Want to pass the CDMP in ONE MONTH? Click here

How to kick start data governance the right way

data governance roadmap

Are you tasked with drafting a data governance roadmap, but currently staring at a blank page? Perhaps you’ve been given a budget or a new role in data governance, and now you need to work out where to begin. Daunting, isn’t it?

Maybe your firm has tried and failed with data governance before, and now you’re trying to re-boot things. This is even more daunting because you’ve got to make a plan to overcome the challenges your predecessors have faced as well as to deliver value.

If you’re scratching your head wondering what to do next, read on, as this is for you.

Creating a Data Governance Roadmap

data governance roadmap
An example roadmap – don’t blindly copy this

The first thing to realise is that most of the training and documentation out there is theoretical. It lists out the stuff you’ve got to do, and sometimes the sequence in which you do it, but the vast majority of what’s written is not practical. This is a challenge for you because theory won’t get you where you need to be. When drafting a data governance roadmap we need to identify the right sequence of tasks for your specific role and firm – which means there’s no “one size fits all” approach.

In order to know where to begin we will make 2 assumptions:

  1. You have already got funding – if not check out how to secure funding for data governance
  2. Your management is bought in/telling you to “get on with it” – if not check out how to secure management’s support for data governance

Step 1 – prioritise your work

Data governance is not a project. It’s a program. It involves cultural change which means you’ve got to change people’s behaviour with data. This means it’s hard – you can’t just pick data governance tools and hope they’ll solve things for you. They won’t.

In order to have any chance of success, you need to prioritise things. Not all data are created equal. Let’s focus your time, budget and efforts against the data that will give you “the biggest bang for your buck”. If you spread your bets and try to govern everything at once, you’ll run out of time and money before you deliver any value. Prioritisation is key to success.

How to prioritise your work

In order to focus your efforts, we need to pick data use cases. Some might talk about building data products. Either way, it’s important to identify which data you need to start with. Think through the following:

  1. What are the most important decisions made by the business?
    • These could be “bet the farm” decisions where the whole company depends on the outcome, or;
    • Small and frequent decisions like “should we ship overnight or is 7-day delivery OK for our customers?”
  2. Which data are used to support those decisions?
  3. What level of quality does this data have today?
  4. What level of quality do we need to make better decisions, and how much is that worth?
  5. How much effort is required to improve the data?

This should narrow your focus to a few critical use cases where improving data quality will show real business value.

Step 2 – setup/tune-up your framework

A data management framework lays out your target operating model – who needs to do what and why. You might already have this in place, but this is an ideal opportunity to tune it up or start it up. Three layers of decision making are usually put in place:

  1. At the working level, we need to establish data governance team(s) that will work on each of your use cases
  2. These are empowered by the most important layer – your data governance council
  3. At the top, we have an Executive Steering Committee to fund, steer the direction and break any deadlock

The data governance team

This is normally pretty easy to work out. You’ll need business SME’s that understand the data to play their part – usually as data stewards. They will be responsible for documenting the data and doing the work to produce data governance outputs. We need them to set expectations of what “fit for purpose” means for the data. This means they determine what high-quality data looks like.

They need data management support – technical people from your IT team that can write data quality rules, build data models, and change the underlying system architecture to ensure data is fit-for-purpose.

The data governance council

This is the critical layer. Data owners come together here to provide enterprise-wide oversight of the organisation’s data. The data owners feel the pain when data is bad, but also have the funds and resources to fix it.

The data governance team will report up to the council. They make recommendations and request support to fix problems. The role of the data governance council is to break silos and ensure data improvements proposed by the data governance team are workable and do not cause problems elsewhere in the firm.

The data governance steering committee

You probably already have steering committees that discuss IT projects, investments, or business challenges. We’ll hook into these existing committee structures and co-opt some of their time (maybe 5%) to discuss the funding and support we need to make data improvements.

Make sure you have a strong Executive Sponsor here, from the business. Many firms have now appointed a Chief Data Officer – but it does not need to be a formal title. Your CFO or COO could own this just as well.

