How to Use Data to Evaluate TCO Before You Buy

Are you one of the lucky IT leaders whose budget is growing by 25% this year? If so, congrats. If not, or if you want to maximize those dollars, we have some data-driven advice to help you do more with less.

Whether you’re one of the fortunate few or the unlucky majority, it’s always a good idea to take a few extra minutes to get the biggest bang for your buck. And when it comes to purchasing new equipment and services, it’s critical to keep Total Cost of Ownership (TCO) in mind. So how do you make the right purchasing decision? There’s a wealth of knowledge to be gained by taking a closer look at the publicly available product data that’s out there.

Unfortunately, it’s easier said than done for a few reasons:

  1. True performance data is hard to find.

  2. If you are able to find the data, analyzing it to get insights is difficult. It’s either Pivot Table hell in Excel (or some similar desktop BI tool), or a long wait in line for help from analysts and the BI team to do a small report for you.

For example, industry performance benchmarks from spec.org can be used to identify the best performing processor based on max throughput. Here’s an example of an analysis we did recently on the top performing dual socket Intel CPUs from last year based on max throughput and max latency measurements.

Based on the max speed chart, you can quickly identify the top performer as:

Intel Xeon Processor E5-2643 v3.

Based on the max throughput chart, the winner is:

Intel Xeon Processor E5-2699 v3.

With this performance ranking and the vendor price quotes, you can now create your custom TCO comparisons tailored to your price/performance tradeoffs.

Another set of benchmarks that people rely on often are those of solid state drives (SSDs).

This particular benchmark gives a score to each product based on a weighted index of performance for sequential and random block read/writes to these drives. We found that OCZ provided the best performance for a sample set of weights that we tested.

Like all startups, we like to move fast. Both of these analyses took us fifteen minutes. To quickly gather these insights, we simply exported data from spec.org into a CSV, then loaded it into our product (using some of these tips, if you want to try it for yourself).

So what’s the key take-away? In an increasingly resource-constrained environment, IT teams are always on the lookout for creative ways to save costs. The best way to do this is with the help of data -- whether that’s performance data, failure data, usage data, etc.

For all the hoopla about self-service, shouldn’t there be easy ways to get answers from your data? We think so. At ThoughtSpot, we’ve built the world’s first search-based BI tool so that you can gather instant analytics from your data without depleting precious IT resources. With our Relational Search Appliance, any user can access and use company data to make better business decisions without relying on analysts or long training sessions.