix, then compare narrow slices before editing.

Does longer reading prove success? No. Longer time can mean useful study or confusion. Add the missing condition or next step and check whether the task was finished.

Ship work that moves pipeline — every Tuesday, in your inbox.

Get the Weekly Traffic Brief