Promises rather than tasks

Task management seems perfectly natural. Something needs to be done. You break it down. You assign it to someone. You set a date. You track progress.

The method works well when the work is repetitive, perfectly understood and stable enough to be described correctly in advance. It works far less well as soon as the work carries a significant share of uncertainty. Which is precisely the case for much of what we do.

To assign a complex task correctly, the person assigning it must know in advance its content, its dependencies, the likely difficulties and the real capacity of the person receiving it. In other words, they must often hold knowledge that no one holds yet.

The organisation then fills the gaps with assumptions. A duration estimated very early goes into a schedule. The schedule becomes a budget forecast. The forecast becomes a client milestone. Then, a few weeks later, an assumption made before the problem was understood is treated as a commitment someone had personally taken.

We want to break that chain.

The commitment replaces the task

At Cercle, the fundamental unit of work is not the task. It is the commitment: a request, a promise, a delivery, an acceptance.

A commitment begins with a request. Someone needs a result. The person able to produce that result examines the request, checks that they understand well enough what is expected and decides whether they can reasonably promise it. They then deliver the result. The person waiting for it accepts it or points out what is missing.

The difference with a task is small in words. It is considerable in practice. A task can be allocated to you without you ever having stated that you could carry it out under the planned conditions. A promise requires your agreement.

It also means that no one can commit your time on the organisation’s behalf simply by assuming you will find a way.

Knowing, estimating, promising: three different things

An estimate remains useful. It helps us reason. It can serve to build a scenario or grasp an order of magnitude. But an estimate is not a promise. And we do not ask for artificial precision at a moment when no one yet holds the information needed to produce it.

A precise estimate is only asked for once the work is understood well enough for it to mean something. It only becomes a promise when someone decides to turn it into a commitment.

As the work progresses, uncertainty shrinks. A dependency is delivered. A prototype reveals a constraint. An option is eliminated. The scope becomes clearer. That is when a precise commitment starts to make sense.

This logic can feel less reassuring than a big schedule filled in months ahead. Above all, it is more honest. We prefer uncertainty explicitly acknowledged today to fictitious precision that will become tomorrow’s delay.

When the impossible appears, it must become visible

A commitment can become impossible to keep. It will happen. A dependency can break. An assumption can prove wrong. A supplier can fail.

The real problem is not that someone discovers the original date no longer holds. The problem is discovering it and carrying on as if nothing had happened.

When a promise becomes untenable, we arbitrate between the variables actually available: the scope, the date, the resources, sometimes the objective itself. The rule also applies to commitments made by the founders.

A system in which employees can theoretically flag the impossible, except when the date comes from leadership, is not an honest system.

Sales follow exactly the same rule

The principle becomes particularly important when a client is involved.

In many companies, the sales organisation negotiates a scope and a date. The contract is signed. Then the team in charge of delivery discovers the promise it now has to keep.

The conflict is almost built into the system. The salesperson receives the benefit of the sale. The people doing the work then carry the cost of any overpromise.

At Cercle, the salesperson carries the client’s request, never the promise. The scope, the date and the conditions of success must be accepted by the person who will carry the responsibility for delivery. Without that promise, we do not sign.

It means someone can say: “I cannot promise that under these conditions.”

The answer is not automatically to ask them to try harder. We can reduce the scope. Move the date. Add resources if that genuinely changes the situation. Or walk away from the deal.

Losing a sale we could not honestly deliver is not poor performance. Signing up a debt in the hope that someone will absorb it later is far more so.

What everyone gains from it

This system offers the teams something very concrete. No one should wake up one morning to discover that someone else has sold their December.

It also offers something to the client. The person promising them a result actually knows what they are promising.

And it finally offers something to the company. An order book made of tenable promises is worth more than a sales volume artificially inflated by commitments no one yet knows how to fulfil.