Blog post

QuickSight: A powerful BI tool for a fraction of the cost

Are you overpaying for your BI tool? Many competing BI tools offer similar functionalities, meaning other factors like accessibility and pricing are much more prominent when selecting your tool. And there’s finally a consumption-based pricing option that could significantly reduce your cost. 

The annual subscription license

Pricing of cloud web services is often based on consumption. For example, when signing up for your Snowflake On-Demand account to begin your cloud data warehousing journey, you are only paying each month for the computing and storage costs that you used.

Logistically, this is a headache since companies can’t know ahead of time the precise level of spend for applications in a given future month or year. However, the financial savings are considerable enough that consumption-based pricing is widely viewed as a consumer-friendly model. 

For the last few years, self-service BI tools have been priced under a different model: the annual subscription license. Regardless of whether a single user is using the tool every day or whether they have not used it for the entire year, their company is still paying the same price for their seat.

The annual subscription license model is most valuable for companies when their end users are using dashboards that change quickly and require frequent interactive analyses. For example, let’s say someone builds dashboards for the business managers of trading desks at an investment bank. These users use multiple dashboards every day to follow trading activity across asset classes and counterparties while responding to daily ad-hoc questions from their compliance department. Those users would surely get their company’s money’s worth from the cost of their annual license seats.

Conversely, annual subscription licensing might feel more frustrating to clients when most of their end users are using dashboards that change slowly and require less frequent interactive analyses. There are many scenarios in management and board reporting where the dashboards only need to be updated and used a few times per year.

Part of the allure of self-service BI is that these tools can be used to analyze most questions from any structured dataset, including questions and even entire data sets that you didn’t even know were coming when you purchased your BI tool’s license. A company’s healthy data culture should promote wide use of analytics tools. That means the annual subscription licensing model of BI will continue to be valuable and serve most organizations well for the foreseeable future.

But what about external deployments?

Some key assumptions of the above thought process break down when we focus our study on BI deployments for end users residing outside of the company. Consider an example: A company conducts quarterly surveys between thousands of organizations to measure the health of working relationships. It’s critical for this company to show multiple stakeholders at their 25,000+ clients the aggregated results of their survey data in a self-service fashion in order to cut down on customer service requests.

Since these surveys are only conducted quarterly, there are 50,000 users who require BI tool access but only need to log in once per quarter when their data is refreshed. Since this external BI deployment has no new analyses or different data sets expected in the future, there will be no “additional use cases” for these external users to make use of their annual subscription license.

For this reason, annual subscription licenses are less valuable to most external deployments. BI companies know this and for the last few years they have allowed a price break on externally deployed licenses – often around 25%.

Pricing

Let’s compare different pricing models for the above scenario. Another BI tool’s license costs are on their website, so we’ll use that for our subscription example.

Externally deployed viewer subscription license: $108 per year

108 x 50,000 users = $5.4M per year

Since the above pricing is so outrageous, most BI providers also allow a second pricing option for large deployments known as a core-based model. In this model, there is no license limit on the number of users because you are licensing based on the power of the hardware supporting the BI deployment. However, that hardware can only support a certain number of concurrent users at once before you need to buy more hardware to support more concurrent users.

With a properly architected dashboard, this often comes out to around $1k per concurrent user. Unfortunately, in our example, all 50,000 users tend to access the survey data application on the same four days of the year, since they are excited to see the results when their new data comes online. A good rule of thumb is that when planning for concurrency, about 3% of your daily users will be clicking around the application within a few seconds of each other. We need to plan for 50,000 users x 3% concurrent x $1k per concurrent user = $1.5M per year.

Enter the consumption-based pricing model

Since most of their users are only logging in four times per year and at the same time, this survey company would get the most value from a consumption-based pricing model. That way, they are only paying for the small amount of time that they actually spend using the tool rather than the large percentage of the year that they do not use it.

There is only one BI tool that currently offers this pricing model – Amazon QuickSight. Its pricing is public, so let’s finish our math and comparison.

Full pricing: 

50,000 users x 4 sessions per year = 200,000 sessions  

For QuickSight, 200,000 sessions = $57,600 per year

Summary of estimated pricing for this external use case:

• Consumption-based pricing: $57,600 per year

• Core-based pricing: $1,500,000 per year (26x more expensive)

• Annual subscription-based pricing: $5,400,000 per year (94x more expensive)

Conclusion

Price should not be the primary factor in selecting the BI tool for your external deployment. Ultimately, end users need to use dashboards that they can immediately feel comfortable with, otherwise they don’t explore their data and the expected ROI is lost.

Still, the industry for self-service BI tools has matured to a point where there are fewer and fewer true differentiators between tools. Of the differentiators that do still exist, even fewer of them apply to external deployments where end users are less likely to be creating their own dashboards.   

In a world where many things between BI tools are equal, price is a factor that will increasingly become the deciding one, especially in situations like the one we outlined here where the price difference is so extreme.

Keyrus’ QuickSight Service Delivery Program helps clients leverage QuickSight to see tangible business results. Learn more about how we helped Premier Brands of America use QuickSight to quickly analyze sales growth of products and brands.