How Do You Track Your AI Search Visibility Over Time?

I'm Ryan Goering, and I run BaaDigi, a marketing agency for contractors and local service businesses. We spend a lot of time on a question most business owners haven't caught up to yet: when someone asks ChatGPT or Google's AI "who's the best roofer near me," does it say your name? And more importantly, is it saying your name more often this month than last? This post is about the second question, because that's the one that actually moves revenue.

Why track AI visibility over time instead of checking once?

A one-time check is a photograph. Tracking is a video. Here's why the video matters more.

AI answers are not static. The same prompt can return different businesses on different days because the models update, the sources they crawl change, and your competitors are publishing content that gets pulled into answers. If you check once and see your name, you might celebrate and stop. Two weeks later a competitor's new comparison page gets cited instead, and you never notice because you never looked again. By the time your phone stops ringing, you've lost months.

Tracking over time answers the questions that actually drive decisions:

• Are you trending up or down? One data point is noise. A line over eight weeks is signal.
• Did that content change work? If you rewrite a service page or publish a new FAQ, tracking shows whether AI engines started citing it.
• Is a competitor eating your share? You can only see this by watching who else gets named across repeated checks.
• Which engine matters for you? You might be strong in Perplexity and invisible in Google AI Overviews. You only learn that by measuring each one repeatedly.

If you want to see where you stand right now before you commit to tracking, start with our free AI Search Visibility Checker. It gives you the snapshot. Tracking is what you graduate to once you know you have a gap worth watching.

What metrics matter in AI visibility tracking?

Not every number is worth watching. Here's what I tell clients to actually track, in plain terms.

Are you named at all?

The most basic metric: for a given question, does the AI mention your business by name in its answer? This is a yes/no you check repeatedly across a fixed list of the questions your customers actually ask. Track the percentage of your prompts where you show up. That percentage moving up is the headline number.

Which engines cite you?

ChatGPT, Perplexity, Google AI Overviews, Gemini, and Copilot all pull from different sources and behave differently. Being cited in one is not being cited in all. Track each engine separately so you know where to put your effort.

What sources get cited?

When an AI names you, it usually pulls from somewhere: your site, a directory, a review platform, a news mention. Tracking which sources feed the answer tells you where to invest. If Perplexity keeps citing a directory listing instead of your own site, that's a fixable gap.

How do you compare to competitors?

Share of voice is the honest scoreboard. Across your tracked questions, how often does the AI name you versus the three or four competitors you care about? A rising share of voice is the metric that most directly maps to more leads.

How often should you check AI search visibility?

For most local businesses, weekly is the sweet spot. Daily is overkill and mostly measures noise. Monthly is too slow to catch a competitor moving on you. Weekly gives you a clean trend line and enough data points that a real change stands out from random variation.

A few adjustments I make depending on the situation:

• Weekly baseline for steady monitoring once you're established.
• After any content change, check more often for a couple weeks to see if the change landed, then settle back to weekly.
• Monthly review where you zoom out, look at the eight-to-twelve-week trend, and decide what to build next.

The key is consistency. Same prompts, same engines, same day of week. If you change the questions every time, you're not tracking, you're just re-checking, and the history becomes useless.

Free vs. paid AI visibility trackers: what's the difference?

You can start free and it's a smart move. A free checker or a manual spreadsheet gets you going. The tradeoffs show up as you scale.

Free and manual tracking works if you have a short list of prompts and the discipline to run them yourself on a schedule. You open each AI engine, ask your questions, and log yes/no in a spreadsheet. It costs nothing but your time, and honestly for a single-location business with ten key questions, it's a perfectly reasonable start.

Paid trackers earn their keep when the manual work becomes a chore you skip. They run your prompts automatically, log everything, chart the trends, watch multiple engines, and flag competitor movement without you lifting a finger. The value isn't magic data, it's that the tracking actually happens every week instead of falling off your plate. I've watched too many owners "plan to check" and never do. Automation removes the discipline problem.

If you're weighing options, I put together a rundown of the best AI visibility tracking tools so you can see how the field compares. And if you're deciding between the big-name platforms and what we offer, we have head-to-head breakdowns: BaaDigi vs. Ahrefs Brand Radar and BaaDigi vs. Rankscale. The short version: the enterprise tools are powerful but built for brands tracking national terms, while our approach is built around the local, near-me questions that actually feed a contractor's pipeline.

Why the local angle changes everything

Most AI visibility tools were built for big brands checking whether ChatGPT knows their product category. That's a different game from a local contractor. When someone asks an AI "who's the best plumber in Huntington Beach" or "reliable electrician near me," the answer is intensely local, and the sources feeding it are local: your Google Business Profile, local directories, review sites, and geographically relevant content.

For local businesses, tracking has to be location-aware. Being named nationally means nothing if you're invisible for the city where your customers live. That means your tracked prompts need to include the location language your real customers use, and your share-of-voice comparison needs to be against the businesses that actually compete in your service area, not national chains.

This is also why the source-tracking metric matters more for local. If an AI keeps citing a competitor's local landing pages or a directory where you have a thin profile, that's your roadmap. Fix the profile, build the local pages, earn the citations, and watch the mention rate climb over the following weeks. That feedback loop, build then measure then build again, is only possible with tracking, not one-time checks.

How I'd set up AI visibility tracking this week

If you're starting from zero, here's the sequence I'd follow:

• Write down the ten to twenty questions your customers actually ask an AI when they're looking for what you do, with your location baked in.
• Run every one of them today through ChatGPT, Perplexity, and Google AI Overviews. Log whether you're named and who else is.
• That's your baseline. Save it.
• Repeat the exact same run next week, same day. Now you have a trend.
• After a few weeks, look at the pattern, find the gaps, fix the content or profiles feeding those answers, and keep measuring.

Once the manual routine proves there's a gap worth closing, that's when a real tracking tool pays for itself by keeping the measurement going without you.

Want to go deeper on the analysis side of this? I've written companion pieces on the AI search visibility analysis tool approach and on choosing an AI search visibility checking tool for the initial snapshot. Read those alongside this one and you'll have the full picture: check where you stand, analyze why, and track whether your work is moving the needle.

The businesses winning in AI search right now aren't the ones who checked once and hoped. They're the ones treating AI visibility like they treat their Google rankings: measured on a schedule, watched over time, and improved deliberately. If you want a hand setting that up, start with the free checker and reach out for a quote on ongoing tracking.