targetbay-onboarding

Use when onboarding a store nobody has assessed yet and the first question is what this store actually is — what TargetBay knows about it, what it already has configured across all three products, and which of those facts are measured rather than assumed. Reads the derived store context and reports what is derived, what is provisional, what is simply absent, and which questions only the store owner can answer. Produces no plan; establishes what a plan could honestly be built from.

Use when an onsite change needs proving rather than asserting — designing an A/B or multivariate test, deciding whether the traffic can resolve it at all, setting the stopping condition, or reading a result honestly. Also use when a test is being called early, when results look too good, or when several changes are being bundled into one test.

Use when deciding who should see an onsite offer, popup, banner or bar — the audience, the trigger, the frequency, the exclusions, and whether the offer should exist at all. Also use when popups are hurting the experience, when offers are being shown to people who already converted, or when discounting onsite is being considered.

Use when a store is new to TargetBay and the question is what to set up first across all three products — onboarding email, SMS, reviews, loyalty and onsite personalization as one sequence rather than four independent setups. Decides what each product contributes, what it must wait for and why, who owns contact with a customer when two products both want to reach them, and what observation would justify each next step. Produces a sequenced ninety-day plan with every borrowed value labelled; provisions nothing.

Use when onboarding a store whose setup depends on facts the platform cannot observe — what the store is trying to achieve, what discounting it can afford, how much work it can sustain, what its brand does and does not permit, which channels and markets are restricted, and when its real peaks are. Asks only the questions the derived store context could not already answer, and stores the answers back onto the store record so every later skill reads them as constraints instead of asking again.

Use when an approved onboarding blueprint needs to become real configuration in TargetBay — the segments, journeys, templates, review triggers, loyalty configuration and onsite placements it describes. Dry-runs every resource first and shows what would be created or changed and how many people each can reach, applies only what a human approved, verifies what was created against what was approved, and stops before anything can reach a real person.

Use when search on the store's site is underperforming — searches coming back with no results, high-volume queries that do not convert, synonyms and redirects that need setting, or search results that do not reflect what the store actually wants to sell. Answers "our searches return no results" and "why does site search find nothing?". Also use to mine search queries for catalogue and merchandising gaps. Use personalization-audit when search is only one suspect among several and nobody has named the problem.

Use for an open-ended assessment of a store's whole onsite experience — when someone asks what to do about personalization, what is wrong with our site, where to start, or what to fix first, without naming a specific problem. Use surface-inventory when the question is only what already exists. Scans surfaces, recommendations, offers, search and measurement together and ranks what is most worth fixing.

Use when deciding which product recommendations belong on which surface, which strategy should drive them, what must be excluded, and what a placement is actually worth. Also use when a recommendation widget shows irrelevant products, when engagement is high but revenue is not, or when a new placement is being considered.

Use when the question is what onsite personalization currently exists and what it is doing — which surfaces carry placements, offers or search configuration, what each one is worth, where attention is being spent for nothing, and which surfaces host a decision but carry no help. Answers "what's actually personalised on our site right now?" and "what have we already got running?". This is the assessment skill every other personalization skill routes through; use personalization-audit when the question is what to do next rather than what exists.