A to B: How Testing Drives Search Marketing

September 23, 2019 Dylan Brooks


Pencils up.

If you’re like me, those two words bring you chills.

When I was growing up, they signaled that a test was coming. The kind with Scantron sheets or light blue booklets filled with lined paper.

It didn’t matter what I absorbed in class or how much I’d prepared. If I couldn’t put down the right answers from memory, my grade was toast.

While I tended to do well in school, I didn’t embrace testing as a student. I thought it was a poor way to measure success.

But now, my thoughts are different.

As a digital marketing professional, I recognize testing is my ticket to ride. Independent insurance agents would be wise to embrace it too.

Here’s why.


Test For Success

The digital revolution has transformed marketing for independent insurance agents. It used to be difficult for agents to tie campaigns to bottom-line results. But that is no longer the case.

Insurance agents can measure performance on their websites with precision today. And with these capabilities come new marketing measurement opportunities.

Some consider A-B testing to be the most revolutionary of these opportunities.

A-B testing assesses the viability of a feature on a website or mobile app. This can be a webpage, an icon, or a similar digital element.

During an A-B test, some consumers see an original version of the feature. This is version A. Others see an altered version. This is version B.

Once the test is complete, agents can compare the data to decide which version won. In other words, they can see which version best drove business results. They could use this data to fine-tune their marketing campaigns.

A-B tests often cover a minute detail, such as a button color or a few words of website copy. But even though these changes seem small, they can impact bottom-line success.

This is why businesses in several industries have embraced A-B testing. One of these businesses is a large company based in Mountain View, California.

You might have heard of it. It’s called Google.


When Testing Meets Search

As the world’s largest search engine, Google has an unwritten mandate. Provide the best answers to what people are looking for online.

This ethos has driven the company’s search ranking philosophy for more than two decades. But it’s also allowed Google to dive into an A-B testing culture.

Google’s commitment to testing is equal parts self-interest and self-preservation.

Like an amoeba, the internet is evolving. And the needs of searchers evolve with it.

It’s difficult for Google to keep up with this moving target. So, the company tests and iterates.

Any time you search on Google, you’re taking part in a massive A-B test. Google runs several simultaneous tests a day worldwide.

Some searchers might see a new feature on search engine results pages (SERPs) as part of these tests. Some might see a SERP with a new layout. Some might see search results laid out in a different order than usual.

In the past, Google engineers and user experience designers would design the tests. They would then feed them into the company’s ranking algorithms for implementation.

But today, the algorithms run on machine learning. So, some elements of these tests might even get launched without human intervention.

The algorithm implements tests at random. So, it can be difficult to recognize them while they’re happening. This also means there is no such thing as one true ranking anymore. Even if your website tends to appear in the top spot for a certain search phrase, it might not show up there for everyone.

This is a significant problem for businesses large and small. Search engines drive about 70 percent of all traffic to websites in the United States. So, these inconsistencies make it difficult to measure success online.

But you don’t have to bury your head in the sand. You can use these same tools to your advantage.


Embracing The Testing Mindset

The test-and-iterate model has proven to be effective. Now, it’s time to put it to work for your agency website.

Install analytics tracking on your website, if you haven’t already. Two great options for this are Google Analytics and Adobe Analytics.

Collect 30 days of website data in your analytics platform of choice. Then, take a close look at your website experience reports. These are dashboards and flowcharts that show how web users interact with your website’s pages.

See if there are any troubling patterns to address. Are consumers heading to your car insurance page but not clicking on the quote button? Are they heading to the employee directory page without leaving their contact information?

These are two potential problem spots that could emerge. You might find others that are specific to your website’s design and your business workflow.

Once you identify these bottlenecks, consider tweaks to those areas. Then form a hypothesis for each of them.

A hypothesis is a statement that you believe is true, but is unproven. Your hypothesis could state that a certain Get A Quote button doesn’t stand out from the page background. Or that some wording on the website is ambiguous.

Once you have your hypothesis in place, set up an A-B test to prove it. Show some users a version of the webpage without the changes, and others with the changes.

There are two ways to do this:

  1. Simultaneously. Under this method, you clone the website page you’re testing. Then, you make the changes you’re testing on the new version of the page. After this is set up, you’d use third-party software to determine which visitors see each version of the page.
  2. Chronologically. Under this method, you’d leave the website page untouched for a week or two. Then, you would make changes to the page and leave them up for the same amount of time.

The simultaneous testing model is most ideal, as it shows an apples-to-apples comparison. But the chronological method takes less time and money to put into action.

Once testing is complete, look at the data for each version. Does one group interact with the page your testing better than the others? Are they more likely to take any action that meets your business goals?

If so, move forward with the version users favored. If not, look for another element to test.

By investing in this process, you can improve your website’s user experience. This can improve your lead conversion rate. And it can help your search rankings.

Search engines are looking to provide the best resources for searchers’ needs. So, user experience is essential. Good user experience positions your agency as that top resource for a related search.

Taking the time to test bottlenecks on your website expedites this process. It can give your agency a competitive edge that drives bottom-line results.

So, don’t be like the younger version of me. Don’t dread testing. Embrace it.

You’ll be glad you did.

About the Author

Dylan Brooks

Dylan Brooks helps ITC clients improve the visibility of their agency websites, working directly with them to improve their search engine rankings. Dylan has a bachelor’s degree in communication from the University of Miami and an MBA from Southern Methodist University. Dylan has extensive experience with writing, strategy and marketing analytics. In his spare time, Dylan enjoys cooking, participating in 5K races, and spending time around Dallas.

Follow on Twitter Follow on Linkedin More Content by Dylan Brooks
Previous Article
ACORD Forms Update (September 2019)

Next Article
10 Minutes to an Updated Website
10 Minutes to an Updated Website

Here's how to keep your insurance agency website updated in ten minutes or less.