A second view before an important technology decision.
The useful time for an independent view is before the decision becomes difficult to reverse. Afterwards it is an inquest.
This is the smallest useful piece of work I do. One decision, examined properly, by somebody with no stake in which way it goes.
- A significant budget is about to be committed.
- A strategic supplier is about to be selected.
- An AI programme is waiting for approval.
- A major system is being replaced, or kept for another cycle.
- An architecture choice would be expensive to unwind.
- A critical capability is about to be outsourced.
- Sensitive data is about to move onto a new platform.
- The board has asked for something nobody internally can answer neutrally.
Why the moment matters
A technology decision of any size arrives with confident answers already attached. The supplier has one and a reason to have it. The internal team has one and usually a good reason. The consultancy has a programme. Each is partly right, and none of them approves the spend or lives inside the result.
What makes these decisions unusually unforgiving is the asymmetry of reversal. Choosing wrongly is rarely fatal on the day. It becomes expensive eighteen months later, when data, process and habit have settled into the choice and unpicking it costs more than the original purchase. The window where an independent view is cheap is narrow, and it closes quietly.
A second view is not a veto and it is not a procurement exercise. It is one person with no stake in the outcome, reading the same material as the board and saying what it actually commits you to.
What is actually being decided
Often the stated decision is a proxy for a different one that nobody has written down.
What it assumes
Every proposal rests on assumptions about volume, cost, capability and behaviour. Most are reasonable. Some are load-bearing and untested.
What the alternatives are
Including the two that rarely make it into a board paper: do the smaller version, and do nothing yet.
What it costs to reverse
Not the purchase price. The cost of being wrong, eighteen months in, with data and process already inside it.
Who benefits from the recommendation
Suppliers, internal teams and consultancies all have legitimate positions and none of them is neutral. Naming the incentive is not cynicism.
What the organisation is ready for
Capability, data quality and operating model decide whether a good decision survives contact with the place it lands in.
- A written view of the decision: what it commits you to, what it depends on, and what would have to be true for it to work.
- The risks worth raising at board level, separated from the ones that are somebody's job to manage.
- A recommendation, or the criteria to decide by if the answer genuinely depends on something you know and I do not.
No fixed template. A decision that turns on one supplier contract does not need the same document as one that turns on whether the data is good enough.
Why me for this
Nine anonymised problems on the Work page show the reasoning, including a legacy estate retired on a deadline and a marketplace model that removed a costly step from a real customer process.
Read the ten casesRelated
The useful time for an independent view is before the decision becomes difficult to reverse.
Two or three lines are enough: what is happening, what decision is coming, and when it matters. You do not need to know which kind of engagement this is. LinkedIn is the single route in, on purpose.