Analytics rock stars: Roll with the pros to rock your analysis:
The first breakout session at the EMEA Summit I attended was about tips and tricks for working with Adobe Analytics & Discover. Brent Dykes who led the session, started with breaking up the analytics profession into two different areas. There’s ‘Setup-land’ which focuses on translating business goals to analytics requirements, tagging the website and building reports. The other area he defined was ‘Action-land’, this is where you focus on the optimizing – which is usually the fun part, getting value from your insights.
So the session initially started with some really basic tips on using Sitecatalyst, dashboard and such. I was still looking for the ‘Rockstar’ part, but about half way through the session it got more interesting and advanced. I ended up leaving with two key take-aways from this session that I wanted to share with you:
One of the experts on the panel, Willem Corbijn (Philips), showed us how they had standardized dashboards within their organization. Most of you must be thinking – every dashboard serves a different business goal, how do I go about standardizing them?
But the way Philips had standardized the dashboards, they still left room for the necessary customizations.
The biggest advantage by standardizing most components in the dashboard is that it makes it easier for the business to understand and compare the data. This way fewer questions arise, which leaves you with more time in ‘Action-land’. Apart from that it can also be a huge time-saver on producing these dashboards, which again leaves the expert with more time creating actual value for the business.
The rule is “Standardize when possible, Customise when needed”, with 4 key questions you should ask yourself if a customization actually is required.
1) Will new reports provide more insights than standard reports?
2) Are they sustainable once adopted?
3) Are you making up for a lack of data or processes?
4) Will your customizations prevent internal benchmarking?
Scatterplot quadrant analysis
When Brent had the floor again, he showed a very powerful technique to break up data in different types of categories. He used a scatterplot for this, in which he plotted different products based on their volume versus. success. In this example he used conversion rate (y-axis) versus the product views (x-axis). By dividing the plot in four different quadrants he was able to create some powerful insights on each product.
For instance the products in the quadrant with a high conversion rate but low amount of pageviews were defined as the hidden gems. These products are very successful on converting but are currently getting a low volume on the website. Possible actions for this quadrant could be to feature the products on the homepage or setting up a campaign to give them more exposure.
The products in the top right quadrant, having a high success-rate and high volume are defined as the rockstars amongst the products. They convert well and at the same time receive a large volume of traffic. These are the top products for a commerce website and you’d want as many of your products to end up in this quadrant.
The bottom right quadrant was defined as the pigeons. These products get a large amount of exposure, but simply do not convert that well. To upgrade these products to the ‘rockstar’ quadrant you can consider setting up an A/B test for the product pages. You can then play around with different elements on the website, such as the images, text or pricing to increase the conversion rate.
Although the session started off a bit basic, it ended up with some great insights. If you want to check out the full session, you can do so here: