Avoiding client rejection later, annoy them now…

One pattern we saw in early SenseMaker projects was challenge and (on occasion) rejection of results by the end client. What I’d failed to do was spend enough time with them, ensuring that they “owned” the signifiers, saw the data coming in, got to play with SenseMaker themselves. My tendency had been to let them sit back early in the process, for fear of disrupting an early relationship with them.

I don’t do that any more.

Tomorrow, we’re going in to a client to generate material from which we’ll create the signifier set for a new project. And then I’m back in on Friday afternoon to finalise and sign them off. Only once we’ve got a signifier set that has been agreed by the client will I even start the translation process (this one is being translated into multiple languages, but thankfully all share the Roman alphabet).

Under time pressure, it can often be easier to start translation on elements of the set while you’re still finalising the client’s native language version. (“Oh, we’re only still working on these last couple, let’s translate the other seven…” Version control becomes a nightmare – which of these various files contains the right set to work on now…)

The other piece that we don’t allow people to wriggle out of either is testing the website entry. Not just trying a few test entries, but really going through, checking the options. (Ron has been excellent at this – I used to think pedantic, but don’t any longer – it’s been too valuable.) And then getting full entry from the field researchers, producing a test dataset and looking at it carefully.

Checking, for instance, that filters aren’t coming out as questions (bad xml coding on my part, that one), that odd alphabets are coming through legibly (not helped by invisible formatting characters in some word processing programmes), etc.

In an ideal scenario – and where the project is critical I’ll push harder – I’d want a full test session to be run by researchers (Anecdote Circles, etc) and then have a test dataset and conference call debrief afterwards.

From what I’ve experienced, some clients may be unhappy you weren’t gathering material the minute they said “yes” – but they’ll be more unhappy if you find later in the process that their researchers have been gathering the wrong material…