Data warehouse and OLAP cubes were invented many years ago to allow reporting on large volumes of data. Technologies like incremental update and overnight batch update of cube data have been a de facto standard for years.
Some people would argue that reacting to yesterday’s data is basically like driving a car on a curvy road while only looking in the rearview mirror. More and more companies are therefore demanding real-time analytics in order to spot immediate changes in behavior or trends.
In-memory business intelligence
With Microsoft Dynamics 365 for Operations we have achieved in-memory processing of data, which means that you can build dashboards looking at transactional data which is updated when you refresh your screen. This is an improvement over previous static reporting. However, is this close enough to achieve the required insight?
If you are a professional marketer, you leverage different tools, such as Google Analytics or social media, to monitor the impact of traffic on your business when you push a campaign via social media to your customers. Social media campaigns typically have a short-lived impact on traffic and are quickly regarded as yesterday’s news. As such, it is becoming increasingly important to measure business impact in real time.
We have all seen stories go viral when a retailer posts something on their Facebook page. I remember a story from a grocery store which made an error and bought a pallet of yeast instead of a pack of yeast. The story went viral and was shared and liked on Facebook, but did it actually generate more in-store traffic?
What if you could measure the impact of social media on your business in real time?
Microsoft and Cortana / Power BI are setting new standards for what you can achieve on the BI frontend, and concepts like streaming analytics are becoming available for a broader audience at a low cost.
Power BI allows different kinds of data visualization. A simple gauge could make a big impact on store managers if they could see in-store traffic counters in real time.
In-store traffic counters
The store manager would be able to make quick decisions, such as: do I need to add more staff? Or measure the impact of the store’s morning briefing on how to change behavior towards customers.
Marketers would be able to follow growth in traffic in real time, and look at the impact from social media campaigns.
I have built a small model to illustrate the concept of stream analytics with Power BI, in which data is being fed directly from SQL to a Power BI dashboard. The actual dashboard is taking sample data from a Dynamics 365 for Operations database.
We can take the concept a bit further and look at Microsoft Dynamics 365 Sales, which offers broad functionality for customer service, campaigns and promotion management, without forgetting marketing automation. You will find many places where Stream Analytics and Power BI change the way you look at data.
- A customer service agent would be able to look at incoming calls and open a service request in real time, greatly reducing the time from registration to resolution.
- A marketer could look at click behavior for an email marketing campaign which shows how many people have opened an email or responded to a survey.
- A sales manager can see inbound leads or changes to opportunity ratings in real time.
Microsoft Power BI Stream Analytics still has its limitations in terms of data modeling and visualization. However, companies which understand the concepts of real-time analytics will have a huge advantage compared to their nearest competitors.