web analytics

By avoiding our online marketing confirmation bias or Oedipus Complex we lay the path to greater insights and greater profitability.

This is a guest post by Craig Andrews.

Sorry, no salacious stories about sons killing their father and sleeping with their mother. But in Oedipus Rex, Sophocles had insights about Online Marketing more than 2 millennia before the internet. The play’s hero had a tragic flaw that plagues all of us. It is our Confirmation Bias or Oedipus Complex.

“My poor children, I know why you have come— I am not ignorant of what you yearn for.” – Oedipus Rex, Sophocles

The Greek tragedy opens with Oedipus, the King, telling the people he knows the source of their pain. It turns out that he is the source of their pain.

At some point, we all approach our internet visitors the same way Oedipus approached the Thebans. We are absolutely convinced we know why our website visitors have come and what they yearn for. But if we don’t continually test and challenge our assumptions, we are just as guilty as Oedipus.

The Origin of Our Complex (our Confirmation Bias)

We are passionate about our business and our customers. We invest time in serving and understanding our customers. When this makes us over confident, we can miss important insights. Even in looking at our website analytics, we can find data that confirms what we believed. This “confirmation bias” can cause us to quit scouring data when we find the data that supports our hypothesis. We must press further.

As an example, a client was convinced their customers wouldn’t visit the website using mobile phones. Yet 5 months after launching a new site, 1/3rd of their website conversions were from mobile phones. In another instance, we saw sustained double-digit organic traffic growth and assumed it was due to Google.

Turns out Google traffic was dropping while traffic from other search engines was rising. Haunted by the words of Oedipus – “I know why you have come” – I was wrong. Continuing with the wrong assumption without correction would have resulted in additional lost traffic from the world’s largest search engine.

No Fate But What We Make

Oedipus thought his ruin was the product of fate. We should not. Rather, we must continually test and challenge our assumptions.

In the digital world, every customer touch offers an opportunity to learn more:

  • Website visits, email campaigns & pay-per-click advertising all enable you to study customer behavior
  • Free analytics tools provide extensive demographic data and even a degree of psychographic data (interests & hobbies)
  • The technology customers use to access your site provides valuable insight into your customers’ context and experience

Don’t Gouge Your Eyes Out!

“You don’t know whether something will work until you test it. And you cannot predict test results based on past experience.” – Eugene Schwartz, author of Break-through Advertising

Take action and put a plan in place. Effective plans should include 2 types of testing: testing a new hypothesis and challenging an existing belief.

New Hypothesis Challenge an Existing Belief
The call to action isn’t clear The increased conversion rate is because of recent changes
Item X is causing friction & will be corrected by doing Y Our website visitors prefer using PCs
Site Navigation isn’t clear Most of our mobile visitors use iPhones
Our visitors don’t identify with a specific graphic Our website visitors are in a specific demographic
Our current graphic isn’t objectionable, but distracts from the conversion Our visitors are looking for bargains

New hypothesis testing is familiar – it’s classic conversion optimization. But testing to continually challenge existing beliefs is what really helps to avert our Internet Marketing Oedipus Complex. Our existing beliefs, held too tightly, can get us in deep trouble.

Our team recently improved a client’s home page bounce rate 10%. Immediately, conversions started shooting through the roof. Initially we thought the changes to the home page were an overwhelming success. It seemed logical. Change followed by success, right? We chose to challenge our belief. After digging into the analytics, we discovered a few things:

  • Indeed, reducing the bounce rate improved site conversions
  • Only about half of the increased conversions were attributable to the home page changes
  • A different high volume landing page (Page “R”) had also seen an increased conversion rate
  • No changes had been made to the other page (Page “R”)
  • The increased conversion rate on the 2nd page (Page “R”) appeared to be seasonal

We found our answer in analytics under Behavior => Site Content => Landing Pages. This lets us track conversion rates based on the first page visited on the site.

Home Page Page “T” Page “R”
Conversion Rate Improvement 48% -8% 39%
% of Total Site Traffic(as a landing page) 57% 15% 28%
Percent of total Conversions(as a landing page) 26% 44% 31%

Our focus was on the Home Page and Page “T” but ignoring Page “R”. Again, it seemed logical. The Home Page receives more than half of the site traffic and one of our home page changes directed traffic to Page “T.” Our confirmation bias initially led us to ignore the 39% improvement on Page “R” where we have a significant number of conversions. But without changes Page “R” seems to be seasonal.

Now We Know (More) Why They Have Come

This discovery put a finer point on what we reported to the client. Instead of promising the client continued conversions at the new rate, we showed them how some of the higher conversion rate would be seasonal. Now the client is happy because they have permanent changes that increased the conversion rate. They also have realistic expectations for the future.

