Experimenting is essential for learning

When it comes to business, as with most things, you know less than you don’t know. So how can we shift this balance? how can we learn? how can we improve?

My approach is to learn by doing. This will also involve making mistakes, which should also provide us a chance to learn.

However, simply doing, isn’t quite enough. You need to come up with an idea of what you think the answer to the question is and then test that theory. And then test and test and test some more.

While at Cadimage we introduced a [paid] student version of Archicad and over a number of pricing experiments we ended up providing the student version for free and finally reached the point where student usage took off.

At the time we did this, it wasn’t necessarily as formal an approach as designed experimentation, but early results were a failure so we tweaked and tweaked until it worked.

In hindsight the answer is as clear as daylight and ultimately providing free student versions, free training for staff and students, and then sponsoring the NZIA Student Design Awards was seen as way of building the commercial side of our business – why we tried to sell something to students is beyond me but the point is the 3-4 years of trial and error led to the answer and not only solved its own situation but had a far broader impact on our business. In my opinion we could never have foreseen this outcome as our focus was very narrow.

Technology has come a long way since we did these early experiments, and there is a huge range of software to support conducting and analysing experiments and the data they generate.

Our Student pricing can be viewed as a fairly simple experiment, but equally I think that is also key to being able to learn. Don’t try to test everything at once. Focus on pieces that will appear to have the biggest impact.

When experiments became automated via technology they began as A/B Testing – ie testing two options. Software allows many more options to be tested simultaneously but I think A/B testing truly encapsulates the approach to take.

Obviously you can experiment with everything from pricing to email subject lines and contents. But stay focused – especially if you are a small team – and don’t experiment with everything just because you can. You simply won’t have the bandwidth to assess the results.


Leave a Reply

Your email address will not be published.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