Data, Trust and Decision-Making – Reflections from the BIG Book Club

“If decision-makers ask questions because they don’t trust the data, then your governance forum becomes a talking shop rather than a decision-making body.” – Chris Walters

The latest BIG Book Club session explored one of the most persistent and uncomfortable problems in modern organisations:

How can governance operate effectively when the information it relies on is fragmented, manually assembled, inconsistently defined, and often poorly trusted?

The session focused on the Data chapter of the Business Integrated Governance (BIG) Body of Knowledge. The discussion built naturally on the previous session around Information, where we explored the idea that governance is fundamentally dependent on good information to support oversight, intervention and decision-making.

What quickly became clear during the discussion is that most organisations do not lack data. They lack integrated, trusted, decision-ready information. And those are not the same thing.

The discussion exposed a fundamental divide

One of the first prompts posed to the group was: “Do you design data from requirements, or inherit it from systems?”

This immediately surfaced a major tension. Most participants agreed with the principle that governance and decision-making needs should define information requirements first, with data structures designed to support them.

However, there was also widespread recognition that this is rarely how organisations evolve in practice.

Instead, many organisations inherit:

• finance systems built for finance
• delivery systems built for projects
• risk systems built for compliance
• HR systems built for administration
• and operational systems built around local process needs.

Any Integrated governance has to attempt to operate by manually reconciling fragments from each.

This is where BIG introduces an important shift in thinking.

Traditional organisations often ask: “What reporting can our systems produce?”

BIG instead asks: “What decisions and oversight does the organisation require – and therefore what information architecture must exist to support it?”

That is a fundamentally different starting point.

Governance by spreadsheet

“The spreadsheets just become… a complete mess… it becomes a bigger job to get the spreadsheet working than it is to actually get better governance working.” – Chris Bragg

A recurring theme throughout the discussion was the operational reality many practitioners experience every day:

• manually assembled governance packs
• duplicated spreadsheets
• inconsistent definitions
• conflicting versions of the truth
• and significant effort spent reconciling information before meaningful discussion can even begin.

Several participants described environments where PMOs and governance teams spend enormous amounts of time manufacturing governance information from fragmented sources rather than governing through integrated information.

One participant described “the horrors of spreadsheets” that become more difficult to maintain than the governance process itself.

Another noted that governance discussions can easily become dominated by debating the credibility of information rather than making decisions.

That observation is critical. If participants in a governance forum do not trust the information in front of them, governance deteriorates into discussion without confidence.

BIG therefore treats information not as a reporting output, but as a core operational capability of governance itself.

Provenance, ownership and traceability

“If I don’t understand the provenance, I don’t believe it.” – David Dunning

The discussion repeatedly returned to several connected themes:

• provenance
• ownership
• traceability
• comparability
• assurance
• and trust.

Participants stressed the importance of understanding:

• where information came from
• who owns it
• who is allowed to change it
• how it has been validated
• and whether it is consistent with related information elsewhere in the organisation.

An especially strong point emerged around provenance: if a number appears in a governance dashboard, somebody should be able to explain where it originated, how it was transformed, and whether it can be trusted.

Without that, governance bodies are left relying on interpretation and persuasion rather than evidence.

The session also surfaced an important distinction: governance is not only supported by data – governance is also required over the data itself.

That includes:

• ownership
• permissions
• stewardship
• assurance
• security
• and semantic consistency.

From a BIG perspective, governance and information architecture cannot be treated as separate disciplines.

The “single source of truth” problem

“I see a lot of organizations struggling and messing about, trying to botch data together, because they haven’t settled on a common way of defining the fundamental things.” – David Dunning

Another major theme was the idea of a “single source of truth”.

Many organisations claim to have one.

In practice, this often means: “one dominant operational system”.

The discussion highlighted why that interpretation is too simplistic.

Different operational systems will continue to exist for good reasons:

• finance
• HR
• delivery
• operations
• customer management
• risk
• engineering
• supply chain.

BIG does not require a single monolithic operational platform. Instead, BIG promotes:

• integrated information models
• common references
• common semantics
• traceable reconciliation
• explicit ownership
• and governed integration.

The issue is not necessarily the number of systems. The issue is whether governance can trust the relationships between them.

Several participants highlighted the importance of:

• common data dictionaries
• common coding schemes
• consistent definitions
• and agreed meaning.

Without those foundations, organisations frequently find themselves comparing apparently similar information that has been generated using entirely different assumptions or structures.

