Data debt is a concept that John Ladley put forward. It is a framework that places value on Data Management activities. As such you can use the concept when discussing the need for data governance with management, and can even be included in Data Governance Business cases.
What is Data Debt and how do I use it?
Data Debt as a concept borrows heavily from the concept of Technical Debt. You accrue technical debt whenever you take shortcuts writing code today that will cost you more in re-work tomorrow. Data debt follows exactly the same logic. You accrue data debt whenever you make system changes that negatively impact your Data Policies and Standards.
For example, assume your Head of Sales adds fields to the CRM without proper Governance Processes. In this case, you have new Data Capture requirements but no official definition of the data you expect to capture. Data Management didn’t get the chance to implement data validation on capture, meaning poor quality data can be introduced into the CRM. As a result, there will be future costs to cleanse the data as well as costs of misunderstanding.
You can use data debt as a concept to demonstrate there is a cost for every data decision. This cost is either borne upfront, when you push the new CRM field request through the Data Governance processes, or later when you need to clean up the mess. As with most preventative measures, incurring the cost up front will save more money later. But there are also good reasons for why you may be forced to accept some data debt. For example; when a new regulation is imminent, and you lack the time to do things right before the deadline hits.
By including this concept in your planning, you acknowledge the fact that Data Governance will come at a cost. The choice your stakeholders need to make then becomes “do I pay now or do I pay later?”.
What are the 4 quadrants of Data Debt?
I want to make minor modifications to the naming that John uses, because I feel some terms can be misunderstood. In his article, John breaks down the quadrants into:
- Illiterate – incurring the debt because you don’t know any better
- Resistance – wracking up the debt wilfully without accounting for future cost
- Realisation – where you see the mounting costs of poor data quality and time wasted and tally this after the fact
- Acknowledged – where you know you’re incurring a cost but do so out of necessity and plan to pay it back
Personally I think “Illiterate” is somewhat unfair. I’d call that “Immature”. If you’re level 1 or 2 in Data Maturity you’re probably just unaware of these costs. If you have no formal Data Management plans in place, how would you know the costs you’re incurring? For “Resistance” I’d probably switch it to “Deliberate”.
Where do my Customers come into this?
The costs described above are those of traditional data management. You either spend money planning and preparing data today, or you spend more money dealing with the consequences tomorrow. Where you spend the most money is when these Data issues are exposed to your Customers. Examples of how this happens include:
- Over-billing due to data quality issues (and where you over-bill you can bet you’re also under-billing)
- Inability to solve a customer complaint in a single call (because you lack the right data)
- Sending out service staff to fix an issue they can’t solve (because you lack data on the problem)
- Sending products or invoices to the wrong address
- Shipping the wrong products to the wrong Customer
- Over-selling and creating Customer dissatisfaction
This short list shows some of the ways poor data practices can hurt Customers. Every time a Customer experiences these things, their trust in your brand diminishes. Customer Satisfaction plummets, and your Customers leave you. Customers leave negative reviews online – reviews you can use to identify data opportunities.
What does the research say about this?
Cognopia ran research into the Telco industry in 2020. We used our DDT Method to evaluate data quality and place a dollar cost on the result. We read through 47,907 online Customer reviews in 51 Telcos across 22 Countries in Asia. Here’s what we discovered:
- 76.5% of companies had Billing data issues
- 80.4% of companies had Account data issues
- 41.2% of companies had poor Customer contact data
- 52.9% of companies incurred unnecessary Truck roll costs due to bad data
- 72.6% of companies had issues with their Product data
You can read through the report here, where we pull together Data Maturity and Data Quality.
How do Customers pay Data Debt?
Customers don’t pay the Data Debt directly, they experience the negative results of letting the debt accrue. By experiencing bad service, inaccurate bills, or receiving the wrong products your Customers feel your Data pain. These Customers will leave if you do not get your house in order. This is where the Customer pays the data debt; by taking their money elsewhere.
We use this approach in our Data Design Thinking sessions to place a dollar cost on this bad data. By tying together Customer Experience and Data Quality we can calculate the real cost of this Data Debt.
Book a call with our team using the link below if you want help calculating your own Data Debt.