Every organisation past a certain age carries positions that were decided once and are now maintained as procedure: an approval threshold, a supplier arrangement, an escalation route, a gate in the investment process, a rule about what may be published and what may not. Each was a trade-off, and at the time it was very likely the right one. What changes is not the decision. It is the conditions the decision assumed.
Those conditions are rarely written down anywhere, so when they lapse, nothing announces it. The procedure continues, and the reasoning behind it gradually becomes something people repeat rather than something they hold. Asking about it from inside is expensive: the question "why do we still do this" is heard as a challenge to whoever introduced it, and the person who asks pays for that, whatever the answer turns out to be.
This is the work I do. It applies to medium and large businesses and to public bodies alike, and in the public sector it applies with more force, because the arrangements are older and the cost of appearing to question them is higher.
A question that can be re-opened at nobody's expense. An outsider absorbs the political cost of asking. That is a large part of what is actually being bought, and it is worth more the longer the arrangement has stood.
A decision that belongs to your own people. A conclusion brought in from outside tends to be resisted on its origin rather than on its merits, and a position that a team did not build is a position that team will not defend when defending it costs something. So the people who own it make the decision, in the room, on material drawn from their own landscape. My role is to make sure the option set is complete and that the options they reject are recorded together with the reasons for rejecting them. An unrecorded rejected option is next decade's unexplained rule.
A written record of the trade-off. What the current position assumes, which of those assumptions were tested, which of them still hold, who owns each one, and the date each should be tested again.
The position may well not move. A re-examination that ends in "we have looked at this, and we are keeping it, for these reasons" is a good outcome and worth having in writing: it converts an inherited routine into a decision with an owner. Boards, auditors, regulators, and insurers increasingly ask for exactly that distinction.
The judgement I supply is not what you should decide. It is why the current position is anchored, and why the anchoring is difficult to see from inside. Three mechanisms recur.
Premises are carried by architecture, not by people. Delegated authority levels, escalation routes, accountability lines, and performance measures all encode an assumption about the people they govern. Recruiting someone who holds a different assumption does not by itself change the assumption the structure runs on. External recruitment is commonly treated as a capability intervention; it cannot by itself alter the governing premises embedded in the architecture that the recruit enters. When the intervention then disappoints, the disappointment is usually attributed to the individuals rather than to the theory of change behind the appointment. This is the subject of my current research, at present under peer review.
An assumption is visible only against a contrast. Where few people inside an organisation have worked under the alternative arrangement, the organisation does not experience its assumptions as assumptions at all. They present themselves as the way work is done. This is a matter of scarcity of contrast, not of anyone's competence.
Not knowing is often a property of the measurement apparatus rather than a choice. Different instruments count different populations on different cycles, and frequently nothing in the machinery places their numbers side by side. The gap is then invisible without anyone having decided to look away from it.
None of these are attributions of failure to the people currently holding the roles. The original reasons for most frozen arrangements were legitimate, and often still are in part. Naming the layer that carries the premise, rather than the individuals operating inside it, is what makes the question askable at all.
Procurement and supplier arrangements. Which assumption about supplier behaviour is your regime built on, and does it still hold? Regimes designed around the cooperative counterparty behave differently once counterparties optimise against the disclosed criteria. Related: concentration that accumulates through individually sound award decisions.
Portfolio and investment governance. How many things can your leadership actually run at once, who set that number, and when? Beyond a threshold, each added initiative or partnership consumes the coordination capacity meant to deliver the others. Long-horizon capital commitments made on forecasts that cannot be reliable over that horizon are the same question in a different form.
What the board actually receives. Between the operational floor and the board, warnings are moderated, and caveats accumulate, each step locally rational. The question is whether what arrives at the top can still support a decision, and whether a director could reconstruct, three years on, the reasoning behind a decision made today.
Disclosure, explainability, and transparency. Openness is designed for the party acting in good faith. Where explanations can be aggregated at population scale, the same mechanism serves those optimising against the system. What to publish, and at what granularity, is a design choice worth making deliberately.
Security and remediation work under fixed capacity. When skilled people are unavailable at any budget, the problem stops being how much to spend and becomes what order to work in. This is the one strand where the output is a sequence rather than a question, and I say so plainly.
Delegation to AI. Which reasoning tasks has your organisation quietly handed over, who now owns the framing of the options, and what capability would atrophy if the present arrangement continued for two more years?
It starts with a decision already in the diary: a policy refresh, an investment gate, a committee's terms of reference coming up for renewal, a procurement cycle. Where there is no scheduled point at which the question lands, the work turns into a report, and reports of this kind are softened on their way up. That is the same mechanism the work exists to examine, and it applies to me.
The contract names the access: which documents, and how many interviews. Elicitation in the room is structured so that seniority does not set the anchor, using equal weighting, simultaneous disclosure of positions, and deliberately coarse scoring. The output is a decision record in your format and over your people's names.
Terms. A fixed end date agreed at the outset. No implementation phase, no retainer, and no participation in any subsequent tender arising from the work. Whether the engagement succeeded is answered six months later, by whether the assumption register is still being maintained.
Two routes, and they end in the same place: a fifty-minute first call, at no charge.
Book the call. calendly.com/vshabad. The booking form carries a field headed "Please share anything that will help prepare for our meeting." That is the one that matters. What goes into it is what I read before we speak, so the call can start from your situation rather than from my introduction.
Or write first, and book afterwards. If the background is easier to set out in prose, or if you would rather establish whether a call is warranted at all, send it to vshabad@vshabad.com and book at the same link once you have.
Either way, the useful material is roughly this, and none of it needs to be polished or complete.
That first call is about whether there is a question here worth doing this to. That there is not is a legitimate outcome of it, and a quick one.
The analytical structure draws on published and peer-reviewed work: Freedom of Information research across NHS trusts, police forces, fire services, universities, and water utilities; National Audit Office and parliamentary evidence; and practitioner interviews in government and critical infrastructure. One finding worth stating on its own, because it invites re-examination of an assumption many organisations still hold: across 171 NHS trusts, security compliance did not track resource (p=0.88, Cohen's d=0.02).
The apparatus was developed largely in public services, healthcare, and critical infrastructure. Where it carries over to other sectors, that is established case by case rather than assumed.
This page describes how I work. It is not an offer capable of acceptance, and nothing here forms a contract or a commitment to take on any particular piece of work. Terms, scope, and fees are settled in writing for each engagement.
I check every enquiry for conflicts of interest before anything is agreed. Existing commitments, prior engagements, and the confidentiality attached to them may require me to decline, and I may have to do so without giving a reason, because the reason itself can be the confidential part. A declined enquiry carries no implication about the enquirer.