How tracking works
One budget of AI answers a month for your workspace, shared across your brands, and how daily, weekly, twice-weekly and mixed checks spend it.
Every paid plan gives your workspace one budget of AI answers a month, shared across all its brands. Each time an engine answers one of your questions, that is one answer, whichever engine it was. You decide how each brand spends its share.
The modes#
For each brand you pick one of these in Tracking. Each comes with the reason to choose it:
| Mode | Why | Solo | Starter | Growth | Agency Starter |
|---|---|---|---|---|---|
| Mix (recommended) | Fast where it matters, steady everywhere else. | 3 daily + 16 weekly | 5 daily + 46 weekly | 20 daily + 184 weekly | 50 daily + 576 weekly |
| Daily | Catch changes the day they happen. | about 10 | about 25 | about 100 | about 300 |
| Weekly × 3 answers | AI answers vary from ask to ask; three answers give a steadier number. | up to 23 | up to 57 | up to 230 | up to 692 |
| Twice weekly × 3 answers | Steady numbers that still move quickly. | 11 | 28 | 115 | 346 |
The counts above use a plan's whole budget on the five included engines. A workspace with several brands splits the budget between them (evenly by default; you can give a brand more).
- Mix (recommended) checks your most important questions daily and the rest weekly with 3 answers. This is what every brand starts with.
- Daily gives each question 1 answer per engine every day. Your numbers and Proof count the last 7 days, so a daily question still rests on up to 7 answers per engine.
- Weekly asks each question once a week, with 1, 2 or 3 answers per engine.
- Twice weekly asks each question twice a week, with 3 answers per engine.
Your question limit#
Your plan's question limit is how many questions its answers cover weekly with 3 answers on all five engines. Fewer answers per question, or fewer engines, never raise it: they leave answers for more daily or twice-weekly checks instead. If a change means some questions no longer fit, they stay saved and are checked again once there is room. Nothing is deleted.
Engines#
Every plan includes ChatGPT, Google AI Overviews, Google AI Mode, Gemini and Perplexity. Turning engines off for a brand means each question uses fewer answers, so the same budget covers more daily questions.
Claude is the Claude add-on on every plan, in two tiers. Claude on the questions you choose gives the workspace a Claude answer budget each month: 130 (Solo), 325 (Starter), 1,300 (Growth) or 3,900 (Agency Starter) Claude answers, each with up to 2 web searches, shared by your brands. You choose the questions and how often: daily (1 Claude answer a day, through the daily check) or weekly × 1, × 2 or × 3 (on the weekly runs; a twice-weekly brand gets Claude on every second run). On Growth that is about 100 weekly × 3, 300 weekly × 1, or 42 daily. Until you choose, Prefera picks your most important questions weekly × 3: the questions you check daily or watch first, then the highest priority. Claude on every question answers every question you track on every run, in each brand's tracking mode. Both come on top of your AI answers. Turn one on in Team and billing.
Setting it up#
Right after you pick a plan, onboarding shows Set up tracking with the recommended setup and how many questions it covers. You can change it there, or skip and keep the recommended setup. You can change it any time in Tracking: the screen shows how many questions each choice allows before you save.
When the answers run out#
Your plan's AI answers are a hard cap: a run never goes past them. When a scheduled run would, it checks only your most important questions that still fit; Run now says how many answers are left instead of starting; daily checks pause until your next billing period starts, and a banner says so. Proof still re-checks a fix you published, so it is measured. Need more? Add Extra AI answers in Team and billing, or check fewer questions daily. See Plans, add-ons and billing.