Skip to contentAnemo
EN
Contact

How to Scope an R&D Project So Your Team Stays Out of It

· 4 min read

Scoping is where most capacity-saving arrangements quietly fail. The discovery phase runs on workshops, and workshops run on your senior engineers' calendars, so the exercise meant to free your team spends three weeks of it before any code is written.

It does not have to work that way. A competent external team can scope from artefacts you already have, plus one day of questions. What follows is what that requires from you, and what it gets you.

Key takeaways

What an external team can read without you

Most of what a workshop extracts is already written down somewhere:

Giving read access to these takes an hour of someone's time. Explaining the same material in meetings takes weeks and produces a worse picture, because it is filtered through what people remember to mention.

What you do have to supply

The business outcome

One or two sentences, written by you, describing the change you want in the business rather than the software you want built. This cannot be derived from the schema and nobody else can write it.

One day of questions

A block of time with the one or two people who know the system best. Real questions from people who have already read the material, not an introduction.

The test of whether an external team is doing this properly: the questions should be specific and slightly uncomfortable. "Why does the orders table have two status columns" is a team that has read your system. "Can you walk us through your business" is a team that wants you to do the work.

A decision-maker for the boundary

Someone who can answer interface questions within a day. This is the only ongoing commitment, and it is hours a month.

Your constraints

The things that are not negotiable: a regulatory requirement, a platform you must integrate with, a date that exists for a real business reason, a technology you have standardised on. Say these at the start. Discovering them at week six is how scope gets redone.

Decide the interface, not the architecture

The question that actually needs answering is not "how should this be built". It is "where does this new thing end and our system begin".

Settle four things and the rest can follow later:

  1. What the new product may read, and through what route.
  2. What it may write, if anything.
  3. What it must never do. Usually no schema changes, no writes to core tables, no heavy queries against production.
  4. Who decides when either side needs to change.

This is a few hours of work between two technical people. It is also the thing most often deferred, because it feels like a detail next to the architecture discussion, and it is the only part that has to be agreed before work can start.

What a good scope output looks like

You should be able to refuse it. That means it contains specifics: what is in the first release, what is explicitly excluded, what assumptions it rests on, what would change the estimate, and a range rather than a single date.

A scope document that could describe any project is not a scope, it is a proposal. The useful version names things about your business that would be wrong if applied to someone else's.

It should also state what the external team still does not know, and what would resolve it. That list is a better indicator of quality than the confident parts.

Frequently asked questions

Is a paid discovery phase worth it?

Often, if it produces something you own and can take elsewhere: a scope, an interface definition, an estimate with assumptions. It is not worth it if the output is a proposal from the same supplier, which is a sales document you have paid for.

What if our system is badly documented?

That is the normal case, and it is why reading the code and the history beats reading the documentation. A team that cannot work without good documentation will struggle throughout the project, not just during scoping.

How long should scoping take?

One to three weeks for a typical business product, from access to a written scope. Substantially longer usually means the boundary question has not been settled and the exercise has expanded to fill the time.

Can we scope it ourselves and then get quotes?

Yes, and it gives you comparable proposals, which is genuinely valuable. The risk is scoping around what you already know how to describe and missing the part that turns out to be hard. If you want a second read on a scope before you send it out, get in touch.

How we would work on this

Related services

Related reading

Building the product for what comes next

We would rather deliver one product that holds up than three that have to be rebuilt. That standard is the same on every project, whatever its size.

Ali Boran GazelCEO

Contact us