As a conversion optimizer, we now have a new hypothesis to test. There may some seasonal surge in this one category. We record that and study it next season. If it is indeed a seasonal surge, then we can tune the website and email campaigns for this newly discovered seasonality, giving the client a strategic advantage they didn’t have before.

This is the power of challenging existing assumptions. In addition to perfecting our view of reality, it opens new opportunities. With these new opportunities, we can find new and innovative ways to increase conversions in our digital media channels.

Do you yearn for success?

When Oedipus discovered the reality of the situation, he gouged his eyes out so he wouldn’t have to see it. To be effective, we can’t fear the truth. We must pursue understanding even if it results in abandoning a strongly held belief that we want to remain true. That means we need to:

  • Commit to a thoughtful, structured and methodical process of testing
  • Regularly compile a list of hypotheses and rank them
  • Regularly identify currently held beliefs and rank them
  • Perform systematic testing on these hypotheses and beliefs
  • Always learn from each and every test – especially if the test appears to be a failure

When Oedipus proclaimed he knew what the Thebans yearned for, he was speaking from his gut and intuition. Intuition is a useful tool, but let’s let the data proclaim what our website visitors are yearning for. By avoiding our confirmation bias or Oedipus Complex we lay the path to greater insights and greater profitability.

About the Author

Craig Andrews is the Principal Ally and founder of internet marketing agency allies4me. Andrews brings extensive scientific and marketing expertise to allies4me. Over the last 25 years, his experience has spanned search engine optimization, internet marketing software, biomedical and semiconductors. Andrews is backed by a team of marketing allies who support start-ups to Fortune 500 companies.

In recent years, clients have been seeking guidance from allies4me on Social Media strategies. Rather than jumping on the latest hype, Andrews sought to understand Social Media through solid metrics across large data sets. The result is an unconventional and insightful approach to Social Media. Testing and data driven decisions advise all areas of allies4me work. Solid metrics and disciplined parsing of data is where allies4me clients find results.

You can connect with Craig Andrews on Google+ and LinkedInYou can find allies4me on LinkedIn, Google+, Facebook, and Twitter.

For more information about testing hypotheses that convert, check out the latest Conversion Scientist’s Podcast,  “Writing Test Hypotheses That Make You Money”.

It’s time to stop boring people with how good your open rates and click-through rates are. Tell them what each and every person on your list is worth in dollars by measuring Revenue per Recipient (RPR). When you track the results of your emails down to the dollar, you track your own value down to the dollar.
From Marketing Land: Marketing Power Processes: Tracking Email To The Dollars by Brian Massey

Subscribe to the Podcast

Tweetables: Click to Tweet

Power Process: Ignore Email Open Rates & Click-Through Rates
Revenue-Per-Recipient (RPR) Ties Marketing to the Money
Like trees in the winter, it’s important to prune and shape your email list.
Most email clients now show the subject line and the beginning of an email in the inbox view.

We all rely on our analytics a great deal. Whether it’s checking out your growth in readers or customers, or seeing where your traffic is coming from, website owners like knowing they can see real measurements for their site.
But what do you do if you analytics seem….well….. off? How do you trust your reports when they aren’t producing correct results?
Brian’s new post on Search Engine Land is all about debugging your Google Anayltics installation to ensure you are getting the correct data from your analytics. He will take you through the steps he uses with his clients to get accurate information from Google Anayltics.
Read his full column on Search Engine Land and start debugging your analytics today.

Feb 11, 2013 12:05 pm


  • PayPal is a preferred method of payment for many of your visitors. Even if you’ve got a merchant account and gateway all setup, you should consider PayPal as an alternative source.
    However, measuring transactions through PayPal is problematic. Or it was.
    Here is an excellent post on how to configure your PayPal account and Google Analytics for full-transaction tracking, includkng Ecommerce tracking. – Brian Massey

by: Brian Massey
read more


Google Analytics Tips: 10 Data Analysis Strategies That Pay Off Big!

Jan 02, 2013 11:43 pm


@Avinash Kaushik has a unique ability to make analytics human. I don’t share many analytics posts with you, as I don’t want to scare you off. But I fear I may be underestimating you.

Here are ten very good ways to get to know your visitors through Google Analytics. I believe you will be energized and excited if you open these reports in your own Google Analytics account.
This is a great way to start appreciating your visitors in ways that will make your site more successful.

by: Brian Massey

Karon Thackston: Phrasing Discount Offers for Maximum Results by Getentrepreneurial.com

Dec 29, 2012 01:47 pm


How you phrase a discount is a powerful way to increase conversions. Some visitors do not like to do math, or will do it wrong. Therefore, offering 20% off is less effective than save $18. However, high discounts (50%, 90%) may draw buyers more powerfully than the dollar value.

