Product ManagementRivi

Reports are Dead. Long Live Embedded Analytics

By Rivi Aspler

In the old days (just a few years ago …), data-reports were a mandatory feature to any business application. Offered with no other choice, end users had to settle for reports (html, pdf, excel etc. ) that contained tabular-formatted information.

Embedded Analytics is different. Usually, these are graphically-formatted pieces of information, residing not only within the application, but in the relevant page (in-context). Embedded analytics can guide users or recommend specific actions to perform. e.g.  If data trend is A, you may want to do the following…

A good example of a company that embeds analytics into each of its screens is Workday. Below you can find one example of Workday’s Expenses Management page.  As one can see business understanding and actions are driven by these graphically based data objects , which can be investigated further more, in real-time, using drill down capabilities.


The great thing is that embedded analytics  are becoming so popular that a simple Google search will yield many more great examples. Just try it out….

Best Practices for Embedding Analytics into your Applications

Regardless of the industry you are in, the embedded analytics should have the following characteristics:

  1. In Context – No longer an isolated  reporting area. Embedded analytics appears near the data that requires a decision. From an application perspective, this means you have to design your pages to have embedded analytics near important pieces of data.
  2. Decision Supporting – As product managers, this is the most interesting job that you will do; re-thinking  your application using an operations point of view, i.e. “based on this data, I would like to change this value from A to B.”
  3. Visualized – The default view of embedded analytics is graphical. Nevertheless, a user can still change the view type to show data in a grid.
  4. Real Time – Goes without saying. Data should be up-to-date; best if it’s real-time. And yet, in most cases there is little harm if the data is a few minutes old.
  5. Deep Data Comparison – Embedded analytics offers deep Drill Down capabilities, enabling a user to easily see the weighting factor that sets a trend-line. For example, a specific person/group/location/geography that influences average-data more than others.
  6. Socially Enabled and Actionable – Easily done, make sure your users can share/email/sms/export to Excel/print etc. the graph and the data behind it.
  7. Impact Analysis — if you change the value of parameter A, the value of parameter B would change in +X% – Still not there as a mandatory feature but certainly a key differentiator if you are able to offer it to your customers.
  8. Industry Benchmarking – Same as above, not there as a mandatory feature but if you are able to offer it (or build it if you are a start up), this could be a major key differentiator, enabling your users to see what similar businesses are doing, hence change their operating decisions accordingly. A simple example of this benchmarking issue can be the percentage of new customers acquisition in your industry in each quarter.

Adding embedded analytics into my company’s applications, I can admit to the R&D and management reluctance to deal with these types of requirements; effort and budget wise.

On the other hand, seeing the users satisfaction and their growing demand for more embedded analytics will ease your persuasion efforts with both R&D and with management, getting the required budget to make your business application as rich as possible with Embedded Analytics.


Tweet this: Reports are dead. Long live embedded analytics. http://wp.me/pXBON-3Td #prodmgmt #roadmap #analytics

About the author

Rivi is a product manager with over 15 years of product life-cycle management experience, at enterprise sized companies (SAP), as well as with small to medium-sized companies. Practicing product management for years, Rivi now feels she has amassed thoughts and experiences that are worth sharing.