Bringing it together

You take the use-cases from your prioritisation work (step 1) and identify whether you have the right people in the right roles. You may have execs that state they’re bought into this program, but now is the time to test this. Go back and explain:

  1. What their role is
  2. How much of their time you’ll need
  3. Which of their team members you’re going to need to work with
  4. How much time do those team members need to devote to this work, and
  5. How much time and money it’s likely to cost to start seeing results

Is this met with rejection, disengagement, or disbelief? Then you may not have picked the right use cases or your data governance business case was weak. Go back and repeat those steps, but focus on data your executives care about. Your data governance roadmap needs to start right in order to succeed.

LEARN MORE: How to create Ironclad Data Strategy your Execs can’t Ignore

Step 3 – ensure your team has the right skills to succeed

In a recent engagement, the team we worked with all believed that data governance was essential to their project. On average they rated the importance of data governance at just over 9 out of 10. When asked what data governance was, confidence dropped. The same team were only 60% sure that they knew their role in the data governance process.

This is to be expected. If you have brought business users into the project they might not understand what data governance is and why they should do it. Your IT team members will understand the data quality issues from a technical perspective, but may not understand the business impact. Both teams will probably look for a quick fix by buying some new technology or outsourcing the problem. Do not allow this to happen.

Change management skills

You’ve got to change behaviour to change data outcomes. The business team members have a day job to do. Data governance is just one of many tasks they need to do. If you want their time, you’ve got to make the case for change:

  1. Align the data use case value to items high on their agenda
  2. Prove the dollar value of data to their business outcome
  3. Clearly explain what you need their help with, how much time it will take, and why it is important

For the IT team members, you’ll need to take a different tack:

  1. Stop them dead in their tracks from immediately shopping for technology
  2. Highlight the fact we need business input to tell us the impact of bad data on business performance
  3. Bring in your business peers to help document data quality expectations, so they know what to shoot for

Start small, demonstrate value, build a library of “war stories” to talk about your victories, and improve incrementally.

Step 4 – review where you are, and refine

Prioritisation framework
Work out how easy it is to get data and how easy it is to use

Where have we got to in creating the data governance roadmap? Well, you now have a list of use cases for data. You can use these to document data quality expectations and prove business value. Rank the data use cases by business value (data risk vs data value).

You have also identified who needs to do the work, how capable and engaged the team is in the work. Rank the different use case teams according to the engagement and capability of these teams.

Lastly, you need to consider how complex the use case is. Does it require new technology or capabilities, or is it within your existing reach? Do you have the data today, or will you need to go and acquire (or improve) it before you’ll get value? Rank the use cases by complexity and availability of data.

Working out where to begin

Now you’ve got things prioritised, you can go about ranking them. Anything high value and high-risk score 7. If it’s high value/low-risk or high-risk/low value, score it 4. Low value and low-risk scores 1.

You can do the same for the other approaches – if your team has low engagement and low capability – give it a 1. High engagement and capability gets scored 7. If the data is hard to get and hard to use; 1. Easily available and simple to use; 7.

Add these numbers together to get an overall ranking. Whichever use case scores highest is going to give you the biggest bang for your buck and is where you start your data governance roadmap.

Final thoughts

With prioritised use cases and a clear understanding of what needs to be done, you’ll need to lay out the next steps. What gaps are there in your capability today, and how do you intend to close those gaps? Your use case deliverables will need to include:

  1. Data quality expectations – business rules and current capability gaps
  2. Data policy changes – how you intend to enforce the changes so people know the rules of engagement
  3. Master data improvements – what changes need to be made to create golden records?
  4. Metadata documentation – define the data you have today, what it means and what it can be used for
  5. Risks and controls – which data risks exist and how will you mitigate them?
  6. Data architecture changes – how might “the data plumbing” need to change to support your ambitions?

The depth and complexity of each output will depend on your use cases and data maturity today. As we said at the beginning, there’s no “one size fits all” approach – diagnose the problem your business faces with data, then lay out the steps you must take to fix that problem.

Next steps

If you need help writing your data governance roadmap, get in touch with the form below. However, if you want to go it alone, consider deepening your skills and knowledge with the courses on offer at the Cognopia Academy. Each course is laid out to be easy to understand, practical to use, and contains the templates and tools you’ll need to succeed.

Good luck!

6 Easy Steps to Win at Data Governance

Advance Your Data Career

Practical, bite-sized training that you can immediately use to engage your business and make more progress. Ditch the theory and gain the skills you need to advance your data career.

Learn more