Governance, bureaucracy and organisational reality

“Data quality can get too big and be a bureaucracy…” – David Mitchell

One of the most valuable aspects of the discussion was that it did not become unrealistically idealistic.

Participants also acknowledged the risks of excessive governance overhead – which include the accusation of excessive governance.

A mature enterprise-wide information governance environment may improve trust and consistency, but it can also introduce:

• operational drag
• increased cost
• additional process layers
• and slower adaptation.

One participant observed that data governance can become “a bureaucracy” if not approached carefully.

This is an important challenge for BIG.

Integrated governance cannot simply mean: “more process”.

The aim is not to create an administrative machine that consumes the organisation.

The aim is to improve decision confidence, alignment and coordination while remaining operationally practical.

That balance remains an important open question.

The role of AI

“We should be able to make this a lot easier than it has been in the past.” – David Dunning

The session began with the intention of considering how the Data chapter might evolve in light of AI.

Interestingly, the discussion naturally exposed why AI raises the stakes significantly.

AI can potentially:

• reduce manual reconciliation effort
• identify hidden relationships
• support sentiment analysis
• surface emerging patterns
• and improve integrated visibility.

However, AI also amplifies existing weaknesses.

If organisations already struggle with:

• inconsistent semantics
• fragmented ownership
• poor provenance
• duplicated truth models
• and weak integration

then AI may accelerate confusion rather than improve governance.

The issue is not simply whether organisations adopt AI.

The issue is whether they possess sufficiently coherent information foundations to use AI responsibly and effectively.

This remains an important area for future BIG exploration.

What the discussion reinforced from a BIG perspective

“The systems must come and support decision-making. They’re a supporting element, they’re not a driving element.” – Chris Walters 

The session reinforced several core BIG principles.

First, governance should define information needs – not merely consume whatever systems happen to produce.

Second, organisations require integrated visibility across:

• objectives
• performance
• delivery
• risk and uncertainty
• resources
• finance
• and operational reality.

Third, provenance, ownership and traceability are not technical nice-to-haves. They are governance requirements.

Fourth, governance requires operational support and information architecture – not simply meetings, committees and reporting rituals.

Most importantly, the discussion reinforced that governance maturity and information maturity are deeply connected.

Fragmented governance tends to produce fragmented information. Integrated governance requires integrated information.

Questions still to answer

The session generated strong discussion, but it also surfaced important unanswered questions:

• What does a practical BIG-aligned governance information operating model actually look like?

Editors note: Clearly, such an operating model has to be specific to an organisation state and situation, but in informed by good current state analysis, information and data governance good practice, governance expertise and Business Integrated Governance knowledge.

• What is the minimum viable integrated information backbone?

Editors note: The BIG CIC developed possible operating agendas for a number of Governance Nodes – including board, portfolio / programme / project, department, and product group. If is clear that while specifics of decisions that come before boards may need specific data, the core information underpinning governance includes very obvious data types -which it should be possible to commonly define and integrate.

Furthermore, it is  clear for every organisation there is a set of metadata to categorise information with – from employee data, to finance codes, project codes, department codes etc.

Both of these can emerge in discovery exercises (if not documented already)

• How much governance rigor is enough before it becomes bureaucracy?

Editors note: This is the usual challenge governance faces. However, a BIG Argument is that if governance is designed to be integrated, then boundaries and interfaces reduce, and possibilities from integrated information open up.

• How should organisations evolve incrementally from spreadsheet-driven governance toward integrated governance operations?

Editors note: The question is – should they evolve incrementally – or even will they? The question more useful might be whether the spreadsheets should connect to reliable data sources.

• Who should own enterprise governance information architecture?

Editors note: This is a situation specific decision, but this could easily fit under a CFO / COO / CIO – depending on existing remits.

• What should governance bodies genuinely expect from management information?

Editors note: The BIG proposition is that governance bodies should be able to explain their standing agendas and be able to indicate their for information needs, and from where that is expected. An operating challenge is selection, summarisation and contextualisation of that data. This requirement will include specific information from time to time, and it will develop.

Governance bodies should genuinely expect to be listened to and engaged for the management information needed – and clearly used.

• How should AI participate in governance ecosystems?

Editors note: This is another large topic to be explored. For examples see BIG Blogs

• How do organisations balance flexibility with semantic consistency and control?

These are not purely technical questions. They are governance questions. And increasingly, they may become some of the defining organisational questions of the AI era.

What do you think?

Join the Community Discussion

What next?

Subscribe to receive notification of blogs and events

Become a Member to access BIG Materials

See the BIG Solutions we have connected with already.

Contact us for deeper insights on anything BIG.