Convincing an organization to act on data quality subjects might prove difficult. Asking your superior for resources to address data deficiencies that cannot be otherwise justified by regular means for project launch, is just one of the reasons.
Unpacking the problem
An example could illustrate this better. Your lead analyst discovers a data quality issue in the marketing database. After her analysis and explanation, you, as a marketing director, understand that it can distort the client overview reports that you send to the CEO for the monthly business review meetings. You immediately instruct the analyst to correct it. The analytics team spends some time on it, and they manage to repair the data. Moreover, they find the root cause and set up a prevention mechanism. This is an encouraging news, but it leaves you with a concern. If one problem has been found, there is a good chance for more hidden deficiencies. Next month you would have to send the report to the CEO again and you would not be sure if the underlying data can be trusted. After all, if your analyst had not identified and tackled the issue in the first place, the CEO could have made a wrong decision based on misleading information that was, ultimately, provided by you.
The challenge
After the issue has been remedied, you sit down with the analyst to discuss your concerns. You explain that such data gaps could have a serious business impact and you believe something should be done about it. Your idea is such gaps to be proactively identified and timely fixed. She agrees with your line of thinking and is very excited that you acknowledged this as a threat too. She explains that such issues are what she has been struggling with for a long time but has never managed to attract enough attention to them. She also seems to know exactly how to address them. She needs three more data analysts to build a capability that tests the complete marketing data and highlights any potential quality gaps. Going further, this cannot be a one-off check, so they should stay as a dedicated team in your department. The rationale is that data is changing constantly, hence quality issues should not be fixed just once, but the data should be continuously monitored.
You know that she is right. On the other hand though, hiring three additional employees to just look after data quality, is a tough sell. You already know you would need to make a business case and pitch it to the CEO. Initial estimations show that you will require an additional one hundred and fifty thousand euros budget per year.
Step-by-step towards a solution
Organizations are usually challenging to steer, especially when the ask is to spend without the prospects for tangible returns. It is true that high data quality could have some quantifiable benefits, but also there are others that are indirect and subtle. Therefore, we have created the following five-step guide to successfully convince the organization on the value of data quality improvement initiatives.
Step 1 - Understand the problem well
Finding data quality issues might be a rare occurrence for you or may be happening daily. In any case, if you would like to trigger a response from the organization beyond just fixing the issues, you need to understand it first. This might not be that easy or obvious, so we suggest the following approach:
First: find out where it happened. Since data usually moves across departments, if you have found issues in your database, they might not originate there. More details you have, would make it easier to sell the initiative to the stakeholders involved.
Second: make sure you fully understand the consequences. What could go wrong due to these gaps should be carefully examined - not only the direct impact on your department or database, but also the impact they could have on the whole organization. Your investigation should cover processes in other areas, dependent on the data in focus; decisions that are based on it and could eventually prove wrong; and systems that could stop “talking” between each other. All of these are domains of potential business impact.
Third: estimate the monetary impact of the feasible outcomes. There might be the possibility for achieving results, which would have a direct impact and can be easily approximated. For example, “fixing the data quality issue will save x-amount of man-day effort” can be calculated by multiplying the price of labor per day by the number of people involved. Some consequences, like preventing negative intangible business impact like loss in reputation, however, could be more difficult to estimate. One good way to do that is the “cost avoidance” technique.
Step 2 - Start developing a business case
Once you are aware of origin and location of the data quality issues, you should also be able to define the root cause of the data gaps. Understanding the business consequences they might follow, could give you an idea about the scope of the problem. Last but not least, estimating the financial impact on the company would summarize the real costs associated with them. These three elements are what you should start your business case with.
For an executive though, the bottom line is what is important. The euro estimate you have created in the previous step can be used as the “net income” part of a Return-of-Investment (ROI) calculation.
Still, a word of caution: even though we might use ROI as an approach to represent the financial benefits of a data quality initiative, we would advise you to restrain yourself from adopting hardcore financial techniques to judge upon or prove the viability of your project. The risk here is it could quickly turn into an investment for your company which will inevitably be benchmarked against an already accepted process of prioritizing the allocation of resources. This could work in companies where data is the core business, but the majority of executives might not be too familiar with comparing data related initiatives to regular project proposals. Therefore, we would recommend using the ROI approach lightly and just for illustrative purposes, while still acknowledging that majority of the benefits are intangible and spread across time.
Step 3 - Define the cost and effort required
So, the benefits side of your business case is complete. It is time to get a price on how much it would cost to build and operate such a solution. We would suggest following the corporate guidelines by the book here. Organizations are usually very good at estimating costs thus, we do not see any issue in adopting any proven technique. To increase your buy-in however, we would recommend elaborating several options to finance this initiative: 1) with a focus on the euro cost, 2) on duration, 3) on feasibility, and 4) on quality. In our experience, executives are much more receptive to new initiatives when they have options to choose from.
Step 4 - Find support
The ROI calculation of your initiative at this stage should be vastly positive. And rightly so. Fixing data quality issues could have a significant impact across the organization with, comparatively, less effort. But at the same time, it will be full of assumptions, especially on the “What are the positives?” side. Despite having been validated, these assumptions could play against you and your initiative since it is easy to disprove benefits that could not be precisely measured.
Therefore, to increase the likelihood for a positive outcome of your pitch, it is best to first gather support for your initiative. Our advice here is to connect with the decision makers in the teams where the data in question either originates, is modified or passes through. There is a good chance that the same issues you experience are also a focus in those teams. This is the most trouble-free way to get support - since your proposed solution would also benefit others, they should, and in most cases will, support you.
Step 5 - Jump!
By now you should know the benefits and the costs of your initiative. You should have gathered the support from multiple teams and have developed several options to deliver it, each with its pros and cons. It is game time now! The final step is easy. Jump!
Schedule a meeting with the decision makers and do it as soon as possible. You should ride the momentum that you have generated across the organization. Also, do not expect an immediate “Yes” but rather treat your first meeting as the start of the conversation with the leadership team. If they agree within the first half an hour - great! You have done a fantastic job to prepare! But often, you should be ready to embark on a journey. If this happens, our advice is to keep the talk about your project going and, more importantly, listen to the organizational feedback.