Why Procurement Finds It So Hard to Say “Our Data Is Wrong”

The gap between what teams privately doubt and what they publicly present, and how to close it without anyone losing face

In the last edition I wrote about the inherited assumptions organisations carry into every tender, prompted by a widely shared saying about putting down the weight of other people’s opinions.

Several readers wrote back to make a fair point, which is that most procurement and estates professionals already know their data has weaknesses.

The difficulty lies less in ignorance than in the fact that saying so out loud, in a board paper or a contract review meeting, carries a professional risk that keeping quiet does not, and the opinions people fear in that moment belong to their directors, auditors and peers.

After more than twenty-five years of auditing FM contracts, I would go further. The reluctance to acknowledge flawed data is one of the most expensive habits in our sector, and it persists largely because the cost of honesty is felt immediately by an individual while the cost of silence is spread across an organisation over many years.

The private doubt behind the public figure

The research on this is striking. A joint SpendHQ and Procurious survey found that three quarters of respondents doubted the accuracy of the procurement data they present, which I think deserves a moment’s reflection, because it describes people who are well aware of a problem yet present figures they do not fully believe, since the organisation expects figures and the alternative is to explain why they cannot be provided with confidence.

More recent studies point the same way. Graphite Connect’s 2026 benchmark found that 47% of procurement leaders do not trust their supplier data, and ProcureAbility’s survey of 160 senior procurement leaders, published only last month, found that half regard poor data quality as the main obstacle to scaling AI, 51% see it as the largest gap in their analytics and 44% name it as their main supplier management challenge.

Gartner adds a revealing detail, noting that 59% of organisations do not measure data quality at all. When the problem is not measured, it cannot be reported, and when it cannot be reported it is very easy to leave out of the conversation altogether.

How defensiveness gets built into the contract

One thing I’ve noticed is that the reluctance rarely shows up as a single decision to conceal anything. It appears instead as a series of small, reasonable-looking choices. A specification is written loosely enough that nobody can be held to a figure that might be wrong. Key performance indicators are chosen because they can be reported from existing systems rather than because they measure what matters. Variance between budget and outturn is explained through narrative rather than reconciled line by line. A benchmark is selected because it supports the current position rather than because it is genuinely comparable.

Each of these choices protects someone from an uncomfortable question in the short term, and each of them weakens the organisation’s ability to manage its supplier. A contract with vague volumes cannot be enforced with precision. A KPI regime that measures activity rather than outcome cannot identify underperformance. And a supplier who recognises that the client is uncertain about its own data holds a quiet advantage in every negotiation that follows, from annual indexation to the scope of the next variation.

The compounding cost of being confidently wrong

There is a principle often cited in data management, sometimes called the 1x10x100 rule, which holds that an error corrected where the data is first captured costs a fraction of one that reaches the point of decision, where the cost can be a hundred times greater.

FM procurement is a textbook illustration. A miscounted asset category corrected during mobilisation costs a few hours of survey time. The same error discovered three years into a contract has been priced into every monthly invoice, embedded in the maintenance schedule and used to justify capital decisions, and correcting it involves a commercial negotiation with a supplier who has every reason to resist.

In the public sector the stakes are rising further. Under the Procurement Act 2023, contracting authorities must set and publish at least three KPIs for contracts above £5 million and publish a contract performance notice at least annually, rating supplier performance against them. Those ratings are visible to competing suppliers and to the public. An authority that has defended a weak baseline now risks publishing performance judgements built on data it privately doubts, which is a far more exposed position than acknowledging the weakness before the contract is let.

Reframing honesty as a leadership strength

Experience has taught me that the organisations which handle this well are distinguished less by the sophistication of their systems than by leaders who have made it safe to say that the data is not yet good enough, and have separated the quality of the data from the competence of the people who inherited it.

A procurement lead who reports that the asset register is sixty per cent reliable, explains how that was established and sets out a plan to improve it is demonstrating exactly the judgement a board should want. The difficulty is that many governance cultures still reward the confident figure over the honest qualification.

Changing that culture does not require a large programme. It requires a handful of deliberate practices that make honesty the easier path:

• Attach a confidence rating to every material figure. Ask that board and committee papers state how each key number was derived and how far it can be relied upon, using a simple scale. Once qualification is expected, it stops looking like weakness.
• Separate the data from the individual. Frame data reviews as an assessment of the information, not of the people who manage it, and say so explicitly when commissioning the work.
• Commission independent review at fixed points. Build an independent data and contract review into the contract lifecycle, at mid-term and twelve to eighteen months before expiry, so that scrutiny is routine rather than a signal that something has gone wrong.
• Measure data quality as a performance indicator. Include register accuracy, timeliness of updates and completeness of service records within the supplier’s KPI regime, with the obligation to maintain the data written into the contract.
• Reconcile rather than narrate. Require variances between specification, invoice and outturn to be reconciled at line level each quarter, so that gaps surface as data rather than as explanations.
• Report the improvement, not only the position. Give boards a trend in data confidence over time, which allows a team to show progress rather than defend a static number.

The saying that started this series is about not living under the weight of other people’s judgement. In procurement, the judgement people fear most is the moment someone discovers that the figures were never as solid as they appeared. The evidence tells us that most teams already suspect it, and the most reliable way to remove that fear is to be the person who raises the question first, with a plan attached.

Questions worth asking of your own organisation:

1. Which figures in your last board paper would you struggle to defend under independent scrutiny?
2. Does your governance process reward a qualified answer, or only a confident one?
3. If your supplier knows more about your estate than you do, who really holds the negotiating position?