Google Fritz Update (July 2003): What It Was and What It Means Today

The Google Fritz update, in July 2003 (some lists say “summer 2003”), was when Google stopped rebuilding its index once a month and began updating it incrementally, close to daily. It was a change to indexing, not a penalty, and it is the reason rankings have moved a little every day ever since.

Fritz ended the monthly “Google Dance” and started what webmasters called “everflux”. If you are researching it, you probably want to understand how Google’s update model developed, and that is what this page focuses on. For individual updates, see my full list of Google algorithm updates.

What the Fritz update changed

Before Fritz, Google worked in batches. It crawled the web, rebuilt the whole index, and rolled it out roughly once a month. Rankings jumped during the rollout and then stayed broadly still until the next one.

After Fritz, Google added to and adjusted its index in small pieces, close to daily. Search Engine Land describes it as Google moving to daily indexing and starting “everflux”, the forum term for constant small ranking movement. Wikipedia describes it as the shift from batch to continuous index updates.

The practical effects:

• Rankings changed in small steps every few days instead of in one monthly jump.
• New and updated content could be indexed and ranked much sooner.
• The ritual of watching data centres for “the dance” lost its point.

Timeline

Who was affected

Every site, in the sense that rankings started to shift continually. No category of site was targeted, and nobody was penalised by Fritz itself. Sites that published and updated content often benefited most, because their changes could be picked up faster.

How to tell continuous movement from a real update

Fritz is why you need to separate everyday fluctuation from a genuine change. This is how I do it:

1. Set a baseline. Look at 8 to 12 weeks of daily clicks in Search Console’s Performance report. That shows your normal range of daily movement.
2. Look for a break outside that range that lasts more than a week. Small daily wobble is the everflux Fritz introduced.
3. Check whether a confirmed update overlaps. If a drop starts within a known rollout window, it is a candidate cause. If not, look at your own changes first: migrations, template edits, robots.txt, redirects.
4. Check indexing. With continuous indexing, pages also drop out continuously. The Pages report in Search Console shows whether important URLs have been excluded.

How to recover from the Fritz update

There was nothing to recover from. Fritz was not a penalty. Webmasters had to adapt to continual small changes, and sites that updated regularly gained because fresh content got indexed faster.

What adapting looks like now:

• Make it easy for Google to find changes quickly. Keep an accurate XML sitemap with real lastmod dates, clean internal links to new pages, and a fast server. My website speed service and technical SEO service cover both.
• Update content when facts change, not just to change a date. Faster indexing rewards genuine updates, not cosmetic ones.
• Report on trends weekly or monthly, not daily, so normal movement does not trigger rushed changes.

Is it still relevant today?

Yes, in principle. Google’s index still updates continuously, and Caffeine in 2010 took that model much further. Every time you see your rankings shift slightly from one day to the next, you are seeing the system Fritz started.

Frequently asked questions

Sources

• Search Engine Land: Google algorithm updates
• Wikipedia: Timeline of Google Search
• Search Engine Roundtable: Google updates list
• Ranked.ai: Google updates and business, part 1

If your new pages are slow to index, or you cannot tell normal movement from a real drop, I can help. My SEO audit service covers indexing, crawl health and traffic trends. Request a free SEO audit.