Benchmark Definition For Analytics, Website Marketing & Statistics

Benchmarks are performance measurements taken before pitching or beginning a new web optimization project.

Benchmarks are important for both projecting the ROI of an initiative you’re pitching, or for measuring the success of a project. 

By establishing accurate benchmarks before a program begins, you can show all of the progress that you have made over set time frames, which will help other members of your team better understand the progress of a project.

Why Benchmarks Are Essential For Website Optimization

Let’s say that you want to make your website perform faster – user experience will be better, rankings will improve, what could go wrong?

Unfortunately, most people in your organization will not know what minifying & grouping JavaScript & CSS files means, but you know that it is important.

Before making the optimizations, you would want to be sure that you had benchmarked data on the existing page load speed, as well as the number of relevant resources & how long it took each to load.

Assuming that you were making no other optimizations to the site content, you would also want to know the current rankings, search impressions & traffic data.

In the case of a UX improvement, you would certainly want to include the qualitative data with the quantitative traffic data in your reports as well.

Once you make the changes, you can then measure the impact over time that these optimizations made.

Proper Benchmarking Strategy In Analytics & Optimization

Keeping the example above in mind we can review the best way to handle benchmarking & measurement strategy.

We initially looked at the real-time code execution of the files that we want to optimize, as well as the overall page’s load speed performance that they are accessed by.

This way we can have an apples to apples comparison of the newly updated content once it has been published live.

Always be sure to have times & dates, as well as other notes such as what the optimizations were that you did included with each measurement.

For the analytics data we need to think a little more.

We need to find a sample size of data that is large enough to be meaningful, while also being small enough to report on sooner rather than later.

With this in mind, analytics platforms offer a “Compare Periods” function which can be beneficial in this instance.

By choosing the pivot date as the date that the optimizations went live, you can report on prior periods easily at whatever cadence you decide.

This is beneficial in the near-term where you’ll want to be watching each day, as well as later when the cadence changes to weekly or monthly check ins for status reporting.