You have to find out for yourself.
Karon offers a nice list of alternatives and some links to research for your enjoyment.

by: Brian Massey

The Top 5 Website UX Trends of 2012 | UX Magazine

Dec 29, 2012 01:40 pm

While good UX (User eXperience) does not always translate into higher conversion rates or revenue per visit (RPV), these trends point to excellent hypotheses for what MIGHT increase the performance of your site.

  1. Single Page Sites: Simplicity is often a great way to increase conversion rates
  2. Infinite scrolling: Consider this for category pages. I haven’t tested this yet.
  3. Persistent top nav: I am very curious to see if this increases CR and RPV. Let me know if you’ve tested it.
  4. Web 2.0 Aesthetics: I hope this includes the rotating banners at the top of so many sites.
  5. Typography Returns: Your message is the most important part of your conversion optimization plan. Typography can help… or hurt

by: Brian Massey

Even though the election season is over, I’ve opted to extend the debate and move it into the realm of online marketing. In 4 Anti-Science Marketing Attitudes That Keep Us In The Stone Ages I compare the anti-science efforts used by politicians and their campaigns to the anti-science efforts stacked against you, my fellow marketer.
The most egregious of the anti-science foes is the person who actively campaigns against data-driven marketing, using any of several excuses. You may have heard them yourself.


  1. Everybody else is doing it.

  3. It’s not “on brand.”

  5. I wouldn’t respond to that!

  7. It’s not creative enough.

  9. Give visitors the facts and they’ll figure the rest out.

  11. We sell to everyone!

Read the entire article at Search Engine Land, or listen to the article via the Conversion Scientist Podcast.

Listen | Download | Subscribe

How does engagement affect conversion rates? Could it impact them positively or negatively? Is it possible to increase conversions by decreasing engagement?

We’ve just finished some very interesting research here at Conversion Sciences labs, and we love it when our deeply held beliefs get blown out of the water.

It’s happened again.

Most of us assume that if our pages are “engaging” to visitors, that they are more likely to convert to leads or sales. If they are engaged, they have more time to take action. If they are engaged, they will truly understand our value and become a lead or a customer.

Look at the following graph of three videos. These three different videos appeared on three otherwise identical landing pages. The graph is “Viewer Attention” as is recorded by YouTube. Basically, these graphs tell us how many visitors were still watching at any point in the video. It tells us how engaging a video is.

YouTube’s Viewer Attention metric would predict that “talking head” video would deliver the lowest conversion rate. In fact, it is the highest converting style of video. In this case, engagement doesn’t predict conversion.

YouTube’s Viewer Attention metric would predict that “talking head” video would deliver the lowest conversion rate. In fact, it is the highest converting style of video. In this case, engagement doesn’t predict conversion.

Clearly, we would expect whiteboard style video to be the highest converting video, since viewers are more engaged for the entire length of the video. We expect slides to be almost as successful. However, we expect plain old talking head videos to perform poorly.

Now take a look at the following graph. This is a graph of the conversion rates of the same videos.

YouTube conversion rates graph. Can You Really Increase Conversions By Decreasing Engagement?

YouTube conversion rates graph.

This graph tells us that plain old talking head video is getting more visitors to click on our call-to-action button. This style of video is almost twice as likely to convert a visitor than the slide-style video found in most webinars.

How Does Engagement Affect Conversion Rates?

Clearly, engagement doesn’t predict conversion in this case. Here, engagement is actually distraction.

Learn more about the relationship between engagement, distraction and conversion in my article Can You Really Increase Conversions By Decreasing Engagement?

I think you’ll be surprised by what you will learn.
Brian Massey

There’s nothing like an arch nemesis to get the “good guys” motivated and engaged in a quest. If you are trying to get your conversion project approved, consider finding a competitor who seems to be making all the right moves.
This is the topic of my new column on Search Engine Land How To Get Management Excited About Conversion Optimization. In it I talk about the power of an enemy.
I discuss the components of your competitor story, including appearing to have greater resources than you, appearing to be smart, and appearing to have an evil plan.
I introduce you to some Spy Equipment that tells you just how smart they are.
I show you how to tell if they are they optimizing their site, using analytics, page analyzers or testing packages.
You can also tell if they are using ratings and reviews, recommendation engine or video players.
I tell you how you can catch them monologuing, revealing their secrets.
Finally, I show you how to cast them in their best light, making them more threatening.

image image

With a little research, you can clearly draw a picture of your competition as an evil genius, working to take food from the mouths of your family. You can develop the story of the powerful competitor using some spy equipment, their public “bragging,” and some page analysis tools. This will make your conversion project seem more critical to the survival of your business.


