We ran our own portfolio brand through the AI-visibility audit. It scored zero, four times. Here is the result, and how we checked it was real before we believed it.
BeAligned is a co-parenting reflection platform built by a solo founder. It has real users, a real product, and a real practitioner behind it.
Ask an AI answer engine to recommend a co-parenting app, and it does not exist. Not ranked low. Not named. Across 18 probes in a single run — every one of which returned a parseable recommendation list — the engines named BeAligned zero times.
That is the measurement that matters, and it is worth being precise about why. Eighteen of eighteen probes came back with a usable list, which means the engines had every opportunity to answer the question. They were not confused, rate limited, or silent. They produced recommendation lists eighteen times out of eighteen, and BeAligned was in none of them.
Absence measured under full opportunity is a finding. Absence under a broken probe is a bug. We checked which one this was before we wrote a word of it.
| Signal | Value |
|---|---|
| Times named by an answer engine | 0 — all four runs |
| Probes returning a usable recommendation list | 18 of 18 |
| Competitors the engines did name | 29 — this run; see the footnote |
| Composite visibility score | 0 |
| Channels lit | 2 |
Run gfa_bealigned-ai_1, 2026-08-06, engine build 03tjqq5. Engines queried: Gemini, ChatGPT, Perplexity.
TalkingParents was named in 16 of the 18 answers, with AppClose and OurFamilyWizard tied behind it at 13 each.
The honest footnote on that count. Answer engines are non-deterministic. Four runs of this same audit returned 33, 31, 33 and 29 competitors. The same handful of names comes back every time and TalkingParents leads every time — but the second and third places trade, and the exact count moves by several. We report the run id and the date with every number because a bare “29 competitors” implies a precision the instrument does not have. Anyone selling you a single unqualified number out of a non-deterministic system is either not measuring repeatedly, or not telling you that they are.
There was a version of this result that would have killed the story: that the zero was an artifact of our own naming detector failing to match the string “BeAligned” — in which case the brand might be perfectly visible and our instrument simply blind to it.
That is the version we most wanted to rule out, because it is the version where the case study does not exist. So it was checked before anything was written, not after. The signature of that failure is specific — citations found for a brand that no probe reports naming — and it appears in none of the four runs. The detector is behaving. The zero is real.
We ran the confirmation before deciding how to frame it, because a result that does not reproduce changes whether there is a story at all — not merely what number the story reports. Reversing that order is how organisations end up framing findings they have not yet confirmed.
If you are reading this because you want to know whether our audit is honest: this is our own portfolio brand, and we published its zero.
We are not claiming BeAligned is invisible because it is bad. It is invisible because being good is not the input. Answer engines recommend what they can retrieve, parse, and corroborate. A product can be excellent, in-market, and genuinely helping people, and still be structurally unavailable to the systems an increasing share of people now ask first.
That gap is measurable. We measured ours. We will measure yours the same way — and if the number is zero, we will tell you it is zero.
Figures on this page come from a single named run and are not continuously updated. Because the engines are non-deterministic, a re-run will move the competitor count and may reorder the names below the leader; the zero has held across every run we have taken. Published under our editorial policy.
