HookGet Open dashboard

Every brand, one screen

Run more than one brand, site or client? Each one is an isolated HookGet project — and the executive view reads across all of them: revenue, ad spend, leads, calls and AI cost per brand, totalled for the org, with the anomalies the engine flagged. The numbers come from the same verified events the per-brand dashboards read.

In short

  • One row per brand: revenue, ad spend, leads, inbound calls and AI cost over 30 days.
  • Org totals and an AI-spend run-rate — labelled run-rate, because that is what it is.
  • Recent pattern breaks across every brand, from the deterministic anomaly engine.
  • Brands stay isolated: separate projects, separate keys, separate dashboards. The executive view is a read, not a merge.
  • Session-only by design — no API key can read across brands, including your assistant’s.

Who reads it

You areWhat the screen answers
A CEO with several brandswhich brand earned, which brand burned — this month, on one screen
An agencyevery client’s funnel health without opening ten dashboards
A holding companyorg-wide revenue and spend, with the anomaly list as the morning briefing
A solo operator with two sitesboth sites, and the AI bill for each, side by side

What it deliberately is not

Not a data merge: events never cross projects, and a per-brand dashboard can never show a sibling brand’s numbers. Not a forecast engine: the run-rate line is arithmetic over the same 30-day window, labelled as such. And not an MCP surface — see the FAQ for why that is a security decision, not a gap.

Questions

What counts as a "brand"?

A project. Every HookGet project is an isolated tenant with its own sources, keys and dashboards — so a holding company runs one project per brand, an agency one per client, a solo operator one per site. The executive view is the org-level read across all of them: same numbers the per-brand dashboards show, summed side by side.

Can my assistant read the executive view over MCP?

Deliberately not. An MCP key belongs to one project, and a project key that could read its sibling brands would be a cross-tenant read with extra steps. The org-level screen authenticates the way humans do — with your signed-in session. Per-brand, your assistant can do everything.

Where do the anomaly flags come from?

The deterministic engine that watches every project: each hour is judged against the same hour of the previous 14 days using median and MAD — robust statistics, no model, no per-alert cost. The executive view lists the recent flags across all brands, so the one screen a CEO opens is also where pattern breaks surface.