In a past IAIDQ/UALR salary survey one of the biggest issues cited was the ability to demonstrate the value of data quality. This typically involves a business case that attempts to convince sponsors that the benefits of your data quality efforts will outweigh the costs incurred.
This is a big area and one article isn't going to come close to tackling this subject but in the spirit of continuous improvement I wanted to share some techniques that may just help tip that next business case in your favour.
Gathering Data Quality Evidence
How will you gather the material evidence used to demonstrate the need for action?
- Practice inside-out and outside-in approaches to data quality assessment. Most business cases start with some form of assessment which demonstrates the need for improvement. Many fall into the trap of relying solely on data profiling stats gleaned from the data (inside-out assessment). This alone is unlikely to compel sponsors into action. Instead, trace the information chain and associated impacts of poor quality data across your business. Record anecdotal evidence, speak to customers, create video interviews, promote an information worker satisfaction survey, measure impact of data quality on external business metrics other than cost (eg. lead times, external customer satisfaction etc.). Combine your detailed data assessment with the human impact of data quality, it can add serious weight to your case.
- Use full data volumes for any assessments (and during the presentation). I don't believe in assessing or presenting samples of data in the business case. The technology exists to assess data across full volumes so make use of it. Sponsors sit up in a presentation when you can drill down from a high-level impact story straight into the data that directly powers their business. Using a sample often means your presentation breaks down as you can't demonstrate issues in real-time and questions go unanswered.
- Gather historical evidence. Showing the trend of data quality issues over time is a powerful technique but how can you go back in time with your assessment? Most datasets have timestamps which show when records were created or updated. You can use this to demonstrate how the business has failed to manage data quality over time. Another technique is to take discrete snapshots from historical backups. Most DBA's will keep 12 month's worth of backups with each month typically having a full backup. Extract tables relating to your data subject area under review and create time periods for the key business objects. This sounds like a chore but it can often demonstrate that there are endemic problems with the way data is managed and not just sporadic recent events that sponsors can easily dismiss.
- Link the information chain to the service value chain. Probably the most powerful technique I have used for building the business case is to integrate a model of how the business drives value with the underlying data. By assessing the information chain that supports this service chain you can build a far more compelling story than simply measuring data quality in isolated data stores. Sponsors like to see that you are not just a data wizard but that you understand their business intimately.
- If the case is not compelling, look elsewhere. Creating a business case is actually a valid data quality technique in its own right. It forces you to view data quality issues through a business-focused lens so the business case effectively becomes a prioritization tool. The cost of any data quality initiative can be considerable, particularly when organisations fail to implement a prevention strategy, so be 100% sure the project will deliver clear benefits in a timely fashion. Missed It By That Much (Jim Harris), Data Quality Project Selection (Steve Sarsfield ) and Pick The Low Hanging Fruit of Data Quality and Data Governance (Mark Goloboy) may give you some inspiration here also.
Constructing The Data Quality Business Case
You have the evidence but how will you plan your initiative and create a compelling business case?
- Link your assessment to enterprise, local and personal drivers. Quite often your business case may be passed on to a more senior sponsor for review. It pays to understand the wider benefits that your data quality initiative will provide so emphasizing the strategic benefits of your initiative is important. How will improving the health of your CRM data help the wider organisation? Do you have the evidence to back this up? You need to understand the local drivers for the initiative but what about the personal drivers for the sponsors? What are their "hot buttons"? Have you discovered what they are measured on by their seniors? Will your initiative help save them from embarrassment or perceived failure? People buy with their heart and justify with their head so push those personal buttons. One tip I've found useful as an external consultant is to dissect the company annual report. You will often find nuggets of strategy that the board have divulged to shareholders.
- Have a Plan B. If all else fails and the sponsors reject your proposal, do you have a fallback plan? Many companies have bureaucratic procurement processes which mean your revised proposal may take several months to re-submit. By creating a range of options you at least stand some chance of getting a project off the ground.
- Be absolutely clear about your costs. Best-practice data quality means continual improvement (although many data cleanse projects are one-off). Don't shy away from providing total accuracy about the costs of the project. Far better to be honest at the outset than have to re-submit a proposal for additional funding in the next project review. Most data quality business cases estimate the potential revenue benefits of data quality improvement and sponsors may rigorously question this estimate. If you also have a rough estimate on the projected costs then the sponsors may view the project as an unnecessary risk.
- Use the acronym "S.T.R.I.F.E" as a crude starting point. If you're searching for inspiration for business case benefits then Stress - Time - Risk - Income - Fear - Expenditure are useful high-level starting points for linking data quality to business impacts. I find that creating a mind-map with these as a focus point in a workshop environment can often help to shape the case and link the various pieces of evidence together. Be wary of ignoring other issues however as these are very high-level motivators.
- Focus on competitive advantage. A strong motivator is to demonstrate where competitors have the edge. Not a simple feat to achieve but if you can demonstrate a competitor has faster lead times and you can close the gap with improved data quality, can certainly raise the attention level.
