Track AI visibility through a company rebrand by locking a dual-name prompt set before the announcement, then running it through ChatGPT, Gemini, Perplexity, Claude, Copilot and Google AI Overviews at fixed checkpoints for at least 90 days after launch. The old brand name keeps surfacing in AI answers long after the new site is live, because training data and citation caches refresh on their own schedule, not on your launch calendar. Treat the transition as a monitoring project with a start date and an end date, not a one-time dashboard swap.
- Run old-name and new-name prompts side by side from 30 days before launch through at least 90 days after it.
- AI citations lag behind website changes because training data and citation caches don't refresh in real time.
- Google AI Overviews, ChatGPT, Perplexity and Copilot update at different speeds — track them separately, not as one blended score.
- An AI Visibility Score from a platform like Koalr lets you baseline the old brand and compare it against new-name citations checkpoint by checkpoint.
Why this matters
A rebrand resets your domain, your logo and your homepage copy overnight. It doesn't reset what large language models learned about your old name during training, or what's cached in an AI visibility platform like Koalr versus what's still sitting in a model's memory from last year's crawl.
Check AI visibility only once, right after launch, and you'll see a dip and assume something broke. Most of the time nothing broke — the citation graph hasn't caught up yet. Marketing teams that skip a rebrand-specific tracking plan in 2026 tend to either panic and consider a second rebrand too soon, or stop watching after week one and miss a genuine drop in Share of Voice.
A rebrand is also the one moment your entire prompt set goes stale on day one. Every prompt built around the old name needs a parallel version built around the new one, run from the same day forward.
How to track AI visibility through a company rebrand
- Freeze your current prompt set 30 days before the announcement and run a full baseline audit under the old name, category terms, and product terms.
- Build a parallel prompt set under the new brand name — same category terms, same product terms — and start running both from launch day forward.
- Segment every citation and mention by which name triggered it. Blending old-name and new-name results into a single AI Visibility Score hides the exact thing you're trying to measure.
- Track platforms separately. Google AI Overviews pulls from a live search index and tends to reflect domain and name changes faster than models with static training data.
- Re-run the full audit at 30, 60 and 90 days after launch and compare the old-name and new-name trend lines against each other, not just against a single post-launch snapshot.
How fast each platform picks up a name change depends on how it sources answers:
| Platform | How it sources answers | What that means for a rebrand |
|---|---|---|
| Google AI Overviews | Live search index | Usually reflects domain and name changes within weeks of re-crawl |
| ChatGPT | Training data, plus live browsing in some modes | Old name can surface for months without browsing enabled |
| Perplexity | Live web citations with its own refresh cycle | New name appears once its crawlers re-index the pages citing you |
| Claude | Training data, largely static | Similar lag pattern to ChatGPT without retrieval |
| Microsoft Copilot | Bing index plus model | Update speed tracks how fast Bing re-crawls your new domain |
30 days before launch: build the dual-name baseline
Run a full AI visibility audit across major AI platforms under the old name before the announcement goes out. This is the number every later checkpoint gets measured against — without it, you have no way to prove the rebrand cost you visibility or that it didn't.
Document which prompts currently cite the old brand, which ones name a competitor instead, and where you show up as a mention without a citation. That split matters more once the new name enters the picture.
Launch week: track both names in every prompt set
The day the new name goes live, add it to every existing prompt rather than replacing the old name outright. Tracking your brand in ChatGPT and AI answers works the same way during a rebrand as it does normally — the only change is running two names through the same prompt set instead of one.
Expect the old name to still win most citations in week one. That's the model's training data talking, not a signal your rebrand failed.
30-90 days after launch: watch the citation migration
This window is where the real signal shows up. Track the ratio of old-name to new-name citations at each checkpoint rather than the raw citation count for either name alone.
A rebrand that's working shows the ratio shifting steadily toward the new name by day 60. A rebrand that's stalling shows the old name holding flat or even gaining share, which usually points to backlinks, directories or review sites still using the old brand.