Read the full article.

This is a guest post by Katleen Richardson of Marketing AdvantEdge
If you’ve been in business for a while and have only recently started to develop your online presence, the idea of having to measure your online marketing performance can be daunting. Even for those who have more experience in the world of conducting business online, trying to compare your online and offline measurements can seem tricky.  It’s important to remember, though, that these two aspects of your marketing strategy are fundamentally the same.  For both, ROI is the most important factor, and in both cases you’re going to be most interested in these core metrics:


  • response rate

  • lead conversion rate

  • sales conversion rate

  • average deal size

  • gross revenue

  • expenses

There is one aspect exclusive to online measurement that gets a little complicated, however – social media performance.  Calculating your social media influence, one of the core social media metrics, can be done in a number of ways.  Here’s one that works well:
volume of content x number of comments x number of shares x net reach = influence
When assessing your net reach, take the deduped audience you have across all social media channels.  In other words, a person who follows you on both Facebook and Twitter is only counted as one person.
In order to get meaningful measurements across online and offline channels while, at the same time, managing your work load, start by choosing a core of five metrics that you’re going to track from month to month consistently.  This creates a solid base from which to make comparisons, and will help to keep you from getting overwhelmed.  Keep in mind that whatever you decide to track needs to tie back to both your marketing objectives and your overall business objectives.
Here are a couple of principles to help you tie it all together:


  1. Make sure each offline component can be tied back to an online component.  Let’s take, for example, QR codes used on brochures.  If you embed QR codes with URLs created with a URL shortening service, you can then track the performance of the shortened URL.  Look at the number of hits on the URL against the total distribution run of the brochures to measure the overall response rate.

  3. Let KPI and media mix reports help steer you.  These reports should be reviewed every month, both within marketing and across the company, to help you make adjustments in the course of your marketing strategy.  The KPI report should include whatever metrics tie back to marketing and company goals.  For the media mix report, take the metrics listed above and apply them to each of your marketing channels, including SEO, advertising, direct mail campaigns, landing pages, and so forth.

If you really want to get strong analytics in place that will cover all your bases, your best bet is most likely going to be a marketing automation system.  These days, there’s no reason not to – there’s a variety of choices out there that cover a full spectrum of requirements, even if you’re on a tight budget.  All you need to do is find the system that gives you the most appropriate options for your situation, and you’ll be well on your way.
Kathleen RichardsonKatleen Richardson (marketing-advantedge.com) is an experienced leader who builds integrated strategies combining research, data analysis and creative thinking. She has delivered successful solutions for the publishing, financial and telecommunications industries, as well as for conference and training companies, and professional associations. Her approach is to design customer focused, cost-effective solutions based on cross functional collaboration and results-based metrics.

iStock_000008395106MediumNow, if you’re interested at all in website conversion, I should know that you don’t wake up at night saying, “Wow! I need to accelerate my online business!”

You certainly don’t spend your day in a cold sweat, wondering how to “optimize for buyer personas.”

So, why did I name a webinar “Accelerating Your Online Business by Optimizing for Buyer Personas!?”

Because sometimes I make the same mistakes you are making in your marketing. I’m trying to come up with titles and headlines and copy that appeal to everyone, when I should be focusing on you, a real business with real conversion concerns.

So here’s my promise: I’m going to tell you who should be part of this webinar, straight up, and you can go about your business if it isn’t you.

If you want to know what color your Add to Cart button should be, you’re going to hate this webinar.

If you want to know how to get your team to stop writing titles like “Accelerating Your Online Business by Optimizing for Buyer Personas,” you’re in the right place.

If you’re looking for ten steps toward better landing pages, you’re going to be disappointed (but you will be entertained).

If you’re frustrated that your website isn’t doing better, and you’re not afraid of change, you should be here.

If you like presentations named “Accelerating Your Online Business by Optimizing for Buyer Personas”, you’re not going to like Accelerating Your Online Business by Optimizing for Buyer Personas.

We’re going to have more fun than that title suggests.

We’re going to talk about personas, which I know is very conceptual for many of you. You are dismissed.

For the rest of you, I offer a clarity of direction and purpose that few website teams can claim.

And we’re going to see if you learned anything, by finishing up with “What to Test: The Conversion Quiz.”

Yes, online marketing should be fun.

So, please ignore the title and come if you want to learn:

  1. An easy way to model your best visitors into four “Modes of Research”
  2. The limits of demographics and how to overcome them
  3. How to use analytics to uncover persona behavior
  4. The Conversion Quiz: Test your ability to apply personas to on-page tests

Pretend we named the webinar something more interesting.