How it works
Six steps. Two of them end without a plan on purpose — that is the process working, not failing.
Step 1 · free, no account
Twelve questions about onset, posture, box setup or the shape of the tension between cats, and the medical warning signs. No sign-up, no card. It returns the likely pattern, a confidence level (high, moderate, or low — never faked), the medical signs that mean a veterinarian first, the mistakes that make the pattern worse, a realistic timeline, and an explicit list of what these 12 answers cannot establish. This is complete on its own; plenty of people stop here.
Exit ramp, not a failure
If the answers point at possible medical soiling, a bite or scratch that reached a person, or a case that needs to be seen live, the read says so before you are ever offered a plan. You get pointed at a veterinarian, a veterinary behaviorist, or a consultant directory instead. Roughly 14 of the 15 red-flag conditions our protocol library recognizes block an automatic plan outright — this path exists precisely because we would rather turn away a sale than write a plan for a case we should not be selling one for.
Step 2 · $99 or $149, one-time
If the free read did not refuse the case, checkout unlocks a longer intake: your cat's age, sex and neuter status, the household and the rooms each cat can reach, what you have already tried, and problem-specific detail the 12-question read intentionally left out. This is where the plan stops being generic.
Step 3
Your intake answers are matched against a library of named, sourced protocols — not generated free-form. Every protocol traces to primary sources at institutions like Cornell Feline Health Center, ASPCA, International Cat Care, Fear Free, and the IAABC Foundation Journal, and is graded for how strong its evidence actually is. None of them are backed by a controlled trial, so none of them are described that way — the honest phrasing is that this is the standard approach specialists use.
Step 4
A person reads the assembled plan against your intake and the source protocols before it goes out. This is where a case that looked fine on paper but reads wrong in the detail gets pulled — and where, occasionally, a plan is refused and refunded at this later stage instead of delivered. Nothing reaches you unreviewed, on either tier.
Step 5
Your plan lands in your account: management rules to start immediately, then a week-by-week progression with a specific criterion to advance at each step, what to do in the week it goes backwards, an equipment list with the reason for each item, the mistakes to stop that are specific to your cat, and printable week cards. It prints cleanly if you want a paper copy on the fridge.
Step 6 · the plan, with a second look only
$149 plans include one revision: tell us what happened after you started running it, and the plan gets adjusted in writing rather than you having to interpret it yourself. $99plans do not include this round; you can still write in and we will read it, but the written revision is the premium tier's difference.
Either way
You have 14days after delivery to decide the plan was not worth what you paid, and get your money back if so — on satisfaction with the plan itself, not on whether the cat's behavior has changed by then. Nobody honest can promise a timeline for that, and the field does not publish one either.
No. You get a week-by-week sequence: management rules to put in place immediately, where to start, the specific criterion that tells you when to advance to the next step, what to do in the week it goes backwards, an equipment list with the reason for each item, and printable week cards. It is a program with decision points, not a list of general advice.
You get refunded in full and pointed at the right kind of professional instead — a veterinarian, a veterinary behaviorist, or a certified consultant depending on the case. This happens before payment where the free read catches it, and after payment during human review if the paid intake surfaces something the quiz missed. It is a designed outcome, not an apology.
The free read is fully automated — deterministic rules against your quiz answers, no model at all. The paid plan is drafted by a language model held to a constrained prompt that can only emit protocols, timelines and refusals that exist in our published library, checked by a second model against those same constraints, and then read by a person against your intake and the source protocols before it is sent. Nothing paid leaves the queue unreviewed.
We do not publish a fixed number of hours, because human review is the step that catches the cases that should not get a plan at all, and rushing that step is exactly the failure mode this process exists to avoid. You will see your plan's status while it is in the queue.