What I learnt running two communities of practice in government and how they should be structured

Stephanie Marsh
6 min readDec 2, 2019

At GDS I was the head of user research and (performance) analysis, two disciplines working on gaining insight to make data driven, user centred decisions to design, build and deliver services and products that meet user needs. They are also two disciplines with different skill sets, particularly in government they require different expertise.

Performance Analysts straddle the Analytical and DDAT (Digital, Data and Technology) functions in government. It’s only recently been added to the DDAT job family. User research has been in the DDAT framework since the beginning (it’s not that its a competition, but they are thought of as different in government). Many user researchers have quantitative and qualitative research skills, others lean towards more qualitative or quantitative. There is some overlap between quantitative research and performance analytics; such as (but not limited to) A/B testing, and both have a lot of GDPR considerations. But mostly they require different and deep levels expertise to do well. What performance analysts do isn’t limited to A/B or reporting page views. It’s complex, technical and way beyond my understanding. User research isn’t just about ‘having a chat with people’ or deploying a survey to get quant data, it needs a lot of skill and finesse to do well.

These two disciplines need to work closely together to enable both holistic and lean approach to user insight. But that’s for another post (and conference talk). As the design disciplines are separate in GDS, so should research and analysis be, yes they need to work closely together but they need separate leadership. Service Design, Graphic Design and Interaction Design all have separate heads in GDS. Content Design and Technical Writing also have separate heads.

With technical experts at the top of a practitioner community, they are able to set direction, vision, strategy, support development and career pathways, make the right decisions big and small in terms of what is needed for the community in terms of people and the work they need to deliver for the organisation and ensure best practice is developed and applied.

I could do this for the user research community to a much greater level than I could the performance analytics community. I tried and wanted to support them equally, but I simply know much more about user research than I do about performance analytics, especially to the depth that it is done at GDS. I have gone back and forth on about whether it is simply my shortcomings that meant it wasn’t effective to have one head of two communities, but ultimately if you want an organisation to be doing innovative work at the forefront of any discipline you need a technical expert to run the community. And those technical experts running the various communities it takes to deliver services need to work together to ensure their communities can work together efficiently and effectively.

UCD team structures

Thinking about organisations more widely, to be a user centred organisation, user centred design needs a seat at the top table, increasingly government departments are appointing Deputy Directors of User Centred Design (or something similar). As I left GDS they didn’t have this position in the organisation. But they did have Deputy Directors of Product, Delivery, Software Engineer, Architecture. User Centred Design is as important as these other disciplines to build services and products that meet users needs, which means it needs an equally influential voice at the top of any organisation. This doesn’t guarantee the building of user centred, evidence based things but it goes a long way to keeping user centred design embedded in the development and delivery cycles.

I’ve been doing some reading about how to structure user centred design in organisations, it’s not an easy thing to do, there is no simple answer to this. Below is a summary of what I’ve found, for you to consider if you are interested in such things.

Building successful communities of practice

Emily Webber

https://emilywebber.co.uk/building-successful-communities-of-practice/

Here are a few thoughts based on what I learnt from this excellent book:

Allow social learning, learning by doing — social learning could in theory happen ‘spontaneously’ with people within the community helping each other out but in order to be effective you need a technical expert to organise this at the multiple levels at which it is need — community, programme, team, individuals. An expert head who knows what the community needs to focus on as a whole, who needs to develop in which areas and who has those skills already. If the head lacks technical expertise then its difficult to for the community to have a strategic direction of what skills it needs to develop and how, for example.

Sharing knowledge and building better practice — You need to be an expert to support the development of best practice to know what good looks like.

Building a community — you need to expertise in the field to ask the right questions to hire the right people.

Leaders need to be knowledgeable, passionate, respected and empowered. There was a lack of empowerment in the role (but that’s a story for another time). I was passionate and I had respect, what was disproportionate for me was my knowledge. I have an extensive knowledge of user research (I feel uncomfortable declaring this, but I have written a book on the subject so I guess I can/should say this with confidence). I have a basic knowledge of performance analysis. This basic knowledge was inadequate for the complex work being done at GDS and the complex issues that need to be tackled to make sure that people are properly enabled to do their work to the highest standard. This means I wasn’t able to support each community in the same way, not for the want of trying, I just didn’t have the knowledge to do it. This meant that I had to delegate much more within the performance analysts community because I didn’t have the expertise to do many of the things that needed to be done and make the right decisions.

People suffer when community leadership is an add on to another job and they become overcommitted and overstretched. Project work always ends up taking priority and the community struggles to flourish.

Who manages an internal UX team?

https://uxstudioteam.com/ux-blog/ux-team-structure/

Management of internal UX teams varies firm by firm. It often happens that UXers have two bosses: the Design or Research Lead of the company, as well as the PM. The Design / Research / UX Lead will also aim to have an overview of everyone’s work, making sure that it complies with consistency and high-quality across the whole organization. They will also oversee professional development, personally mentor, and organize internal events for them. UX design usually only appears in the C levels at bigger companies. In these cases, it belongs with the CDO (Chief Design Officer), also called Head of Design, or VP of Design. They oversee and evangelize design throughout the whole company, make sure it has the right leverage and resources, and that people understand the design mindset of the organization. They don’t actively design or research anything, holding a more representative role.

Good UX Management is User Centric

https://www.interaction-design.org/literature/topics/ux-management

At the heart of UX management is the enabling and valuing of UX resources, researchers, designers and design leaders. Good UX management means pushing the organizational maturity of your company in the direction of peak or consistent delivery of UX return on investment (ROI).

Good UX management puts users at the centre, includes validation and metrics, user centred design, support UX culture

Types of UX Management

UX managers often work in a user centered process, keeping an understanding of the users’ needs at the center of all activities. UX management comprises two dimensions — strategic and tactical. You can be adept at both, at different times.

  1. Strategic — You focus on long-term plans: (e.g.) funding models and UX evangelism. You may also become involved in UX process development, project selection, etc.
  2. Tactical — coaching and addressing everyday issues, working directly with UX designers.

TL;DR: To be a user centric organisation you need discipline specific heads of community and a director level person bringing it altogether at the top provides the strategic and tactical leadership.

--

--

Stephanie Marsh

Currently UX Research Operations Lead at Springer Nature. Wrote a book about User Research for Kogan Page.