The Love and Hate of Maturity Models

Maturity models often fall into love-em or hate-em categories.  Executives often love them because they want to understand where they are and where they want to be.  Non-executives typically hate them because they worry they're being evaluated and criticized, not to mention having their jobs in jeopardy.  And consulting companies love them, because, you know.

The biggest challenge faced with maturity models is that the decision to use a maturity model often comes from the executive level and is not always communicated down.  So, when a consulting company comes along and starts asking questions at lower levels, they're talking to people who don't understand why the consultants are there in the first place.  The first step, therefore, becomes calming people down, communicating the purpose of the maturity model, and explaining what the outcome will be.

How do you calm people down?  You need to be honest and explain to them what's going on.  I've never been involved in a maturity model exercise where the outcome was to fire people.  The majority of the time, the people in place are not the people who created the mess in the first place, so you can't blame them for the situation they're in.  Most likely, the end result is to determine the path forward, which will involve change.  Change can be scary, but change doesn't have to mean letting people go.  The change can be around helping people understand what they need to do to be more focused on data management activities to achieve better outcomes for the business.

There are a lot of maturity models out there.  I've worked with large data management maturity models, but also smaller maturity models focused on a specific issue the business has, such as data governance or data quality.

Generally, I see maturity models with 4 or 5 levels.  It really depends on how granular you want to be.  I've also seen maturity models with numbered levels and named levels.  For instance, we might refer to "level 1" or "reactive".  The difference here is really just perception.  Sometimes, people who are already uncomfortable with the idea of maturity models will be even more uncomfortable being told that they're "reactive", so calling it a number is easier on them.  Alternatively, if they're already uncomfortable, it might not help at all being told they're at level 1 and there are 4 levels.  Again, communication is important so that you get their participation in the activity and explain that you're not there to hurt them.

I'm going to focus here on data governance, since I specialize in that area, but maturity models are similar for other areas.  This is an example, but doesn't necessarily apply to everyone since it includes levels based on being a global company.  It's not intended to be the definitive data governance maturity model.  There are plenty out there, all with the expectation that they're not necessarily out-of-the-box applicable to your organization.  Get a maturity model, analyze it, and modify it, if you need to, to fit your organization.  You might have to change levels, names, job titles, and tasks.  To help in identifying tasks, it can be easiest to classify into standard groupings of people, process, and technology.  Remember though that I said this was about data governance, so I'm not going to talk about things like data quality tools because those would fall under a data quality maturity model, or a metadata tool that would go on the metadata maturity model.  Likewise, these would all be subsets of a data management maturity model, which is too extensive for this article.  These are all subsets of a data management maturity model.  I'm stressing the difference between data governance and data management.

Level 1 - Reactive

People - There are no formal Data Owners, Data Stewards, or Data Governance Organization in place.  There's no formal ownership, so issues are fixed as they are identified, often by users or IT.  There's no formal change management process in place to track changes.

Process - There's no roadmap or plan in place.  Limited governance processes are in place, meaning that issues are addressed individually and randomly, as they occur.  No metrics are in place to measure success.  Processes are not integrated in the enterprise data management processes.

Technology - No tracking or reporting exists to track and report on data issues.

Level 2 - Controlled

People - There are limited Data Owners and Data Stewards in place, often at the level of individual applications.  Changes are tracked on an individual and sporadic basis, and often not shared sufficiently.  A Data Governance Organization exists at a local level only, so not integrated at the enterprise level.

Process - A roadmap or plan exists at the local level.  Governance processes exist at the corporate and business unit levels for critical business functions, but do not exist for departmental applications.  Limited metrics are tracked.  A subset of processes are integrated at the enterprise level.

Technology - Workflow tools exist to track data issues at a corporate level.

Level 3 - Proactive

People - Data Owners exist for each business unit.  Senior management reviews data governance progress and activities are carried out at local levels.  Cross business unit Data Stewards manage and track changes to data.  A Data Governance Organization manages processes across business units.

Process - Governance processes exist across the company.  A formalized data strategy is in place for each business unit.  Metrics exist across business units.  Most of the enterprise data management processes are integrated at the business unit level.

Technology - Workflow tools exist to track data issues at corporate and business unit levels.

Level 4 - Predictive

People - The Data Governance Organization includes members for both business ownership and technology enablement.  Data governance is a part of all levels of the organization and across all business units.  Metrics are tracked.  Governance exists at the global level, overseeing coordination across the company.

Process - A global roadmap or plan is in place and supported by global leadership.  Everyone has access to well documented processes that are implemented at all levels.  KPIs are in place at the global level and drilled down to regional and local levels.  Fully integrated enterprise data management processes are in place at the global level.

Technology - Workflow tools exist to track data issues at all levels of the organization.