Rob Smith, Managing Director of Blueleaf, discusses how his teams are experimenting with different ways of doing things in order to make a 1% gain each week. A gain is a better way of working and just a 1% increase each week leads to a 64% improvement over the course of a year.

Please find a transcription of Rob's video below:

1% may not immediately grab you as a world-changing number, but if you make a 1% gain every single week, meaning something improves by 1%, then the annual net effect is a 68% improvement each year.

The really important thing here is that so many departments and teams go about the 'doing' every single day and never really think about sharpening the saw and making things better. Whereas, if you focus on making something better every single week, it's amazing how much improvement you'll see, whilst not hindering the day-to-day 'doing'.  

Here at Blueleaf, we’ve introduced this ethos and we've been running 1% gains for over a month. We’re an agency so we do things slightly differently but we've improved the way we do scheduling, the way we sign our legal agreements, the way we quote change requests and more. Keeping up momentum is really important because now every single week we focus on what we're going to improve and at the end of the week we look at the results and learnings. I think the same thing can be applied to an ecommerce team / multi-channel team.

To explain the ethos a bit more, you basically use a scientific method to do it - sounds pretty boring but it really, really isn’t. Here's the 5 step process:

Step 1: Observation

We make sure we observe what's going on and look for problems or questions that need to be asked. You’ve probably got a huge list of things you want to improve, floating around in your head right now. You’ve probably been thinking about these improvements for a while. These are the things that should go on your list of observations. 

Step 2: Research

Then we do a bit of research into three or four of the things on our list.  Which ones will make the biggest revenue impact? Which ones are the hardest to do in terms of development time or design time or business change?  These are important questions to ask.

These experiments don't just have to be website changes such as changing the colour of this button or changing this product. They can be things like how to answer the phone in the call centre, how product packaging looks or how samples are sent out. All these things can be improved, nothing is sacred in the experimentation culture. So make sure that you take that to heart and think of all areas that can be explored.

Step 3: Hypothesis

Once you've done your research and picked an experiment, then you need to form a hypothesis. What will happen and why? How will you measure it? It's extremely important that you have a crystal clear way of measuring your experiment. 

Once you've got the hypothesis then it's time to run the experiment. 

Step 4: Experiment

Some experiments won’t run for a week. Some experiments will take longer to get statistical significance.  Think about how long it will take to make sure you don’t see 'fluke' results but are seeing definite, actual and tangible results that have been generated from lots of people going through the process. Now, that’s not to say that you can't start or finish an experiment every week, it may just take longer than a week to run.

Some changes are a no-brainer. Some things don’t need to be tested, they just need to be improved; for example, if you've got an image process for how to send images to your website, it's probably inefficient in some way so could be improved. Or a simple change that doesn't need testing is to move this image process in-house if it's currently outsourced. Whatever it is, the point is there's lots of things that don't necessarily need significant testing - they just need a focus to get done.

Step 5: Analyse

So once the experiment is done, of course you need to analyse the results and see if the hypothesis is true. Did what we said would happen actually happen?  If not, why not? Normally, if it hasn't worked, another area for improvement comes out in the wash.  You find something out and in turn, realise you should probably improve that.  So, put this on your list of  1% to gains to look at. And if it was really successful, well that's fantastic!

Keep a log of learnings, good and bad. When someone new joins the team you can share all the  experiments you've run, as well as your list of things to look at in the future.  A lot of questions they have may be answered through these learnings.

So, this is the 5 step process. What next? You rinse and repeat and you keep going.  You'll get  so much improvement over the course of a year and it doesn't take as long as you think to run these experiments. So give it a go and create a culture of experimentation. Lots of businesses are starting to run in this way so take it on board and see what happens to your business.