Content Marketing Magazine | News, Tips, Insights & Thought Leadership

The Science of CRO - The Case for Testing

Written by Isaac Rothstein | Apr 4, 2014 5:00:00 AM

A paper was published in Feb. 2013 by Harvard Business School professor Sunil Gupta and two of his colleagues entitled "Do Display Ads Influence Search: Attribution and Influence in Online Advertising." The questions the study considered were straightforward - do display ads make a difference when it comes to conversion; how much of a difference do they make, and for how long?

Most online marketing professionals will take one look at the title of the paper, and immediately answer "yes, of course". Some may take a further few seconds to think, then say "yes, but not in certain niches." Such answers will of course be based on experience.

So, what is the answer, and how does it affect conversion rate optimization (CRO) testing? Before we get to that, it's best, perhaps, first to consider a different question. What are the chances of these "gut reactions" being correct? And how do these "gut reactions" influence the decisions that are made about online businesses? In addition, we need to consider how these feelings effect understanding of how customers make buying decisions.

CRO Helps With Making Reliable Business Decisions

When it comes to CRO, there's a common misconception. When entrepreneurs are first introduced to the idea of CRO, they tend to simplify it. They understand CRO as:

- informing decisions, based on data, or

- optimizing specific aspects of online marketing to determine which out of two or more decisions performs the best, and generates the most revenue

On the surface, both these simplified definitions of CRO are correct, at least on a technical level. However, if you look under the surface, you'll understand the psychology that is used to influence your audience; to persuade visitors to your site to perform specific actions. The actions your customers perform relies upon them to make specific decisions.

Effective conversion rate optimization helps you to understand what goes through a customer's head while they are perusing your site. You want them to make certain decisions, such as signing up for a newsletter, clicking on a certain link or actually making a purchase. CRO is the strategy that persuades them to make these decisions, resulting in your chances of obtaining a increase in revenue more likely.

In theory, it doesn't sound too difficult. You use the data you've gathered, along with techniques such as A/B testing, to uncover the correct strategies you need to adopt, then apply them. But there is another layer underneath this common sense approach that's not so readily apparent.

Humans Do Not Always Take The Efficient Route When It Comes To Making Decisions

People tend to think that making decisions is a linear process. Because of "A", we make choice "B", and then influenced by choice "B", we make choice "C", and so on. Unfortunately, the human decision-making process is not actually so clear cut. Professor Jonah Berger, of the Wharton Business School, published a study entitled "Decision Quicksand" which goes some way to explain the how humans do actually make decisions. 

In short, it can be difficult to understand what is really important when it comes to decision-making, as humans often feel paralyzed when making difficult decisions due to the huge number of options that have to be considered, and the importance of making the correct decision.

Think about how this complexity affects Internet marketers, and online customers. There are thousands of studies that help us understand how the human mind functions, but it's clear that understanding which decisions are considered important, and how those decisions are finally arrived at is difficult. When it comes to CRO, only the data that is gathered from structured testing can help paint a clearer picture.

Common Sense to the Rescue?

If we don't have data to hand, we tend to make assumptions. We either trust our "gut instinct" without validating it; blindly accept "best practices", or take advice without really testing it. Each of these is a bad idea, and here's why:

- Using Common Sense

There is a simple idea in psychology that's known as "availability cascade." In the human realm, there's a certain body of knowledge that's referred to as "common sense." It starts with a simple solution to a complex process. This gathers momentum until it begins to be perceived as wisdom. The problem with this is that very rarely is a "common sense" idea backed by testing. It simply becomes the accepted solution as people who are trusted consider it to be a good idea. But then ask yourself this: How many "common sense ideas" when it comes to online marketing have you adopted, only then to discover it has zero impact or even a negative effect on your business?

- The Wisdom of the Wise

Once someone has achieved "guru" status, everything they then say is taken as gospel. There are numerous issues with this. No doubt engaging an expert can help you achieve appreciable results, but by relying too much on expert advice, you end up flattening your own learning curve. In addition, expert advice is often adopted without verification as to whether the advice is correct, and applicable to your particular business model, audience and clientèle.

- Intuition

As we go through life we gain experience, and via that experience we learn to recognize certain patterns. When we encounter these patterns, we tend to address them in the way we think best - in layman's terms we call this behavior "hunches."

Hunches are unpredictable. Some lead to success, others to catastrophic failure. There are two approaches when it comes to responding to hunches: The expensive, time-consuming way it to waste effort and resources in finding out whether your hunches are correct. The cheaper, swifter way is to use your hunch as a hypothesis then use structured testing to get you to where you really need to be.

Will Testing Overcome These Limitations?

You can easily find yourself discouraged by the body of research and anecdotal evidence. You may consider it impossible for you to be able to really get to know your customers. Getting them to take the actions you want them to take is the result of months of hard work, and thousands of tiny factors. Finding influence levers and using them can take significant investment. You've discovered that your tried and trusted routes of using common sense, the advice of experts and your intuition are not the best ways of finding successful strategies for your business.

The good news is that CRO is here to the rescue. CRO is the answer when it comes to addressing all of these issues. CRO allows you to:

- Get To Know Your Audience

The more data your collect about your audience, and the more you analyze their actions, the better you get to know them - and the more you know about them, the better placed you'll be to identify, predict and influence their decisions.

- Define Your Buying Cycle

The buying cycle is the process your customers go through when considering a purchase. CRO will allow you to find out which stages of the cycle are failing, so you can attend to them and up your revenues as a whole.

- Use Common Sense, Expert Advice and Intuition To Create Hypotheses

Instead of using these issues to define your processes, use them to kick them off. Creative ideas can drive tests that really get results.

- Structure Tests and Conduct Them

With testing you can really discover which of your ideas work best when it comes to connecting with your audience. It will help you to implement ideas that'll see your customers help you to achieve your business goals.

- Implement Your Findings and Increase Profits

Use all of the above to improve your conversion rate. Be patient, and you'll discover that your revenue will increase organically over time.

Looking Ahead

You may remember that right back at the start of this article we discussed Dr. Gupta's research on the effectiveness of display ads when it comes to attracting customers. In the report, Dr. Gupta and his team focused on a company in the financial industry. They determined that by concentrating on simple metrics such as click-through rates, companies were over-simplifying the decision-making process.

Later in the report they went on to offer a more detailed analysis which helped them draw up some effective conclusions, while emphasizing that the context in which they worked was very specific to the needs of a bank hoping to acquire customers.

There's no doubt that suggestions from experts can lead your business up some very interesting and potentially lucrative paths, but without looking ahead and seeing where that path leads, you could be walking into a minefield. The key is this: the only way to know what really works for your business is to test first. If you want your business to be successful and to achieve its objectives, then there is one thing you need always remember, and that's to always be testing.

This is a guest post by Isaac Rothstein. Isaac is an analyst at Infinite Conversions, a digital focusing on improving website’s real-world financial metrics through conversion rate optimization.