Platform thinking can be vastly improved by applying the agile analytics architecture.
In our last post, we discussed applying agile analytics to the different building blocks of digital transformation. By applying the architecture of agile analytics, we were able to vastly improve the first building block of digital transformation: design thinking.
In this post, we will make the next building block of digital transformation significantly more efficient: platform thinking.
What is platform thinking?
Platform thinking is the brainchild of Sangeet Paul Choudary as initially expressed in a post he wrote on medium in 2013. Platform thinking states that every business is an engine. It needs to do a certain set of things repeatedly to create value. If you haven’t figured out that set of repeated operations, you probably haven’t created a scalable business yet.”
There are three basic steps to platform thinking:
- Creating the engine: Figure out the repeatable steps necessary to create value.
- Oiling the engine: Test and optimize the repeatable steps to refine the process.
- Stepping on the gas: Repeat these steps over and over again for more and more clients until you’ve reached scale.
Where does agile analytics come into play?
In the same article on Medium, Choudary lays out the three ways a business can run using this formula: Using humans, algorithms, and users.
An example of each is:
- HuffPo: Editors decide the content to prioritize each day.
- Google News: Algorithms decide the top news of the day.
- Twitter: Users’ tweets and retweets decide the top news of the day.
Agile analytics can optimize each and every step along the way regardless of the use case. In fact, in the agile analytics architecture, we like to fuse all three business cases together.
How do we do this? Let’s create a new content site. One that is powered by humans, optimized by an algorithm, and which takes into account how users are engaging with your product.
This would be a new type of content site, one that many publishers are currently experimenting with. Since machine learning algorithms require massive amounts of data to optimize, the site would take a mix of first and third party data to inform the human editors which content to create.
Applying agile analytics
In this example, our first party data is the site use itself. Our third party data would be general internet usage which you can get from companies like BlueKai and even second party data from Twitter.
So where is the user factor here? This is where it gets fun. Based on how users interact with the site, we could use cookie data to create a custom built site for each and every user. This would factor in data from other similar users using first party segmentation data with act alike audiences. It would also use your own browsing data to ensure the content you are getting is relevant to your interests.
Platform thinking does not disappear with the advent of agile analytics but it continually evolves. Instead of three separate thought processes, we have merged them into one and created a hybrid model based on platform thinking that is buildable, testable, and scalable. It is by all definitions, the Toyota Prius of the product design world. We have taken the best of what we had and merged it together to create a longer lasting, better performing, and more customizable platform using the same processes.