A2Go Allows Smaller Companies to Enjoy the Benefits of Augmented Analytics
by Monica Jade Romeri, on Aug 6, 2018 9:54:35 AM
Large organizations in every field have already learned about the power of analytics. Smaller companies need to adapt in order to compete in today’s market. Most data science projects improve performance against the targeted metric by more than 10%. Some typical experiences we have personally observed with clients using analytics effectively:
- Online price analytics driving revenue increases of 20% and profit improvements of nearly 40%
- Inventory optimization analytics reduce inventory levels by 20%+, while dramatically improving customer service performance
- Demand prediction with 20% higher accuracy for many different companies in different businesses and over different time horizons
Among small and medium-sized companies the number of machine learning pilots in 2018 will be twice the level in 2017 and will double again by 2020. Leaders of companies of all sizes are learning and experimenting with machine learning techniques. No one can afford to be left behind.
The focus of Analytics2Go is on augmented analytics. Augmented analytics are similar to data science projects in that:
- Relevant data is collected from many internal and external data sources
- Data is quality checked before it is used
- Analytic applications help analysts and data scientists identify hidden value—so called “machine learning insights”
What is different with augmented analytics is that we go beyond the identification of insights. Our experience has shown that most companies struggle to apply insights effectively across their organizations. What may seem evident or even simple to a data scientist after a few weeks or month of work may not be easily applied to everyday decisions by the company’s workforce.
With augmented analytics, our “micro apps” are designed to support specific recurring decisions like customer quotations, short or long-term demand prediction, aggregated revenue prediction or trade promotion. Rather than expecting the decision maker to connect the data insights to the specific decision at hand, we provide a recommended decision or action based on the underlying machine learning insights.
Augmented analytics delivers optimized decision recommendations in real time to individual decision makers. We complete the insight-to-value process and allow individual decision makers to take advantage of analytic insights without mastering machine learning themselves. We maintain the apps and regularly monitor their performance. We have automated the process of converting data science insights to actions. Our clients do not have to hire their own data scientists. Sounds a lot easier already, right?
How We Work with New Customers
We help customers optimize performance by systematically enabling essential recurring decisions with analytics. We use an analytic horizon model to identify the important recurring decisions within the scope of our project. Below is an example of a high-level supply-chain analytic horizon model that was prepared for a company that rents electronic equipment. The horizon model helps us understand and get alignment with the client on the scope of decisions we will focus on when we are applying analytics to optimize the performance of their business.
The second step is to create an analytic data model. The main objective of the data model is to understand and codify the different internal and external data types that are required to support our planned analytic applications. The analytic data model is designed to enable effective collaboration with our client. We carry out client workshops in advance of any significant data science work efforts to ensure that we can take full advantage of available client insights and use them to guide our analytic work efforts.
The third step is to define the analytic design requirements for each decision. For each in-scope decision, we evaluate current practices to understand how analytics can improve decision making. We also define analytic design objectives. Below is an example that shows the analysis we need to complete in order to specify analytic micro apps effectively and efficiently for the client.
A2Go Augmented Analytics-as-a-Service Can Reduce the Overall Cost of Data Science Implementations by as Much as 90%
Our approach takes full advantage of the business expertise and insights available within the client’s organization. We have proven collaboration techniques that keep the data science initiative transparent and manageable.
Augmented analytics-as-a-service provides a business-like approach to applying analytics to any business model. Since we enable one recurring decision at a time, the initial investment is much smaller. Furthermore, values achieved from initial projects can be used to fund subsequent projects, creating an analytically-enabled continuous improvement cycle across the entire enterprise. The diagram below provides a recap of the advantages we offer our clients.
We think augmented analytics-as-a-service is the future of data science. If you are a small or medium-sized company, please contact us. We would be delighted to tell you more about augmented analytics can bring your organization success.