Solving the problem of dodgy data

In 4 weeks, MES launches v6 of its Membership Database system for Foundation Trusts.  Preparing to roll it out to our 100+ clients has inevitably got me thinking about v1, and the start of it all. 

Back in 2006 when we started spec’ing it out and development began, we were approaching the project from a very different perspective.

Our USP for the system was that we were coming at it from an electoral point of view, not a purely technical perspective.  The technicalities and functional capabilities were of course important, and needed to be better than elsewhere.  However, we were not a software house focused just on the jazzy features.

MES was born in the Electoral Reform Services stable.  Elections are their thing – quality, impartial, accurate project management & delivery of ballots and elections, by post, online, on-site or telephone.

Now, elections are not easy things to run.  They are riddled with risks and pitfalls – which is why so many organisations choose to hand their projects over to a company that concentrates solely on them. 

Perhaps slightly selfishly, in MES we quickly started to think ‘how could we help improve ERSs job’? 

IMPROVING DATA QUALITY

If we could do this, and look after our clients’ data before it gets to election stage, then ERS’s checks would be faster, data could be prepared easier and costs could be lowered for the clients.

So ‘election-ready data’ became, and remains, the first bullet point when we explain our NHS Membership Database to the world: v1 to v6 and beyond.  Keeping it clean, cleansing regularly, building in automated checks… It can make a huge difference to our parent company’s ability to deliver.

So being in my early 40s, male and into my music, here’s my top 5 of what can go wrong with dodgy data.

1.    Delays to the election timetable.  Quality data is imperative, if it takes longer to prepare then delays are more likely.  At best a delay could cause irritation and suspicion amongst candidates and voters, at worst there could be legal consequences.

2.    Given this, it causes stress and extra work for all concerned.

3.    Pressure of time means errors are more likely.  So the extra work caused by bad data in the first place could in fact increase the chance of mistakes being made.

4.    Duplicate records, missing records, misplaced records – all these can cause serious issues.   Bereaved member records can also cause distress to family members on one hand, but also weaken the robustness of the data.

5.    A complete re-run is the worst-case scenario.  If data has been so problematic and if checks have not picked errors up, a re-run can be expensive, embarrassing, wasteful and impacts on turnout.

Luckily, ERS has in place quality checks at all stages to minimise the risks as far as possible.  When running an election with data arriving from an MES database system, this risk is further mitigated and processes speeded up. Better for all!