Slow technology decisions are a design problem, not a people one

When technology decisions stall, duplicate or pile up on senior desks, leaders often blame individuals or culture. The usual cause is design: nobody has defined who decides what. Authority is assumed rather than assigned. Make decision rights explicit, by type of decision and level, and delivery speeds up while conflict falls.

An escalation ladder is not a decision model

Most organisations have an escalation path: if a team cannot agree, the question goes up a level, and up again. That feels like governance. It is really a fallback for when nobody knows who should decide.

The cost is delay. Questions that a team could settle in a day wait weeks for a senior meeting. Senior leaders spend their time on choices below their level and have less time for the ones only they can make. Teams learn that it is safer to wait than to decide.

A ship's captain does not ask the harbourmaster how to steer at sea, and the harbourmaster does not ask the captain where boats may dock. Each knows the boundary of their authority. That boundary is the governance.

Three failures show up when authority is unclear

Decision vacuum. Nobody believes a decision is theirs, so nobody makes it. Work stalls, or an important choice gets made by default when someone simply goes ahead.

Decision collision. Two groups each believe a decision is theirs. They make conflicting choices, and the business pays for duplicated systems, rework and friction between teams.

Decision gravity. Decisions drift upwards because nobody below feels authorised. Senior leaders become a bottleneck, and the organisation moves at the speed of its busiest people.

Each failure has a visible cost: delay, duplicated spending or lost management time. All three come from the same root cause.

Map decisions by type and by level

The fix is to make decision rights explicit along two lines.

The first is the level of the decision. Some choices shape the whole business, such as major platforms, data strategy or large investments. Some affect one area or product. Some belong to a single team. Each level needs a named decision-maker.

The second is the type of decision: for example customer experience, data, security, integration or spending. Different types need different expertise and different people.

Put the two together and you get a simple map: for each type and level, who decides, who must be consulted and who is informed. Then look for the gaps. Where is no one named? Where are two groups both named? Where are decisions sitting at a higher level than they need to?

The map does not need to be perfect to help. A first version, agreed by the leadership team, removes most of the waiting and much of the conflict.

What to do

• List the last five technology decisions that took too long, and note where each one stalled.
• For your most frequent decision types, agree who decides at business, area and team level.
• Publish the result and review it after three months, when real decisions have tested it.

If decisions are slowing your business down, let's discuss who should decide what.