Data issues: building on quicksand
"Sales have dropped by 50% overnight," our retail director announced in the emergency meeting. The data from our newly implemented credit vetting system showed customer rejections had doubled in the past 24 hours. Our sales teams were furious, customers were confused, and executives wanted answers. After a frantic investigation, we discovered the problem wasn't with our customers' creditworthiness at all - it was our data. The system was implemented to unify two legacy systems post-merger, but what appeared to be identical data fields meant entirely different things in each organisation, causing the system to misinterpret customer information and reject perfectly good applications.
Not managing data is one of the most common reasons I see transformations fail. Everyone talks about data, but few truly understand or actively manage it. As a transformation consultant, I've sat through countless presentations about new technologies, process improvements, and organizational redesigns. Yet when I ask about the quality of data that will feed these grand plans, I often get blank stares or vague assurances.
"In God we trust, all others bring data." – W. Edwards Deming
I personally learned just how dangerous this oversight can be during that credit vetting system implementation. On paper, everything looked perfect – we had two similar organizations with similar customer bases merging their systems. What could go wrong?
Despite six months of careful design and successful functional testing, the system started producing bizarre results when we went live. "Post code" in one system referred to work addresses, while in the other, it meant home addresses. Salary data in one system was monthly, in the other, annual. These seemingly minor differences cascaded into major issues that not only delayed the project but required extensive remediation across multiple systems.
Our mistake is one that I have since seen repeated many times. Transformation initiatives usually focus on systems, people, processes or the "plumbing" of the organisation rather than what is flowing through the pipes.
Ignore data at your peril
Data's role in transformation is uniquely pervasive - it's simultaneously the subject of transformation (what we're changing), the enabler of transformation (how we make changes), and the measure of transformation (how we track progress). This three-dimensional impact means that poor data management doesn't just create isolated problems - it can undermine every aspect of your transformation journey.
When Data is the Subject
Many transformations aim to improve how organizations handle data, whether through new systems, better analytics, or improved decision-making capabilities. If your underlying data quality is poor, these initiatives are doomed from the start. It's like trying to digitise a library where half the books are missing or filed under wrong titles - the technology might be perfect, but the outcome will still be chaos.
When Data is the Enabler
Even when data isn't the primary focus, it powers the transformation process itself. Configuration decisions, process designs, and organizational changes all rely on accurate data about current operations. Poor quality data leads to flawed decisions that can derail entire programs. I've seen organizations invest millions in automation only to discover their process assumptions, based on unreliable data, were fundamentally wrong. This is particularly relevant in the era of AI. How can we expect even the best AI technologies to work when we are feeding it unreliable data.
When Data is the Measure
Perhaps most critically, data serves as the compass for transformation - telling us where we are, where we're going, and whether we're on track. With poor data management, you lose the ability to measure progress effectively or demonstrate success. It's like trying to navigate with a broken compass - you might be moving, but you can't be sure you're heading in the right direction.
This triple impact creates a compound effect. Each data quality issue doesn't just create one problem - it spawns multiple issues that interact and amplify each other. The result is often a transformation that appears to be progressing on the surface but is building on quicksand.
Don't bring the answers, ask good questions
The first thing to acknowledge is that data management is a sophisticated field with entire disciplines and career paths built around it. From data governance to data architecture, from quality management to data science - these are complex specialties that take years to master. As a transformation leader, your role isn't to become a data expert yourself - that would be both impractical and unwise. Instead, I think your job is to ask the right questions, recognize when you need specialist expertise, and ensure data gets the attention it deserves in your transformation journey.
Having learned these lessons the hard way, here are the critical questions I now ensure get asked and answered in any transformation initiative - whether by me or by the experts we bring on board:
1. Who Owns the Data?
Not just technically, but who is accountable for its quality? In successful transformations, data ownership is clearly defined and supported with appropriate authority and resources. Without clear ownership, data quality becomes everyone's problem but nobody's responsibility.
2. What's the Current State?
Before launching any major change, conduct an honest assessment of your data landscape. Key questions include:
How complete is critical data?
Where does it come from and how is it maintained?
What quality controls exist?
How much manual intervention is required to make it usable?
This baseline understanding helps set realistic expectations and identify risks early.
3. How Will Data Support Your Future State?
Map out exactly how data will flow through your transformed organisation. What new data will you need? What existing data must be cleaned or standardized? What governance needs to be in place? One financial services client saved months of rework by identifying data requirements for their new digital platform before, not after, development began.
4. What's Your Data Strategy?
This goes beyond just cleaning up existing data. Consider:
How will you maintain data quality going forward?
What standards need to be established?
How will you handle master data across systems?
What training and tools do people need?
A clear strategy ensures data management becomes part of your organizational DNA, not just a one-time clean-up exercise.
Transformation isn't just about implementing new systems or processes - it's about creating sustainable change. Without a solid foundation of quality data, even the most brilliantly designed transformation will struggle to deliver lasting results.
Until next Friday, keep failing forward!
Failing Forward is a weekly newsletter exploring the lessons we can learn from failure. Subscribe to get insights and actionable strategies to drive meaningful change.
