Discussion
  • Technical SEO

A CMS sent inaccurate lastmod dates for years. Has anyone seen Google start trusting lastmod again after it was fixed, and how long did that take?

Published 1 reply

Discussion starter: a technical SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    The short answer: Google has not published a recovery timeline, public evidence on how long recovery takes is thin, and no controlled public test exists that I can point to. What Google has documented is the condition it applies, and that condition tells you how to find out for your own site. Google says it uses lastmod if the value is consistently and verifiably accurate, and it gives comparing the date with the page’s last modification as the example of verification. Google can only check a date against the page when it fetches the page, so in practice trust can rebuild only as fast as Google recrawls the pages you update.

    That changes the useful question from “when will Google trust it again?” to “are genuinely updated pages being fetched sooner than they would be anyway?” Server logs can answer that within a few recrawl cycles. On a small or frequently crawled site, the answer may be that lastmod barely matters, which is worth knowing before anyone spends more engineering time on it.

    What Google has documented, and what it has not

    Google’s sitemap documentation says:

    “Google uses the <lastmod> value if it’s consistently and verifiably (for example by comparing to the last modification of the page) accurate.”

    The same page says the value should reflect the last significant update, and defines significant: an update to the main content, the structured data or the links on the page is generally significant, while an update to the copyright date is not. It also says Google ignores <priority> and <changefreq>, so of the three optional tags in the standard sitemap protocol, lastmod is the only one Google uses.

    If your CMS stamped every URL with the current date, Google has addressed that pattern directly. In April 2025, John Mueller said that setting today’s date across a sitemap does not work in anyone’s favour and called the practice lazy. He later clarified that he meant sitemaps showing the current date for every URL, not normal lastmod use.

    What Google has not documented matters just as much:

    • whether accuracy is judged per URL, per sitemap file, per host or per site
    • how many accurate observations it takes before lastmod is used again
    • whether a long history of bad dates takes longer to overcome than a short one

    Any timeline you read, including in replies to this question, is an inference from someone’s own logs on someone else’s site. It is a data point, not a rule.

    Why recovery speed follows your recrawl rate

    Think of lastmod as parcel tracking notices. If a courier’s “delivered today” messages were wrong for years, you would keep checking the doorstep yourself. You would only start relying on the notices after a run of them matched what you found, and the more parcels you receive, the sooner that run builds up.

    The technical version, offered as a working model rather than a documented algorithm: Google schedules recrawls from many signals, such as how often a page has changed before and how important it looks, and lastmod is one hint layered on top. Each time Googlebot fetches a page, Google can compare the stated date with the change it actually finds, and each accurate pair is evidence. Pages recrawled daily supply that evidence quickly. Deep pages recrawled every month or two supply it slowly, and those are the pages where a trusted lastmod would help most. So the pages that most need trusted dates are also the slowest to re-earn trust.

    The type of inaccuracy tells you what to fix and what the old data looked like from Google’s side.

    Old pattern What the sitemap claimed What a recrawl found The fix
    Every URL stamped with the build or request time Everything changed today Almost nothing changed Derive lastmod from the content’s last significant edit, not from deploys or sitemap generation time
    Dates bumped by template, footer, widget or comment changes Frequent changes Changes that Google’s definition would not call significant Exclude non-significant fields from the “modified” calculation
    Dates stuck at first publication Nothing changed Content that had changed Update lastmod when the main content, structured data or links change

    All three fail Google’s test, since a date that is wrong in either direction is not verifiably accurate. The first two also cry wolf, and the first is the pattern Mueller criticised.

    Check whether lastmod matters on your site at all

    Google’s crawl budget guide is aimed at sites with a million or more unique pages that change about weekly, sites with 10,000 or more pages that change daily, and sites with many URLs in “Discovered - currently not indexed”. Below that scale, Google may already recrawl your important pages often enough that lastmod adds little. Deep, rarely linked pages on any site are the exception, since the sitemap may be one of the few recrawl prompts they get.

    A quick check before you build a measurement project: pull 50 pages that had a real update last month and look up when Googlebot next fetched each one. If most were fetched within a day or two anyway, lastmod is not your constraint, and the fix was worth doing for accuracy rather than speed. If deep pages waited weeks, lastmod is worth measuring properly.

