How to: Avoid licensing trap and customer frustration when selling inexpensive Microsoft Power BI plan

All of us know about the perfect Power BI tool, kindly offered by Microsoft as low as 10 USD per user per month (or even for free in some desktop installations). Many of our customers use it with NAV and some already with Dynamics 365. I guess NAV + Power BI can provide users one of the best experience on current market. However it is not so cheap in reality, as users who operate with that licensing model will sooner or later definitely face a problem of limitation for 1 Gb of data storage on importing a single dataset into Power BI.

Within the typical scenario, customers who are paying low fee of 10 USD per month sooner or later would run into this limitation as the quantity of data would grow with time. (Note here that reaching this 1 Gb limit by customer does not mean that their business grew and they became a bigger customer – it is just growth of the data with time passing).

Imagine customer had completed Power BI project for 5-10 users and starts using the solution. Soon it became their business critical software for analysis and decision making. But suddenly customer runs into a limitation of 1 Gb that prevents him from using the application. Then customer has two options:

1. Upgrade to Premium package for some 16k USD – which is not the case for this customer as he is paying as little as 10 USD per user per month with very few users and simply cannot afford such upgrade.
2. Purchase separately Azure Data Analytics for some 600 USD per month on top of his Power BI licenses, plus the work to set this up with a partner. This is also quite a poor substitution of initial 10 USD per user per month.

So, customer gets quite unhappy with the product. It is a lost-lost-lost case for all the parties: customer starts to consider an option to quit using this product, and neither partners nor Microsoft would get money. What is the probability that customer would upgrade to Premium? In any case, he would not be satisfied with the fact that he needs to do that.

With existing model, when I’m explaining this limitation to customers, they usually decide not to go for that solution as they are afraid of that trap. I also can’t keep this in secret as sooner or later customer would run into the limitation of 1 Gb, and would curse us all.

So, to my mind, the correct option would be to let customer purchase more storage when it is needed. Customer is eager to pay e.g. extra 10 USD for extra 1 Gb (or whatever ratio it would be), but considering he can expand that when needed. This would be win-win-win for all the parties: customer will be happy as it is reasonable and logical, and both partners and Microsoft will get paid.

I had addressed my question to Adam Wilson, Group Program Manager of Power BI solution, meeting him on Microsoft Inspire 2017 in Washington. Here is the summary:

Unfortunately, there are no specific plans to provide to customers the option of expanding the size of the data storage. The main reason for that is architectural limitations. It is not only about the storage itself, but rather about the computing resources behind the story. Premium package of Power BI runs a completely different architecture in the background, having all the dataset imported into memory – this enables the performance when building the reports and processing the data. This consequently requires that all this memory is firing all the time, thus makes the option expensive. The 10 USD per user per month option does not store the data in memory. Simply expanding the storage would result in the significant performance degradation – so this option is simply not scalable with current design.

The options for small customers are to split the reports into different datasets, to aggregate historical data, to use direct queries – thus by all means perform architectural optimizations in the very beginning of the project.

Summary: There is no licensing and architectural solution so far to eliminate the gap between the cheap solution and premium option, due to current architectural limitations.

With all that spoken, I was left unsatisfied with the outcome. I hope that is some near future Microsoft could provide a reasonable scalable option for the customers to stay cost-effective with their solution.