Confused about conversion optimization? You should be with all the experts sharing their opinions. After all, everyone is entitled to their own spin on the measure, test, analyze, and repeat optimization cycle.

When Jeffrey Eisenberg, my brother and business partner, recently spoke at a WAA conference, he said that you should be able to predict the direction of your tests better than 50 percent of the time. Imagine that, we’ve arrived at the point where we believe that conversion optimization should be less predictive than a coin toss – how sad!

With all the noise, the signal is hard to detect; oversimplification and overcomplication are leaving businesses confused about how to take optimization past the first few tests. The conversion optimization process is simple but not easy; the devil is in the details. The details involve a continuous improvement process that takes into account the realities of your organization and those of the real-time marketplace.

Only the most basic testing is about what variation(s) beat the control, because then all you know is what won. Testing needs to be about why something won. In order to learn, duplicate, and evolve, you need to understand your customers within their context and become more efficient at the real-time marketing tactics that today’s competitive marketplace demands.

Let’s start by better-defining conversion rate so that you’ll realize why the typical two to 10 tests in a month are not good enough and why you should be running dozens of tests a month.

Conversion rate = The number of people who take the action you want them to take divided by the total number of potential people who could have intended to take that action.

Not everyone will convert. Nevertheless, with the exception of bad traffic, everyone has some intention to convert. Our job is to offer every visitor an experience that fulfills their needs and leaves us a measurable conversion, even if it’s not a macro-conversion.

Please follow these steps:

Step 1 – Let’s Understand “Number of People”

Your conversion rate is influenced by how effective you are at attracting the right traffic and then helping them engage with your “funnel” to convert.

How are you bringing these “number of people” to your website? Do they all come by directly typing your URL in their browser? Do some search for your brand? Do some search for your category? Do some search for your products? Do some of those people come from organic search; paid search; e-mails; affiliates? Do these people come from different websites: Google? Bing? YouTube? Wikipedia? Twitter? Facebook?

Who are these people? Are they all the same? Do they have different characteristics, needs, questions? Do all these people have the same amount of product/service knowledge that you do? Are they all at the same stage in their buying process? Do they know you already? Or have they never heard of you before? Do you know your customer segments?

Do you launch new marketing efforts regularly? Are the efforts last week different than this week? Last month versus this month? Is there an important calendar event occurring (Christmas if you’re a retailer; Fourth of July if you’re a seller of flags; Mother’s Day if you sell flowers; etc.)? All of which may induce a “spike” in traffic that is different than usual.

There is no such thing as an “average” person. There is also no “average” marketing campaign. That’s why your average conversion rate is a rough indicator but also virtually worthless as a way to focus your conversion optimization.

You have lots of segments who come to your website. They differ by demographics, psychographics, behavior, or because they came in through very different marketing efforts. You can calculate a conversion rate for each one of these segments/marketing efforts and you should prioritize each one based on impact and value.

Step 2 – Let’s Understand the “Action You Want Them to Take”

If you are in retail, you want them to purchase a product.

If you are in lead generation, you want them to become a lead.

Are there no other actions that are valuable and contribute to the bottom line?

In retail, even in they don’t convert now, would it at least be more valuable to know if they added an item to your wish list, or subscribed to your newsletter, or looked up your retail store hours, or added items to their cart versus just bouncing off the site right away? What are you doing to turn that one-time customer into a repeat customer? Do they only need one product you sell or might they need different ones over the course of time?

In lead generation, if they don’t give you all their information and request to be contacted by sales, is it valuable to have them sign up for a whitepaper, or a demo, or your newsletter? Is it better to download specification sheets, engage in calculators, or print or forward pages rather than just bouncing off the website? These are all steps that move people through their buying process.

These are just some of your macro-actions. What happens when someone comes from one of your ads and gets to a landing page? Sometimes the action is one of those listed above, but what if that page is only meant to help your visitors to choose the right product or service and they still need to actuallyclick on the right one for them? What do you do to help them take that action and not bounce away? These are the micro-actions that need to happen from step to step in the potential customer’s journey.

All of these are actions we need to optimize. You can calculate a conversion rate for each one of these macro- and micro-actions, and you should.

Step 3 – Let’s understand “Total Number of Potential People Who Could Have Taken That Action”

What happened to the majority of visitors who didn’t convert? Have you asked and answered why they didn’t?

Did they land on your site incorrectly? For example, they typed in “shingles” into a search engine and they were looking for roof repair and not a skin condition related to chicken pox. This is obviously a disqualified visitor. Did they try to purchase from your website and something went wrong? Did they have problems accessing the information? Could they not figure out how to take action on your website? Did they not trust you? Did they leave with questions that were not answered? Did you answer their questions, but they weren’t ready to buy? Did you not instill a sense of urgency or desire in them? Did you not make them a great offer?

First , you need to optimize your website experience for these potential buyers through the hierarchy of optimization.

There are thousands of potential improvements to choose from. You need to prioritize these based on the level of impact that improvement can have and the resources available to execute them. If you don’t have a copywriter available at the moment, you shouldn’t focus on copy changes, even if it would be the most impactful. That may sound like common sense, but whenever I say it, it seems to catch many companies like a deer in headlights.

Your average conversion rate is the aggregated conversion rate of how well your website performs for each of your customer segments, and each of your marketing efforts for each of the actions you want them to take. You optimize your conversion rate by first focusing in on the elements that impact as many of these as you can, and then you have to work on these “micro-funnels.”

While you work on conversions, market conditions, competitive forces, and ad copy, your customers’ needs are changing. This is why you need to continuously optimize your marketing efforts. You can’t afford to have a “set it and forget it” mentality to your marketing.

If you only focus on improving a few landing pages here and there, testing a few variations here and there, and tweaking creative here and there, you will never reach your highest potential conversion rate!

Don’t worry, your customers won’t go unsatisfied. I guarantee that sooner or later your competitors will figure out how to satisfy your visitors’ needs. Hopefully that will motivate you to start getting your conversion goals on target by investing in a true continuous improvement process that asks “Why?” as much as “What?”