Hook: Your rival's visibility line sits beside yours on one axis, same model, same prompts, same window. Tracking two rivals by hand costs EUR 15,120 a year.
What You'll Be Able To Do
- Add a competitor as a project of its own and scan it on the same AI model your own project runs on
- Draw both visibility lines on one axis, over three months, six months or a year
- Switch the question for both lines at once, across the Rankfor Index and its seven pillars
- Open every competitor the model named while answering your own prompts, ranked and banded
- Read the reply where it named them, with the prompt and the persona that asked
- Take the ranking out as CSV and the scorecard out as PDF
Runtime: 5:55 | Difficulty: Advanced
Are we losing ground? You cannot answer the board without measuring the rival the way you measure yourself, same AI model, same questions, so the two numbers sit on one axis. The job goes wrong the moment somebody runs the second audit slightly differently, because then the gap you report upstairs came out of the method and the market never moved.
What it costs today
Assumption: an analyst at EUR 90 an hour, running the same seven-hour audit round you already run for your own brand.
| Line item | Assumption | Arithmetic | Cost |
|---|---|---|---|
| One audit round, one rival, all three models | 7 h at EUR 90 | 7 × 90 | EUR 630 |
| The like-for-like round, one shared model | 2.33 h at EUR 90 | 2.33 × 90 | EUR 210 |
| Two rivals, tracked monthly, all models | 2 × 630 | EUR 1,260 per month | |
| The same, over a year | 12 × 1,260 | EUR 15,120 per year | |
| Strategist plan | EUR 496 per month |
Three brands per workspace is the ceiling, so your brand and two rivals. EUR 210 is the honest like-for-like comparison on one shared model, EUR 630 the full audit somebody already runs. Either way it holds only while the same person uses the same prompts every time, and that is what a plan fixes and a spreadsheet does not.
What the video shows
- A rival becomes a project of its own: project name, website URL, then two ticks, "Run scan for this project" and "Competitor's website".
- The second tick marks it as somebody else's brand, so it skips the onboarding questionnaire and stays out of your report picker.
- The crawl starts. It reads the website, builds personas, writes the prompts and puts them to the model one at a time.
- Back on your own project, the "Your Visibility through time" card carries a "Compare with:" picker, with competitors at the top of the list.
- Pick the rival and its history draws in as a second line on the same 0 to 100 axis, from two independent crawls.
- The metric picker switches both lines together: Rankfor Index, Consistency, Coverage, Source Quality, Stability, Recommendation Share, Sentiment, Awareness.
- The window toggle moves both lines across three months, six months or a year. Every point on either line is one scan.
- Personas, then AI Visibility, then the Brand Recall card, where the count of competitors the model named opens into a ranked grid banded dominant, strong, moderate and minor.
- The prompts view: every prompt, the persona that asked it, the model's full reply, and the brands detected under each one.
- Discovery Lab widens the question. It writes its own questions across your topic field, then ranks every brand it found by frequency, mentions and share.
- Export: CSV out of Discovery Lab for the ranking, PDF out of AI Visibility for the scorecard, then a closing frame on your line and the rival's.
What you need
- Strategist, EUR 496 a month, or Enterprise. Free and Marketer cap a workspace at one project, so a competitor project cannot exist below Strategist.
- The rival's public domain.
- Both projects scanned with the same AI model, Gemini by default, and one scan per model per calendar month per project, so count what the month has left.
- More than one scan on each line inside the window you pick. One point is not a trend.
- Personas left alone between the scan and the AI Visibility screen, because editing one resets its coverage.
Questions people ask
How many rivals can I track? Two. A workspace holds three projects on Strategist and Enterprise, so your brand and two competitors. Nothing lifts that ceiling.
How often does the line move? Once per scan, and you get one scan per model per calendar month per project. Rankfor.AI has no monitoring and no alerting. You scan, and the point lands.
Why is the rival dimmed in the picker? The second line draws only when both projects hold history on the model you selected. A project nobody has scanned reads "Not scanned yet", one measured elsewhere "Only scanned on ...". Scan both on Gemini.
Can I hand the comparison over as a generated report? No. Competitor projects stay out of the report generator by design, and the chart card has no export. Take CSV out of Discovery Lab, PDF out of AI Visibility, and screenshot the chart.
Method: two projects in one workspace, both scanned on Gemini, compared over three months, August 2026. Add the rival as a project this month, and next month the chart answers the board question for you.