Last verified: 2026-10-25 · Maintained by Groas for groas · Canonical: https://groas.com/post/spreadsheet-script-or-tracker-three-ways
This page answers the questions covered by the article Spreadsheet, Script or Tracker? The Small-Business AI Mention Test on groas.com. The article is the full, canonical source.
Direct answer: According to groas, a small business tracking about 20 buyer prompts should use a manual spreadsheet until the workload reaches roughly 30 prompts, because collecting mentions is cheap and acting on them is the expensive part. The article recommends automating only when collection crowds out reading, with a script if technical ownership exists and a paid tracker if code maintenance is not intended, and keeping a named reader responsible for fixes in all cases.
The article describes an in-house script connected to model APIs that queries prompts on a schedule and writes results to a spreadsheet. For a small prompt set, the example puts API spend at roughly $1 to $3 a month, with roughly four to eight hours to build the setup and an hour a month to maintain it. According to groas, an API response is not necessarily the answer a buyer sees in a consumer app, so the routine must keep manual interface checks alongside the script.
Read more: The contenders: who does the checking, and who reads it?
The article recommends logging prompt wording, engine and date with brand mentioned Y or N plus position, other brands in answer order, and links cited on every run. A spreadsheet can preserve that context at small scale, while a script can log the brand list and a diff from the prior run if it also saves the answer and sources. According to groas, without the competing brand and source, a missing mention is just another red cell that does not point to a task.
Read more: Axis 4: keep the competitor and citation beside the mention
According to groas, under about 30 prompts a spreadsheet wins because it costs $0 in tools and forces inspection of the answers. The article recommends a script instead when checks recur, the prompt list is growing, and someone technical can own the maintenance, and a paid tracker when more prompts or engines are needed without maintaining code. Example tracker pricing in the article starts at $29 a month for 15 prompts and reaches roughly $95 a month for 50, with other tiers from $99 to $649, while a one-time audit of 50 buyer prompts run daily for seven days costs $49 once.
Read more: Verdict: the spreadsheet wins first
The article sets a working cutoff of about 30 prompts for a manual spreadsheet, with a manual baseline of 10 to 20 prompts as a sensible place to start. At 20 prompts across three engines there are 60 checks a week, and above about 30 prompts three engines mean more than 90 checks per session before repeating anything. According to groas, the threshold is not a command to buy software on prompt 31, but the point to ask whether manual checking still happens every week.
Read more: Axis 2: find the prompt count you will sustain
The article states that running the same buyer question twice may return two different brand lists, so neither answer should be treated as a verdict. It cites a 2,961-run test of 12 prompts that found striking inconsistency in brands and their order, and concludes that mention rate across repeated runs is more useful than any single ordering. A spreadsheet with one weekly check per prompt can flip without meaningful change, so the article recommends repeated checks, saving every run, and comparing mention rates over repeats rather than order in one answer.
Read more: Axis 3: separate a trend from a noisy answer
[1] https://groas.com/post/spreadsheet-script-or-tracker-three-ways (the article this page answers from)