90+ days: confirm the cutover
By 90 days after launch, most brands see AI Overviews and Perplexity fully on the new name, with ChatGPT and Claude still trailing depending on their training cutoff. If the old name is still winning the majority of citations at this point, the gap usually sits outside your own website — old directory listings, syndicated press, or a Wikipedia-style entry nobody updated.
Why AI visibility lags during a rebrand
- Training data cutoffs — models trained before your rebrand simply never saw the new name.
- Citation caches in Perplexity and AI Overviews refresh on their own cycle, not in real time.
- Backlink equity under the old domain takes time to transfer trust signals to the new one.
- Third-party sources — review platforms, directories, comparison sites — update slower than your own site.
- Structured data and schema changes need to be re-crawled before models can act on them.
- Old mentions in training corpora persist even after the new name is live everywhere on your own properties.
Report AI visibility to leadership
Show the pre- and post-rebrand trend line without guessing.
How long does it take for ChatGPT to recognize a new brand name?
There's no fixed number — it depends on ChatGPT's training cutoff and whether browsing is enabled for the query. Expect the old name to keep surfacing until a model update trained after your rebrand ships, which can run well past the 90-day mark used for most rebrand tracking plans.
Does a company rebrand hurt AI Overview citations more than organic rankings?
Usually the opposite: Google AI Overviews pull from a live index, so they tend to reflect a rebrand faster than the citation behavior of ChatGPT or Claude. Organic rankings and AI Overview citations can move at different speeds during the same transition, which is why they need separate tracking lines.
Should you keep the old brand name in your prompt set after the rebrand is complete?
Yes, for at least one full audit cycle after you've confirmed the cutover. Dropping the old name too early means you lose the ability to prove the old-name citations have actually stopped, rather than just gone quiet for one checkpoint.
FAQ
How do I track AI visibility through a company rebrand in 2026?
Track AI visibility through a rebrand by running old-name and new-name prompts in parallel from 30 days before launch through at least 90 days after it. Segment citations by which name triggered them and compare the two trend lines rather than a single snapshot.
What's the difference between a brand mention and a citation during a rebrand?
A mention names your brand in an AI answer without linking or crediting a source; a citation directly attributes the answer to your page or domain. During a rebrand, mentions of the old name often outlast citations of the new one because models recall the name from training even without a live source.
Do all AI platforms update at the same speed after a rebrand?
No. Google AI Overviews and Perplexity pull from live indexes and tend to reflect a rebrand within weeks of re-crawl, while ChatGPT and Claude rely more on static training data and can lag for months.
How often should you re-run an AI visibility audit during a rebrand?
Re-run the full audit at 30, 60 and 90 days after launch, in addition to the baseline audit taken 30 days before the announcement. Checking only once right after launch misses the citation migration that happens over the following months.
Does redirecting your domain fix AI visibility during a rebrand?
A domain redirect helps search crawlers find your new site, but it doesn't change what a language model already learned during training. Expect the old name to keep appearing in AI answers for a period even with a clean redirect in place.
What happens to Share of Voice when you rebrand?
Share of Voice typically dips right after launch because citations split between the old and new name instead of consolidating under one. It recovers as the new name accumulates its own citations, which is why tracking the split matters more than watching a single blended score.
Should you track competitors during your own rebrand?
Yes — a rebrand is exactly when competitors can pick up citations you temporarily lose. Watching competitor Share of Voice during the same 90-day window shows whether a dip is a rebrand lag or a real loss to a rival.
Can you speed up how fast AI platforms recognize a new brand name?
You can't force a model retrain, but updating structured data, press coverage and third-party directories with the new name gives live-retrieval platforms like Google AI Overviews and Perplexity a faster path to the correct citation.
One last thing
The old brand name usually keeps winning citations for a stretch after launch that outlasts most teams' patience — don't schedule a second rebrand because ChatGPT hasn't caught up by week three. Watch the ratio, not the raw count, and let the 90-day checkpoint make the call in 2026 rather than the launch-week dashboard.
