Lean Experiments for day to day problems
Many think Lean experiments are just for start-ups. The reality is that anyone can do them any time.
Lean experimentation is one of the most important skills to develop to foster innovation. Lean experimentation enables the ability to deliver new ideas which are more likely to succeed and avoid investment in ideas which are unlikely to succeed.
Recently, a Satalyst client was sharing some of the struggles they were having with improving their new market acquisition campaigns. They wanted to find a way to increase their acquisition rates and reduce costs, as well as the amount of time they were taking to carry out. Seems reasonable, right?
Introduce lean ideas
Satalyst proposed trying a Lean experiment and explained the benefits. The concept of spending less time and money to test an idea that could improve acquisition rates seemed pretty appealing, so ahead we went.
Traditionally, lean experiments start with an idea. Here we started with a problem that needed solving. The first thing we did was have an ideation session to decide on possible options. We came up with a range of ideas and through critical discussion, we converged on one. The outcome was a theory that we could overlay some of the client’s data with external sales data and try to show a correlation. If the correlation turned out to be strong enough, then we would use that information to improve the timing of acquisition campaigns and improve the conversion rates.
Test the theory
Now that we had a starting point, the focus was to challenge our hypothesis. We started by purchasing a small set of data and building an experiment to verify whether our idea was accurate or not. We used open source tools to overlay the data and then used RStudio to create some visualisations to show the data correlation.
The experiment looked promising, so we purchased a broader set of data and re-ran the experiment.
Running on the bigger data set also proved successful both at an aggregated view as well as when measured by area. Based on this, we were confident that we would be able to progress with the idea.
Enrich the results
At this point we wanted try to create some further insights, adding some extra value to the experiment if possible. Our client was able to provide us with persona data, so we overlaid that on our experiment findings to see what would happen.
This was a great extra dimension of data. We were able to find groupings of people that appeared to be more attracted to the client’s brand. These were coined the BFGs.
Where to target?
The final step was to determine where to target the next acquisition campaign. So we had a quick brain storm and came up with a simple approach. We decided that we could overlay the areas that the client was competitive, profitable and had a relatively lower penetration and which suburbs the BFGs lived in.
We fed this data into Tableau and created a map of Western Australia with a colour coded map showing the desirable suburbs. Gradually adding each dimension – profitability, competitiveness, penetration and finally BFGs, we ended up with a shortlist of 11 suburbs. With a little further discussion, we added weightings to each of the data dimensions and then created a final short list of 3 suburbs.
The client’s marketing department was very enthusiastic and will be testing the experiment’s findings with a campaign soon. The entire experiment took a few days to carry out and was an excellent opportunity to demonstrate that Lean experiments are a quick, cheap and an innovative way to test ideas and solve problems.
The experiment has also created the impetus at the client to challenge some of their existing processes and help provide demonstrable ways of focusing on time to market and capturing opportunities more swiftly. It was an excellent opportunity for Satalyst to help coach some of the client’s staff in thinking in new ways and providing a measurable, structured method to try out new things and approach problems differently.
Watch this space to see if the fairy tale experiment turns out to be as good as it promises…Categories: