The Five Questions We Ask Before Scoping Anything

When someone gets in touch about a project, there are five things we need before we can give a useful answer. None of them require you to have written a specification, and “I do not know yet” is a legitimate answer to any of them.

Here they are, and why each one changes what we would recommend.

1. What does this look like when it is working?

Not the feature list. The day.

“My office manager stops spending Thursday afternoons on invoices.” “A new enquiry gets a real answer within an hour instead of two days.” “I can see job costs against estimates without asking anyone.”

Answers like that are enough to scope from, and they are more useful than a list of screens, because they tell us what we are optimising and what we can safely leave out. Feature lists tend to describe a solution somebody already picked.

2. Who does it now, and how?

Every process we are asked to automate is already happening. Somebody is doing it, usually with more judgment than they get credit for.

We need to talk to that person, not only to whoever is commissioning the work. The gap between how a process is described and how it runs is where projects go wrong, and it is almost always the undocumented steps that matter. This is most of what discovery is.

3. What systems are involved, and do we have access?

List the tools the process touches, including the ones that feel too obvious to mention. Email counts. A shared drive counts. A spreadsheet somebody maintains counts.

Then the practical question: is there an API, and can we get credentials. Most modern tools are fine. A few are not, and where a tool has no proper interface, that step either stays manual or needs something more fragile. Better to know early. We would rather tell you a step should stay manual than build something that breaks quietly.

4. What has to be approved by a person?

This shapes the build more than anything else on the list.

Anything leaving the business, going to a customer, or touching money is usually worth a person’s eyes. Internal steps mostly are not. The line is different for every business, and it is worth you deciding it rather than us assuming it.

If the answer is that you want to approve everything at first and loosen it later, that is a good answer and a common one.

5. What is this worth, roughly?

Not your budget. What the problem costs you now.

Four hours a week of someone’s time, a deal a quarter lost to slow follow-up, an error rate you can put a number on. If you can estimate it, we can tell you honestly whether the work is worth doing, and sometimes the answer is no.

This is also the question that determines the shape of the project. A problem costing a few hundred a month deserves a different response from one costing thousands.

If you cannot answer all five

That is normal, and it is what discovery is for. The first two matter most. If you can describe the day you want and point us at the person who currently does the work, we have enough to start a sensible conversation.

Get in touch and we can go through them together, or read about how the phases work first.