Skip to content

Demo content

Clients, quotes, figures, people and awards on this build are invented for review. Nothing here is real yet.

Free tool

Is your business actually ready for AI marketing?

25 questions across five dimensions, scored on this page. It takes about four minutes, it is capable of telling you the answer is no, and roughly half the businesses that ask us for this work should hear exactly that.

  • No email address required
  • Nothing sent anywhere, ever
  • Result appears immediately
0 of 25 answered—

The task

0 of 5

Automation economics are entirely about repetition. A task nobody currently does at volume will not become one because you have bought a tool.

  • There is a specific task, done by a named person, more than twenty times a week.
  • We could describe that task as a set of steps someone new could follow.
  • The task takes roughly the same shape every time, rather than turning on judgement.
  • We know approximately how many hours a week it currently consumes.
  • If that task disappeared tomorrow, the time saved would go somewhere useful.

The data

0 of 5

Most automation projects that fail do not fail at the model. They fail because the information underneath was worse than anyone admitted.

  • The information this task needs already exists in a system, not in somebody’s head.
  • That data is accurate enough that we act on it today without checking it first.
  • We could export it ourselves without asking a supplier for permission.
  • Records are consistent: the same thing is named the same way throughout.
  • We know where the data came from and who is responsible for keeping it right.

People and control

0 of 5

Every process has somebody who informally catches the mistakes. If you cannot name them, automation will not remove that person, it will remove the catching.

  • Someone would own this after launch, by name, as an explicit part of their job.
  • We can say who currently catches the mistakes in this process.
  • A person would review the output before any customer sees it.
  • At least two people would understand how it works, not one.
  • The team affected has been told this is being considered.

Systems

0 of 5

The build is rarely the hard part. Getting two systems that were never meant to speak to each other to agree on what a customer is, that is the hard part.

  • The systems involved have APIs, or integrations that already exist.
  • We are not planning to replace any of these systems in the next six months.
  • We know what we currently pay for the tools in this area.
  • Somebody internally can make configuration changes without calling a supplier.
  • Security and data protection sign-off is achievable for this.

Intent

0 of 5

This is the section that ends most of our conversations, and it is the one worth answering honestly. Automating around an unresolved problem makes it fail faster and more expensively.

  • We can name the specific number this is supposed to move.
  • We would still do this if it saved time but did not reduce headcount.
  • If we built nothing, the first thing we would fix is not a broken process elsewhere.
  • This is not being done mainly because competitors appear to be doing it.
  • We have a budget in mind, and it is not “whatever it costs”.

Answer all twenty five to see a result

This checklist used to sit behind an email form as a PDF. It works better as something you can just use, and the reasoning is written up in the four questions we ask before agreeing that AI will help.

See the service