- Predict future trends.Can you gather evidence of the likely course of action if the sponsor fails to act? Pushes the accountability button as the sponsor will inevitably be held to account for not acting when the opportunity arose.
Presenting The Data Quality Business Case
The case is fully prepared, how will you demonstrate the value to sponsors?
- Don't use PowerPoint. I don't know if this is a coincidence but since I dropped it I have had far more success in winning proposals. Eliminating PowerPoint forces you to think of far more creative and engaging mediums.
- Demonstrate a link from impacts to evidence using an interactive dashboard linked to data. I typically create a hybrid business model, data model and data quality model populated with full datasets. Using a data visualization tool to present the high level impacts we can then drill-down into the detail as the inevitable barrage of questions come your way. I find that DQ tools are too technical in their presentation so I extract the data quality results from various tools, link it to business data, integrate a costing model then publish this in an interactive front-end in the visualization tool. This technique has dramatically improved the way I present business cases as it lays bear the entire problem. The sponsors can also take the dashboard away with them to speak with their respective teams which is far more powerful than a PowerPoint file. Another benefit is that the presentation has something of value for each different type of participant eg. senior business sponsor, IT representatives, data managers etc.
- Get a senior ally to present and step into the shadows. Simple technique, used it several times and it has certainly worked. I often struggle to speak "executive management" particularly when it is a new sector so swallow some pride, find an executive envangelist to help present the case and take a backseat.
- Keep it simple. Can you sum up your business case in less than a minute? Thirty seconds? Ten seconds? Brevity is key. Don't use ANY technical jargon or DQ buzzwords. When people don't understand a topic they can often lose interest and bring these negative feelings into their judgement.
- Customer testimonials can be powerful. There is something completely unbiased about customer stories relating to poor data quality. Sponsors will find it hard to reject testimony from this source. Just look at IQTrainwrecks.com for an example of what can go wrong.
- Use different presentation methods and drivers for different personalities. This list has recognised limitations (eg. some people can bridge psychological types) but you may find it of some use:
- Analysers: This sponsor want facts, figures, evidence. This is why you need to have the high-level story with the low-level data, they will want to see the facts for themselves.
- Achievers: Action focused sponsor, want to see clear plans, good structure, evidence of previous success. Want to see that you are going to deliver a successful initiative so can you demonstrate the results of a small pilot perhaps to alleviate fears?
- Socialite: These sponsors are acutely aware of their standing in the organisation and they may ask for additional assistance for reviewing the case. Demonstrating how the improvements will improve their importance and acceptance within the organisation is a good presentation tactic.
- Carer: This sponsor is averse to change and risk, a common trait of IT sponsors incidentally. Demonstrate that the project will only create positive change for the organisation and a reduction in risk. Focus also on the personal benefits of the project, demonstrate how it won't impact their career.
- Present a future with or without data quality management. Covered this earlier but using the dashboard technique and historical data I typically demonstrate projected defects and business impacts if the business fails to act. A great way to close the presentation as it leaves a simple take-away message, do nothing and the situation will deteriorate.
- Demonstrate the value to other initiatives. Are there data integration projects planned that will require the data in scope? Is there a downstream data warehouse or corporate reporting interface? Is the data used in any web services or customer interfaces? If you can demonstrate that adding value in one area can have a positive impact in another part of the business you are far more likely to get sign off. The sponsors may also look for small pockets of cash from other dependent units to help soften the financial blow on their own budget.
- Use an A0 chart that demonstrates the core issues and the route forward. Daragh O Brien has recently talked about this on the IAIDQ forum but I love this concept. The ability to lay the sheet out on a table and gather people round is so compelling. It also gives you a simple mechanism to brief each sponsor on the case before and after the main presentation so the issues and benefits stay fresh in their mind.(Tip: if you're struggling for ideas on how to create one there are a few examples in this case study pack from a specialist firm, Cognac, who are masters of the art).
- Practice your presentation. Obvious point but many people ignore it. Test your business case presentation on other senior managers and colleagues. Variety is the key as you want diverse questions and feedback to help you gain confidence and clarity for the live presentation ahead.
- Provide clarity about what will happen AFTER the project finishes. Many business cases focus on what the project will do instead of what the project will deliver. Sponsors are mostly concerned with what they will get for their investment after the project has terminated, from a corporate and personal perspective. If you're pitching a continual data quality programme for example it therefore pays to be clear about what benefits will be delivered at specific checkpoints. Explaining what your clever technology and robust methodology will do during this time is largely irrelevant.
Your business case should be simple and compelling. If you have gathered the necessary data, constructed a clear statement of benefits and created an engaging presentation there is no reason why your business case will not be accepted.
Remember that the business case is effectively a sales pitch and people buy on emotion but justify with logic so take ample time to discover those personal fears and goals that are important to your sponsors.
Sponsors generally prefer to move from a position of risk than to a position of gain when making corporate decisions. If your presentation can demonstrate how sponsors can both reduce personal risk and accelerate their goals whilst providing irrefutable evidence, you should be in a better shape to win the case.