    How to tell whether Google is using your lastmod again

    A page changes in a way Google’s definition calls significant
    ↓
    The sitemap lastmod updates, ideally without a new internal link to the page at the same moment
    ↓
    Googlebot reads the sitemap (Sitemaps report “Last read” date, or your logs)
    ↓
    Is the page fetched sooner than its normal recrawl interval?
    ↓
    Yes, repeatedly: lastmod is plausibly being used. No: not yet, or crawling is not your constraint
    1. Audit the fix first. Sample a few hundred URLs and compare sitemap lastmod with the CMS revision history. Dates should move only on significant edits. Confirm that a template deploy, a price sync or a new comment no longer bumps the date.
    2. Build a baseline from logs. For pages that did not change, calculate the median days between Googlebot fetches, segmented by template and depth. That is each segment’s natural recrawl interval. Verify Googlebot by reverse DNS, because user agents can be spoofed.
    3. Measure update-to-fetch latency. For pages with a genuine update, measure the time from the first sitemap fetch that contained the new date to the next Googlebot fetch of the URL.
    4. Compare against the baseline in slow segments. Pages Google fetches daily tell you nothing. The signal is in segments with long natural intervals: if updated pages there are fetched well before their usual interval, something is prompting the crawl.
    5. Track monthly cohorts. If trust is returning, latency for each month’s updated pages should fall relative to the baseline over successive cohorts.
    6. Cross-check in Search Console. The Crawl stats report splits requests by purpose into Discovery and Refresh, and URL Inspection shows the last crawl date for individual pages. Both are coarse, so use them to sanity-check the log analysis, not to replace it.

    Watch the confounders. Updated pages may gain new internal links at the same time, from “recently updated” modules, category pages or the home page, and those links can prompt a crawl on their own. A manual “Request indexing” in URL Inspection does the same. Keep a list of URLs that received either and analyse them separately.

    Hypothetical example:

    A retailer with 300,000 URLs finds that product pages in long-tail categories have a natural recrawl interval of about 40 days. The CMS fix went live in March. Pages genuinely updated in April are fetched a median of 25 days after their sitemap date changes, and pages updated in July a median of 6 days, while the baseline stays near 40. That pattern is consistent with Google using the dates again. If every cohort had sat near 40 days, the honest reading would be “no evidence yet”, not “Google ignores lastmod”.

    What to do while trust rebuilds

    • Set historical dates from revision history, using each URL’s last significant edit. Do not mark the fix by setting every URL to the day it shipped: a sitewide jump to one date repeats the old pattern at the worst possible moment.
    • Omit lastmod where you cannot work it out. The field is optional, and leaving it out is better than guessing, because a wrong guess is one more inaccurate date.
    • Split sitemaps by section or template if you have not already. Google has not documented per-file trust, so do not expect new files to reset anything. The value is measurement: the Sitemaps report and the Page indexing report filtered by sitemap both work per file.
    • Keep other recrawl prompts working: links from frequently crawled pages to updated content, and “Request indexing” for a small number of priority URLs, since it is quota-limited.
    • Add a regression alert. If a large share of URLs change lastmod on one day without a matching content release, the old bug, or a new one, is back. A daily check on the sitemap diff catches it before Google sees months of bad data.

    How long to measure, and what would change the decision

    Set the window by your slowest important segment. You need several recrawl cycles in that segment to see a trend. A segment Google fetches weekly can give you an answer within a month or so, while one fetched every 40 days needs about four months to produce three or four readable cohorts. Avoid reading anything into a window that includes a site migration or a large template release, since both can change crawl patterns by themselves.

    What you see Likely reading What I would do
    Updated pages fetched well before baseline, with the gap growing across cohorts Lastmod is plausibly being used again Keep dates accurate, keep the regression alert, move on
    No difference, but pages are fetched within a day or two anyway Crawling is not your constraint Stop spending engineering time on lastmod beyond keeping it accurate
    No difference, and deep pages stay stale in the index (old last crawl dates, outdated snippets) Not trusted yet, or the fix leaks somewhere Re-audit the fix, strengthen internal links to deep pages, keep measuring
    Faster fetching only for pages that also got new internal links The links are doing the work Credit the links, and treat lastmod as unproven for your site

    The practical position

    If this were my site, I would stop waiting for a date when Google “forgives” the old sitemaps and treat it as an evidence problem. Make every date accurate from now on, backfill history honestly or leave it blank, and measure update-to-fetch latency against the baseline in the slowest segments that matter commercially. If that latency falls over successive cohorts, the fix is working. If it never moves but those pages are crawled promptly anyway, the fix was still right, and lastmod was simply not the bottleneck.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.