GrowMarket.AI screens

GrowMarket.AI
weebpal.com
The public overview explains that campaigns are written, scheduled and posted automatically while every claim still traces back to a fact the owner signed. A published list of things the system will never do, such as inventing a number or publishing without an approval record, that no setting can switch off. Guided setup walks through website import, business context, brand kit, channels and mandate so the system learns what it cannot safely guess. A daily queue shows only the work that needs an owner decision, so approvals, warnings and stopped publishing all surface in one place. Real photos and video are uploaded with an explicit rights confirmation, so the material used in posts is genuinely owned by the business. The material library separates uploaded photographs, public product screenshots and AI-made graphics so every post shows where its image came from. The truth profile holds the verified facts, boundaries and destinations every campaign must respect, and marks which claims cannot be used yet. A single claim is reviewed and signed by the owner with attached evidence, because a website source is not the same as owner verification. Forbidden lines record statements the content must never publish, and they are checked before every generation and again before every send. The brand kit stores colours, logo files, typeface and tone, and shows detected website changes so nothing is applied without an owner choosing it. A campaign is planned by stating the business intent, the time window, the exact destination accounts and what each post is allowed to use. Before any copy is written, a pre-scan forces every uncertain claim to be confirmed, corrected or excluded, holding generation until it is resolved. The campaign calendar lays out every planned post by date and destination, and moving a slot changes only the timing, not the approved content. A single post shows its approved version, its exact destination account and the recorded publication evidence, with no cross-channel success inferred. Each post is approved one at a time alongside pre-flight evidence showing which verified facts it used and which guardrails it passed. Translated versions are reviewed against the approved source so a language edit changes only that derivative and never the original approval. Connected destinations are listed by channel type with their capabilities and send route, so each account is authorised independently of the others. The posting mandate bounds automatic sending to named accounts, a capped number of sends and an end date, and it can be revoked at any time. For destinations the system may not send to, the approved copy, graphic and link are packaged for the owner to publish and record the public URL. Results separate what actually went out from what came back, keeping tracked events and platform figures apart instead of totalling them together. The campaign report closes the loop with the publication record, the measured response, the stated limits and one single change to test next. Publishing can be stopped for one post, one campaign or the whole workspace, and nothing restarts until an owner decides what happens to held work.