Sentiment into Data into Pictures

Don’t get me wrong, the sentiment, the feeling, the strength of feeling, what you don’t hear about and sometimes the what-you-can-read-between-the-lines is all vital. If you can capture and convey that back to a client, it can be hugely powerful. But increasingly, when I’m running qualitative research, I’m turning at least part of it into data

Why? In part, because people ask. “They talked about work-life balance – how many of them?” “65%” is a better answer than “most”. And in part for my assurance, so that I can have even more confidence in my findings. When I re-review data, I tend to find that I was mostly right on first cut, but that a second more thorough cut produces a nugget or two of fresh insight. So I categorise comments into topics, and then I can group topics in themes.

And once I’ve got that, I’ve got data, and then I can present it in nice ways. Then we’re in a world of charts, which is very much my world. So then I might decide that

Might be better presented as

 

Straight away, you can see that “This” is nearly half of the total, which you couldn’t see as clearly before.

 

And then this might just add that bit more clarity and ease of access, and make it stand off the page a bit more.

The possibilities are endless, but they’re possibilities that I love to explore. It’s a long standing ambition of mine to present back a research project via one chart – just like my lecturer presented a term of lectures on the Properties of Matter with reference to one chart. When I achieve it, I’ll let you know…

Why use the big engagement survey providers?

I like engagement surveys. But that’s often been because the experience of employers using them has been so bad that they then turn to me. The surveys have taken a long time, a large investment, and it’s just produced data, rather than insight. Or, it’s highlighted an issue, but offered no thoughts on how to explore it more. Or, the provider has failed to analyse any of the free comments, massively devaluing the work.

I’ve been happy to step in and provide something more tailored and more useful.

But recently a number of people have approached me to create their own engagement survey. It seems their clients are not signed up to the necessity of benchmarking. They don’t see value in comparing engagement with other employers. They’d rather create something more bespoke, that will give them actionable results, whilst still being able to compare results longer-term.

And so I’ve stepped in here too.

I’ve given a few options. The first of which is to base it on existing, established engagement questions. But this feels like a bit of an unnecessary compromise. You’re confining yourself to fixed questions, without the benefit of benchmarking. So instead, I’ve proposed:

• Measuring my own set of (up to) 20 engagement factors. First their importance and then how well they are delivered at that employer. These factors are based on my research, what others measure and the work of Engage 4 Success. But they’re infinitely tailorable and addable-to.
• A purely free-text survey – usually asking what’s good/what’s not or what would you start/stop/continue.

Or, ideally, a combination of the two. By doing this, you’re getting your consistently-measurable scores, whilst also allowing more input - even-co-creation of solutions - by employees.

It still might only point at the issues, rather than supply the total solution. But surveys have their place, and for me this is a quick, cheap, bespoke way to approach it. (And I’ll always, always give you insight.)

 

What makes good research?

For a couple of recent pieces of work, I’ve had to think about what makes good research. Not so much how you conduct it, as how you conduct yourself. Not what you do, but how you do it.

And I’ve got it down to two things that research absolutely should be:

1. Simple 
2.  Commercial

Yes, simplicity is about turning complexity into clarity. But it’s about making the process simple too. There’s a temptation when you’re the expert to blind with science. But you can really demonstrate expertise by making a knotty question or a complicated process seem accessible and achievable. Experience of understanding what you can and can’t know will help.
There’s a temptation too to think you must get to all the answers straightaway. Often you just need to report the facts, neutrally. It may take time for the client to get up to speed with what you’ve found, so make it simple for them to access. Other times, once you have established the facts, the recommends start to create themselves.

None of this is to say that I don’t like to slip in a big word*– to have “axiomatic” edited out of my last report was a blow.

The commerciality is about making it a profitable piece of work, naturally; no-one buys a freelancer to lose them money. But it’s about making it actionable, for the client. Even if it can’t tell them “what next?” it must tell them “so what?” And – if having proven yourself, if having given them some strong affirmation, if having given them an unexpected insight– there’s an opportunity to suggest the next project, you should. If they take you up, that’s the best feedback you can get.

But commerciality is about being pragmatic; there are very few clients with infinite budgets and patience. Cost and time shape most projects. I believe you have to challenge that, but be more than prepared to just crack on if you can’t shift them. Otherwise you’re making it more about you than the client – and that’s the antithesis (* see?) of commercial.

What do you think? What qualities make good research?