Data Insight in a Local Authority; what have we learnt so far?
Pye Nyunt, Corporate Insight Hub Manager, London Borough of Barking & Dagenham and an Agilisys Transformation Consultant (, 2 June, 2017

Pye Nyunt writes about his experience of setting up a council insight function.

In October 2016, we set up the “Insight Hub” at Barking & Dagenham Town Hall, with the aim to focus on making better use of the council’s data to understand customer needs, forecast future demand and design behavioural interventions.

There is a general acknowledgement across local government that there is value to be gained from better use of data. In just six months, the Insight Hub is driving behavioural change and reducing pressure on council services. Here are five lessons that I learnt in the process of setting up the Hub:

1) Start with people, not data. This is about creating a culture of data sharing. Some staff believe they are not allowed to share their data with other departments. This is mostly because they didn’t realise they were allowed to, their manager has told them not to, or they live under the impression that no-one else in the organisation will understand their dataset. Others have a very protective attitude towards “their” data. We are all on the same team. Staff should be empowered to understand that data should be shared unless there is a statutory reason not to do so before work even begins on your data asset registers. The duty to share data is just as important as the duty to protect its integrity.

2) Challenge what people understand by the terms “insight” and “intelligence”. On my travels to other local authorities, I saw many intelligence functions but very few insight teams. Most of the people I met in these intelligence functions were looking after performance reporting and developing dashboards. It was here I learnt that an insight team should always be outward rather than inward looking. The word “intelligence” has become synonymous with performance reporting and therefore looks at what the council has done rather than what we could do.

3) A data scientist is not a database administrator. Recruiting the right people into these roles is not easy. Some colleagues offered their thoughts; “why don’t you take [insert name] into your team? they are really good at extracting data from [insert council database]” – however, we’ve done this many times before in local authority i.e. centralising and then decentralising information support teams. My focus was on recruiting true data scientists; people who excavate data for discovery of new knowledge.

4) Data alone is not exciting, but using it to make decisions is. On a daily basis, many stakeholders ask the team “can you give me data on [xyz]?” our first response is always “what decision do you need to make?”. Therein lies one of the differences between a data analyst and a data scientist; data analysts focus on modelling data given to them, data scientists start with questioning the organisation on the use of its data. Whilst we hold many different datasets, we don’t get excited about data hoarding, we get excited when a decision, an intervention or a solution comes to completion as a result of experiential evidence.

5) IT is not analytics. I often found myself joking that there is probably more “analytics” software out there in the market than people who actually know how to use them. The day I updated my LinkedIn profile with my new Insight role, I had people as far away as Silicon Valley wanting to connect with me to sell their analytics products. There is great value in many of these technology solutions but frugality is equally important; there are numerous open-source products on the market that do the trick. A good data scientist can use open-source software to code their own scripts to automate various analysis, without the need for expensive IT.

The benefits that have been realised since setting up a true insight function include; securing valuable relationships with other local public sector bodies and socially-driven businesses, using data to attract social investment and most importantly has enabled the creation of a data-driven culture that has shifted people’s narrative from starting sentences with “I think…” to starting those sentences with “I know…”