What I ask before we take on a project — Achintya Verma

When someone writes to us, the first thing I want to work out is not what we could sell them. It is whether we are the right company for the problem. Sometimes we are not, and it is better for both sides to find that out in the first reply.

These are the questions I come back to.

01What is the problem, in your own words?

Not the solution you have in mind. The problem. "We need a chatbot" is a solution. "Customers wait two days for an answer" is a problem, and it may have a better answer than a chatbot.

02What does it cost you today?

If the cost cannot be described, the benefit will not be measurable afterwards, and neither of us will know whether the work succeeded. A rough answer is fine. No answer is a sign the problem needs more thought before anything is built.

03Where does the data live, and who may see it?

This decides the design more than any feature does. Whether data may leave your network, who is allowed to see what, and which system is the source of truth shape everything that follows.

04Who will own this on your side?

A project without a person who wants the result and can make decisions will stall, however well it is built. I would rather wait for that person than start without them.

05What is the smallest thing that would prove it?

I am wary of beginning with the whole vision. I want the first stage small enough to finish and clear enough to judge. If it works, the next stage argues for itself.

06When we say no

When someone wants a foundation model built or licensed.
When the need is a consumer app sold by subscription.
When what is wanted is engineers by the hour.
When the programme needs a thousand people.
When we do not think it will work, and saying so is the useful thing we can offer.

We will say honestly whether we can help, and what it would take.

In short

Start from the problem, not the solution someone has in mind.
A cost that can be described makes a result that can be measured.
Where data lives and who owns the project shape everything.
A small first stage, and an honest no where the fit is wrong.

Questions

Does Quantum Beetle turn work down?

Yes, when it is not the right kind of company for the problem, and it says so in its first reply.

What should I include in a first message?

The problem in your own words, what it costs you today, and where your data lives. A few lines is enough.

Why does he insist on a small first stage?

So that the work can be finished and judged quickly, and the next stage is decided on evidence.

Sounds like your problem?

Tell us about it. We'll say honestly whether the swarm can help, and what it would take.

Talk to us hello@quantumbeetle.ai