SEO Tools

Semrush Rebuilds Traffic and Market Data With a New Model, and Old Numbers Will Change

Semrush has rebuilt the engine behind its Traffic and Market toolkit, and the numbers it reports for competitors will change, including the historical ones. As of November 9, 2026, the toolkit runs on a brand new traffic prediction algorithm, supported by a broader set of data sources and a purpose-built machine learning model engineered for channel prediction. Semrush has recalculated metrics from January 2023 onwards with the new model, so values retrieved before the change may differ from the values the toolkit shows now.

What changed

Traffic and Market is Semrush's competitive intelligence toolkit, used for market research and for sizing up competitors' traffic and channel mix. The upgrade replaces the traffic prediction algorithm underneath it with a new machine learning model, and widens the data sources feeding it. The stated goal is more accurate traffic estimates and channel attribution, so marketers get a clearer picture of where a competitor's visibility comes from and where to focus their own efforts.

The part that matters most for anyone who has already reported on these numbers is the recalculation. Semrush's Trends API has recalculated historical data with the new algorithm, so a traffic figure pulled in October and the same figure pulled in November may not match, even for the same site and the same month. The company notes that values may differ from what was retrieved previously, and invites users to reach out if the update affects their reporting. That is a polite way of saying that past reports built on these estimates are now out of date, and that any trend line crossing the change needs a note explaining why it moved.

Why now

The timing is not an accident. Traffic estimation has got harder this year, and not only because of AI search. From late October 2026, Google changed how it responds to automated queries, and keyword counts and traffic estimates in Semrush dropped, a change noted by independent reviewers working with the tool in November. A traffic model trained on the old shape of search will misread the new one, and a rebuild is the honest response to a measurement environment that has shifted underneath the tool.

Semrush has also been expanding in the other direction, toward AI search itself. The same November release notes include a native Semrush connector for Claude, giving Semrush One and SEO Classic subscribers direct access to live search and market intelligence inside AI conversations, built on a dataset the company puts at 28.8 billion keywords, 43 trillion backlinks and 317 million LLM prompts. A separate MCP connector for Perplexity brings Semrush search intelligence into the AI search engine itself. The company, now an Adobe company, is positioning its data as something that lives inside AI tools, not just on its own dashboard.

How to handle the recalculation

The practical steps are straightforward, but they need doing now, before the next client report goes out. Re-pull any competitor traffic figures that matter, and compare them with the numbers in existing reports. Where a figure has changed, annotate it with the date of the change and the reason, so that nobody reads a restated number as a real traffic move. Re-check any trend line that spans the November 9 change, and be ready to explain to a client why a competitor's traffic appears to have jumped or fallen when nothing about the competitor changed.

It is also worth re-checking channel attribution specifically, because that is what the new model was built to improve. If the new model attributes more traffic to one channel and less to another, the story about where a competitor gets its visitors may have changed even if the total is similar. The channel split is the part of the toolkit that feeds budget decisions, so a restated split deserves a fresh look rather than a quiet carry-over of the old one.

Why it matters for SEOs and link builders

For SEOs, the rebuild is a reminder that every third-party traffic estimate is a model, not a measurement, and models get rebuilt. A competitor's Semrush traffic is an estimate built from a sample, and when the model changes, the estimate changes with it. The right way to use these numbers has always been for comparison and trend-spotting, not as a substitute for first-party data from Search Console and analytics, and the recalculation is a good moment to restate that to anyone who treats a Semrush number as a fact.

For link builders, the change matters because traffic estimates are the screen most prospecting starts with. A site whose estimated traffic jumps under the new model may look like a better prospect than it did last month, and a site whose estimate falls may look worse, without either site having changed at all. Re-run prospecting screens against the new numbers before pitching, and check the sites that moved most, because a restated traffic figure is a reason to look again, not a reason to trust the new number blindly.

What to watch next

Watch how much the restated numbers actually move for the sites you track, and whether the new model's channel attribution tells a different story about where competitors get traffic. Watch also for whether other tools follow with rebuilds of their own, because the same pressures, Google's changing responses to automated queries and the growth of AI search, apply to every traffic estimation model. The tools that admit their numbers changed, and explain why, are the ones worth keeping.

Sources

Sources: Semrush release notes, November 16, 2026.

Related reading on this site: November Core Update Collides With Black Friday, and a Clean Read Will Not Come Until December, Core Update Volatility Started Days Before Google Announced It, November Search Console Click Spikes May Be Bots, SEJ Reports, SEJ Pulse: Core Update, ChatGPT Business Hub and AI Mode's New Model, AI Overview Citations: What 15 Studies Agree On, and Don't, and Black Friday SEO Starts Eight Weeks Out, and November Is Already Late.

More from Linkbaits

Turn this playbook into links

Linkbaits helps you create content that earns citations and backlinks. Start with the toolkit and build your first linkable asset this week.

Start free trial