The Financial Impact of NOT having Data Governance
During a recent customer visit, we got discussing the financial impact of Data Governance. To help explain this point, I thought I’d share some of the more common problems associated with NOT having data governance. By looking at it from this point of view we can get an idea of what the business is doing to overcome these issues, against which we can then associate some value.
The Financial Impact of NOT having Data Governance
This isn’t meant to be an exhaustive paper on the subject, more a sharing of thoughts and ideas. I’d also add that the ideas presented in this blog aren’t suggesting these impacts will happen, more a sharing of some common challenges we see in the world of Financial Services and a way to try to understand the potential financial impact they might cause. These challenges should be seen from the perspective of potentially being part of a broader Data Governance initiative.
To make the explanations simpler, I’ve assumed that the bulk of any Data Governance related issue is concentrated at three points in the Information Supply Chain. These points are:
- Data Ingestion
- When data is initially ingested into the Information Supply Chain or when the data is entered into a system or process (includes physical data input)
- Business Data Consumption
- When data is consumed by the business for the purposes of supporting the operations of the business. This would typically be data consumed within the Lines of Business (LoB) and to support specific LoB functions
- Enterprise Data Consumption
- When data is used either at the enterprise level (for example, for business planning purposes) or when data is required at an enterprise level (for example, for an annual report or for a regulatory report)
I’ve obviously made this is a very simplistic model and Data Governance would also concern when data moves ‘between’ data stores or applications (lineage). In making this model simply, it makes the concepts of where the financial impact of NOT having Data Governance a little easier to explain.
The objective of this blog is to try and help encourage more parts of an institution to become more active participants in the Governance of Data (because the implications are much more visible) and to support teams, that are trying to provide the necessary capabilities, in the justification of budget allocation for these capabilities.
This is the point where data enters the Information Supply Chain. This data could be provided in several different forms but some common ones include:
- Structured Electronic Data Files
- Unstructured Electronic Data Files
- Manual Data Entry (by users of the business)
- Manual Data Entry (by customers of the business)
On entering the Information Supply Chain, a number of checks and controls should be run that will ensure that only the highest quality data proceeds. Now, these checks don’t always happen and when they do, they don’t always pick up all issues with the data. Some common issues that can get missed include:
- Missing data (i.e. missing postal code from an address)
- Duplicated data (i.e. same data entered multiple times when this shouldn’t occur)
- Incorrect data (i.e. picking up the wrong data field when looking for data of birth)
- Incorrectly interpreted data (i.e. assuming a data is based upon UK formatting when it is in a US format)
- Non-standardised data (i.e. addresses are provided in a range of different formats)
So, when these issues get missed I would describe these as ‘characteristics’ of NOT having Data Governance. Because the data is in the very early stages of the Information Supply Chain, there is likely to be a minimal financial impact from this data – basically because it’s not being consumed yet.
There is likely to be some financial cost incurred later, when checks flag up issues with the data which are then processed to correct them. This cost of processing is often passed to specific individuals or specialist teams for correction. Because the data is so early in the information Supply Chain, there are often a relatively small number of places it needs fixing. The combination of a small number of places to fix the data plus the relatively low cost base of the individuals or team who process these corrections, means the financial impact could be relatively low.
Here is a simplified, worked financial example.
- Each error takes 1 person to fix it
- Each error takes each person, 5 minutes to fix
- Each person costs £1 per minute (fully loaded cost)
- Total Cost per error is £5 (5 minutes * 1 person * £1 per minute per person)
If an organisation had 10,000 errors in total to fix across a year, this would result in a cost of £50,000 per annum. For sake of simplicity, I’m assuming that the appropriate application of Data Governance will fix all discovered data issues.
For this stage of the Information Supply Chain, the potential financial impact of NOT having Data Governance is could be approximately £50,000 per annum.
There might also be other factors that will contribute financially, such as data reprocessing costs if using an external agency, so for simplicity I’ve removed these. But be advised, these do exist in many institutions.
Business Data Consumption
This is the stage of the Information Supply Chain where the data is now being consumed and used by the business community to support business operations. Often, the ingested data has
- proliferated to many business supporting systems
- been used as part of business supporting models and calculations
- been used in departmental / LoB reports
- form the basis of customer oriented calculations (risk models, premium calculation, interest calculation etc.)
If there are issues with the data, there are often many places if needs fixing plus it may take significantly more work to identify that there is something wrong with the data. One common action is for Line of Business staff to determine what’s wrong with the data then remediate it themselves (as they probably don’t want to incur the perceived time penalty often associated with returning the data to the start point for remediation there). This is where our financial calculation gets a bit more expensive.
Here is a simplified, worked financial example.
- Each error takes 2 people to fix it
- Each error takes each person, 10 minutes to fix it
- Each person costs £2.50 per minute (fully loaded cost)
- Total Cost per error is £50 (10 minutes * 2 persons * £2.50 per minute per person)
If an organisation had 10,000 errors in total to fix across a year, this would result in a cost of £500,000 per annum. Once again, for sake of simplicity I’m assuming that the appropriate application of Data Governance will fix all discovered data issues.
For this stage of the Information Supply Chain, the potential financial impact of NOT having Data Governance is could be approximately £500,000 per annum.
There could also be other factors that will contribute financially, such as the financial cost of fixing an incorrect premium calculation or the remediation cost associated with an incorrect interest calculation, so for simplicity I’ve removed these. This example has really just looked at the cost associated with remediating data issues and, it could be argued, that there is a much higher cost associated the remediation associated with fixing the decisions made with this data.
Enterprise Data Consumption
This is where the calculations get a bit vaguer, but often larger, than in previous stages of the information Supply Chain. It’s at this stage, where data is used for either strategic purposes or market/regulatory purposes, where the consequences of data issues become very significant if they aren’t caught in time.
Some examples of the types of consequences could include:
- An incorrect regulatory filing that could result in a substantial fine
- An incorrect statement about the health of the business, in an annual or quarterly report, that could result in a substantial reduction in the share price of the institution
- Incorrect data, used to create predictions about enterprise-level future sales, that could potentially result in additional expenditure when it’s not required (i.e. hiring more sales people when they aren’t needed)
- Incorrect data, used in capital adequacy planning, that could lead to greater cash retention than required
I could have included more examples but the main point is that each one of these will often carry a financial burden which will often run into hundreds of thousands or millions. Issues with data alone, that create an incorrect view about the wealth or wellbeing of the institution, could lead to sizable amounts of additional money being utilised as the data has caused ‘under-planning’ or ‘over- planning’ issues. ‘Under-planning’ is where the business hasn’t allocated enough capital for planned operations and ‘over-planning’ is where a business has allocated too much capital for planned operations.
By just taking a few of these examples, it’s easy to get a potential financial impact that could run into £millions.
What I’ve tried to demonstrate here is the financial impact associated with NOT having Data Governance as data flows through an Information Supply Chain. What I’ve tried to show with the examples is that the further un-governed data flows through an institution, the more potentially expensive it gets to address the issues it causes.
The purpose of this blog wasn’t to provide a specific answer to a specific question, more to share some thoughts and ideas on how to explain the potential financial impact of NOT having Data Governance. By providing this explanation, it is hoped that more parts of an institution become more active participants in the Governance of Data and that the appropriate budgets are allocated to provide the capabilities necessary to properly Govern Data. These samples should be see in the context of them being part of a broader Data Governance initiative
If you want to understand more about the ‘Impact of Data Issues’, here is a blog I’ve written previously on the subject.