How to Design, Build and Execute a Data Governance Framework; interview with Guy Harvey

One of the most rewarding aspects of running Data Quality Pro is getting to meet practitioners who are maturing their organisational approach to data quality and data governance.

It is particularly rewarding when we meet practitioners who are achieving these objectives in the not-for-profit sector as in the case of our latest interview with Guy Harvey, a data governance practitioner who has been helping Sanctuary Housing roll out an entire data governance framework.

Guy shares a series of practical tips and advice for anyone looking to not only embark on their data governance journey but achieve it in one of the sectors not always associated with data quality and data governance initiatives.

Interviewee : Guy Harvey, Group Data Manager, Sanctuary Housing
Profile : Data Quality Pro | LinkedIn
Categories : Data Governance | Data Quality

Dylan Jones: Can you summarise the vision for Data Governance at Sanctuary Housing? 

Guy Harvey: Sanctuary’s vision is to create an environment that promotes data quality as an essential asset and where data adds value to its stakeholders. 

Dylan Jones: What were some of the challenges facing you before commencing with the Data Governance initiative?

Guy Harvey: The main challenge constantly faced is the perception of Data Governance itself.

The pushbacks vary from ‘been there done that’ to ‘that won’t work here, we’re different’ and ‘good luck, when do you leave?’.

Having to sell Data Governance is part and parcel of the job but I have found that if you set out to quantify, in no uncertain terms, the financial benefits Data Governance will bring, you are setting yourself up for a fall. I always sell Data Governance as a concept – a discipline for the 21st century.

Dylan Jones: That’s an interesting point as most organisations do rely heavily on ROI justification. How did you educate and achieve buy-in across senior management without creating financial justification? Do you find Data Governance is a term most leaders of a housing association are unfamiliar with?

Guy Harvey: I’m sure most, if not all, Housing Associations already recognise the challenges they face with the management of data. Sanctuary has grown massively through mergers and acquisitions. Sanctuary Executive recognised these challenges early and took action to create the post of Group Data Manager. By taking a personal one-to-one approach to sell the concept of data governance directly to stakeholders, buy-in was gained quite rapidly. But I would say adapting best practice processes and structures to meet the organisational culture and direction was a major enabler. In short, you have to make it fit.

Dylan Jones: Can you share an example of a one-to-one approach you adopted that ensured buy-in from a Stakeholder?

Guy Harvey: I sat down with a cup of tea and listened. Not once did I interview anybody. I spent my first 6 weeks drinking tea and chatting to people in the canteen. My approach was to have conversations with people and during that conversation bring out common questions. I probably never asked the same question in the same words every time but tailored it to the temperature of the interviewee. I sometimes had to tackle the same subject from different angles but always made a note of the answers.

Dylan Jones: What were the drivers for moving forward with data governance?

Guy Harvey: The primary drivers were tighter legislation and compliance and the pending move to a centralised ERP system. Previous attempts at measuring data quality had not been structured or been able to produce quantifiable results, so visibility and clarity of the true situation of data quality became a big driver.

Dylan Jones: What did the implementation of the Data Governance vision look like? What kind of structure did you adopt?

Guy Harvey: I implemented a typical Data Governance Board chaired by the COO with strict top down enforcement. Data Owners were appointed with accountabilities, they were not asked. 
Roles and Responsibilities were very descriptive and authoritarian as this matched the Corporate Governance approach. 

Data Stewards were appointed by Data Owners with clear responsibilities and tasks. Each Data Owner was tasked with putting together objectives, purpose and an action plan for their dataset with a timetable of achievements. 

Other members on the Board were involved, in an advisory capacity, in the areas of Risk, IT and the Enterprise Architecture.

Dylan Jones: How did you determine your Data Stewardship team? What qualities do you look for in your Data Stewards?

Guy Harvey: My message to data owners was always very clear and the same across the board. This is about getting the right people with the right attitude. The technology and the processes we can teach but you can not teach integrity and passion. Sometimes there were limitations on resources available but generally those put forward were of the right calibre. Describing what the end result needed to be helped focussed minds in the selection process.

Dylan Jones: How did you plan and execute your user involvement strategy?

Guy Harvey: Our engagement model for users was built around buying in stakeholders from the outset. The overall plan was to build a consensus around the current issues/challenges and then ask stakeholders to prioritise them in terms of pain. This had the added advantage of being able to use the data I collected to communicate directly with affected users and, when speaking with them, focus on their specific concerns and issues. Users felt they were being listened to and somebody understood what their challenges were and a solution was on its way. 

Dylan Jones: What is the current status of your program, what have you achieved so far?

Guy Harvey: Data Governance at Sanctuary has been extremely successful in identifying and qualifying the main data quality issues and providing focus on resolutions. 

Starting with a PoC enabled us to demonstrate the concept of defining data quality needs and solving root causes was a worthwhile and financially rewarding exercise. All areas of the Group are now heavily involved in Data Governance and defining/monitoring data quality. 

Dylan Jones: What kind of tangible benefits is the organisation witnessing as a result of these new improvements?

Guy Harvey: It’s always hard to be able to attribute a direct cost saving or benefit to increased data quality as there are always other factors and forces involved that also want a share of the cake, but what is clear are the risks that have been mitigated through a better understanding of the data and its purpose. Examples within the HR environment, Right to work in the UK criteria & RTI submissions have all benefited from increased data quality and continual monitoring processes. 



Starting with a PoC enabled us to demonstrate the concept of defining data quality needs and solving root causes was a worthwhile and financially rewarding exercise

— Guy Harvey



Dylan Jones: What final lessons from your experience can you share with anyone about to embark on a similar data governance or data quality journey?

Guy Harvey: A big lesson learnt was getting the right people on board in the first place. I was fortunate to find three very talented and experienced data quality analysts and this enabled me to hit the ground running with experience resources. I don’t think we would have been so successful if we had tried to skill up existing resources. 

Being clear about what is in scope. Data management is a large area and brings all sorts out of the woodwork when somebody’s name is in the frame. A strong sense of direction and strict control over scope enables you to stay focused and deliver what you set out to do. 

Don’t give up. The doubters are out there and may try underhand tricks and diversions, especially where hidden data factories are concerned. Be true to your beliefs and push hard, but also know when its a lost cause and you need to let go!

About Guy Harvey

After spending over 18 years designing and managing Business Intelligence and Data Warehouse environments, I saw the light (thanks John Ladley) and realised that most data quality problems being solved within Data Warehouses are caused earlier on in the process of data management. If I was to make an impact on improving information quality, I realised I needed to work closer with business managers and subject matter experts to influence and improve their data management processes and practices.