
What Good Technology Due Diligence Looks Like
A clean architecture diagram tells you almost nothing about deal risk; the team, the delivery record, the debt and the key person dependencies decide what happens after close.
Blogs
Twenty articles drawn from our services, sectors and delivery experience. Each opens in a new window so you keep your place in the list.

A clean architecture diagram tells you almost nothing about deal risk; the team, the delivery record, the debt and the key person dependencies decide what happens after close.

Training completion rates measure attendance, not behaviour change; adoption has to be engineered through incentives and capability, then measured in how people actually work.

Standups and sprint reviews are not evidence of maturity; flow, quality and predictability are, and most teams doing the rituals have never measured any of them.

AI contracts written for convenience today often become the reason switching costs a fortune tomorrow; portability and exit rights need negotiating up front.

Most benefits cases collapse under scrutiny; the fix is disciplined baselines, honest attribution and a named owner who signs.

Getting an e-commerce platform ready for peak trading is a rehearsed operational discipline, not a one-off load test.

Spend visibility only creates leverage at the negotiating table when taxonomy and supplier master data are trustworthy.

Modernising core payments infrastructure to a regulatory deadline requires resilience thinking, not just a migration plan.

Digitising paper records only pays off when the process around records is redesigned, not merely copied into a screen.

Standards compliance is not the same as a workflow that clinicians trust and use without friction.

A well-designed digital front door for public services only works if the back office truth behind it, and the assisted digital route beside it, are designed with equal care.

FinOps dashboards can highlight overspend but they cannot fix it, because cloud cost is set at design time by data gravity, coupling and tenancy choices.

The most valuable business analysts close decisions rather than merely documenting requirements, and that shift changes how the role should be measured.

Stakeholder confidence is a delivery asset that erodes long before a programme visibly fails, and it can be rebuilt with honest reporting rather than reassurance.

Proportionate governance lets generative AI move fast in low risk areas and slows down deliberately where the stakes are real.

Bad data is rarely a data problem, it is a symptom of an operational process nobody has fixed.

The organisations that modernise successfully rarely do it through a single dramatic replacement.

Most target operating models are diagrams that describe structure but never make a single decision.

Every transformation programme eventually meets a ceiling built from years of unmanaged system integrations.

Most organisations have too many AI ideas and not enough discipline to say no to the wrong ones.