We show you the screens,
not the sales talk
AI search has no rankings table, so it is easy for an agency to say “it’s going well” and leave it there. We measure AI answers across Naver AI Briefing, ChatGPT, Gemini and Perplexity every week with a monitoring system we built ourselves and use in-house. Below are six screens from it. Every one of them measures our own company, so nothing is blurred out.
4
AI engines measured
Weekly
Automatic re-measurement and report
18
Our own overall score (out of 100)
0
Accounts sold to outsiders
SCREENS
Six screens, captured on September 9, 2026
The account being measured is WizThePlanning itself. Client screens belong to clients, so we do not publish them.
Screens are from our Korean-language dashboard.
01
Monitoring dashboard
This is our own account. Our overall score is 18 out of 100, and we are not hiding it. The dashboard shows a 7% share on questions that do not name us and 88% on questions that do, side by side. We keep them apart because blending the two inflates the number.
02
Weekly report
Every week we re-measure against the same criteria and show what went up and what dropped compared with the week before. This particular week recorded 2 newly cited keywords and 3 keywords that lost their citation. The losses stay in the report too.
03
Alert center
When the automatic measurement finishes early on Monday morning, the briefing and warnings land here. If a question that cited us last week stops citing us, a lost citation warning fires right away. This is the first screen the account manager opens at the start of the day.
04
Results by question and engine
We ask the four engines, repeatedly, the kind of questions real customers would ask. For each question we record separately whether the name appeared in the answer (mention), whether it was linked as a source (source), and whether it was used to support the answer (evidence). Being mentioned and being cited are different things, so we count them apart.
05
Site audit
We run all 52 pages of the site to score whether search engines and AI can read and cite it. Bot access, structured data, citability and technical health are scored separately, so the output is a list of specific items to fix.
06
Real traffic check
Showing up in an AI answer is not the finish line. With Google Search Console and GA4 linked, we count the visits that actually arrived from ChatGPT, Gemini and Perplexity. In the last 28 days, 27 visits came to our site from AI. It is a small number, and we report it as it is.
METRICS
We count things separately so the numbers do not inflate
Unbranded questions are the headline metric
Ask an AI "What do you think of WizThePlanning?" and of course it answers. So our headline metric, category share, counts only questions that do not name the brand. Brand questions are checked separately and used only for alerts.
Mention, source and evidence are counted separately
A name appearing in the answer text (mention), being linked as a source (source), and being used to support the answer (evidence) are different events. We record all three for every question. Citations are what bring customers.
Site health gets a score
Bot access, structured data, content quality, trust and technical health are scored axis by axis. If AI cannot read a site, no amount of content will get it cited, so we look at this first.
For an example of what re-measurement looks like on a real account, see our clinic case: re-measuring ChatGPT citations.
ROUTINE
The same routine, every week
Early Monday: automatic measurement
The question set goes to all four engines again, and the site audit reruns. These are not screenshots someone picked by hand.
Alerts first thing
Questions that lost a citation, questions where a competitor showed up, and new problems on the site come in as warnings.
One report a week
Change versus last week, the work done this week, and next week's priorities, written by your account manager.
Fixes in priority order
We work on site structure, content and Naver Place in the order the measurement points to. What we changed and why is recorded in the next report.
NOT FOR SALE
What we do not do with this system
We do not sell it
We built it to manage client work and use it only in-house. We do not take inquiries about the tool itself.
Clients do not get accounts
It is a specialist tool that needs interpretation. Clients get a weekly report with the account manager's reading of the numbers.
We do not guarantee visibility
AI answers vary by engine and by moment. Instead of a guarantee, we keep measuring against the same criteria and report the change in numbers.
FAQ
Frequently asked questions
Q1.Can clients log in and see the system?
No, we do not give out accounts. It is a specialist tool, and the numbers need someone who knows SEO and GEO to interpret them. Your account manager takes the numbers from this system, adds an interpretation, and sends you a weekly report. You do not need to learn the tool.
Q2.Do you sell the tool?
No. WizThePlanning built it for managing client work and uses it only in-house. This page is not a sales pitch for software. It shows the evidence we work from.
Q3.Which AI engines do you measure, and how often?
Four engines: Naver AI Briefing, ChatGPT, Gemini and Perplexity. Measurement runs automatically every Monday early in the morning, and a report follows once a week.
Q4.Why publish your own score?
Because we promise measurement instead of guaranteed visibility, and showing only the good numbers would break that promise. Leaving our overall score of 18 and our list of lost keywords in plain view is proof that the reports we send clients are built the same way.
Start with where your brand shows up today
We measure your current state with the same system first. You are welcome to take the results and stop there.