Data analysis: capability crisis?

I have a pure maths ‘A’ level and a third of my degree was in Economics.  I used to be a banker, had a career in venture capital, and I work at NPC. So I must be really good at statistics, quantitative data analysis, and excel databases, right?

Wrong. When looking survey results I hit my limitations after about 30 seconds. I just about know the difference between percentages and percentage points, and that you need good samples sizes to produce reliable statistical analysis, but beyond that?

I can do financial analysis and financial modelling until the cows come home. So know all those excel functions. And I’m pretty good at thinking about strategy, organisational issues and so forth. But ask me to perform any kind of database function and I run screaming to the nearest NPC geek.

And NPC is crawling with data geeks, right? Well, not quite. We have an excellent mix of skills—consulting, financial, research, life—but only a  specialist minority can really handle the complexities of sound statistical data analysis.

And here’s the rub. NPC is arguing for more and better data to be available to charities from government, and to be collected by charities so that we know what works. But having collected the data, how many charities will be in a position to analyse it? If statistics are a minority sport even at NPC, what chance is there of a small to medium-sized charity having the right skills? The skills gap is daunting, and I don’t mean any disrespect to charities who might be excellent at service delivery.

The solution could lie with the development of online tools, such as NPC’s well-being measure, that does all the maths for you. We need more such measures and I predict that whoever corners the online charity data processing market will do well. Perhaps not in 2013—NPC’s experience with the well-being measure is that we are still in market-building mode – but perhaps by 2020?

In the meantime, I cheered up hearing an excellent presentation from Aimia, the ‘loyalty management’ people who own Nectar, and all the data harboured about us on our Nectar cards. Their analytics function knows more about our tastes and preferences than we know ourselves. Aimia ran a weekend ‘DataDive’ last September for charities such as Place2Be, Keyfund and Oxfam, to help them analyse their data. They discovered some fascinating insights about what works, and are considering running more. Aimia are also experimenting with the idea of longer-term pro bono partnerships with organisations that need their analytical expertise. Kaggle is another option for accessing global analytical skills at an hourly rate.

Maybe more corporates with data analytics departments such as Aimia could offer their expertise to charities? Much more valuable than an away day painting an office!


  1. Lies, damn lies and statistics? Well not exactly. I am market research by professional, trained in one of the top UK MR agencies, and yes I got Statistics O level, used stats in my Psychology degree and hold the MRS Diploma. So here’s my two-penneth for what it’s work. For over 20 years I’ve been analysing survey data and for 12 of these I’ve specialised exclusively on helping charity clients, because I love contributing to this sector. The secret is to have the statistical grounding to design a survey with robust sample design -otherwise there is a danger of “garbage in garbage out”. My caution would be beware the total DIY approach – although I fully appreciate charities have modest research budgets, so sometimes charities to hire an expert (which is where I come in!) to advise on the design of the study and interpretation of the results – even they do the actual interviewing in house. And to have some statistical knowledge to understand how to interpret survey findings is vital else you can get hung up about small differences in percentages which are actually meaningless and due to no more than ‘sampling variability’. I seek to take the mystique out of sampling and survey statistics and present data in an accessible way – vital for charities who may embarking on their first ever research assignment; and crucially to provide insights and actionable recommendations, not just more data! And that surely is the aim – whether it’s about understanding beneficiaries needs or supporters’ motivations. As for truly demonstrating impact, I admit that one has me challenged – surely the Holy Grail. If anyone ever cracks that one, they will be the ‘geek pre-eminent’!

Leave a Reply

Your email address will not be published. Required fields are marked *


Back to top