Bad data in always means bad data out💩➡💻➡💩

If we want data-led organisations, we must start by supporting frontline staff to capture better, more accurate data.

Tom Spencer
2 min readDec 1, 2023

I’ve been talking dashboards recently and I was reminded of a presentation when I first joined Camden Council, that included several emojis. There was a push at the time to add more data fields and options to an IT system used by people in the SEND service. The idea was that with a more detailed understanding of the needs of children and young people, we could commission services that better met these needs. This makes sense. What didn’t make sense was making the IT system more complicated.

Adding more complexity doesn’t help improve the quality of the data going into the system. It probably makes it worse. Too often we focus on improving the IT system and data analysis tools, when more needs to be done to support the people inputting the data to feel motivated and confident to capture good data.

Here are a few things I think can help:

🔎 Focus on measures that matter🔎 — There is always a clammer to capture more data but most of the time there are a handful of key metrics that really matter. If someone wants to add something, then take two other things away! This also makes time for your frontline teams to do the relationship elements of their roles.

Explain the ‘why?’ to the people that capture data❓- Too often we ask people to capture data without explaining why they are collecting it. It might feel obvious but frontline staff often aren’t invited to read reports or be engaged in commissioning services. Their data is invited but they aren’t! Without understanding why it matters people aren’t motivated to capture data effectively.

💻Provide the support and training people need 💻 — IT departments perform an increasing number of functions. A vital one is helping people use the systems they use for their day jobs. The training and support for people in frontline roles need to be much better. We are told that the big IT systems are ‘as intuitive as Facebook’. They’re not. People need training when they start and support they can pull on when they need it.

What else needs to happen to improve the quality of data we capture?

The title says bad data in always means bad data out. Underneath that there is a picture of an emoji poo then an arrow pointing into a laptop. There is then an arrow coming out the other side and another emoji poo

--

--

Tom Spencer

Helping public sector and community organisations deliver great outcomes for the people they serve