Stop by our booth if you plan to attend these events, and if you’d like to schedule a meeting with an Optimost representative in advance, please

The holiday season has officially begun. How do I know? Well, a couple days ago I was talking on the phone to a friend driving in his car who was shocked to report to me that Bruce Springsteen's rendition of "Merry Christmas Baby" had just come on the radio. Pretty remarkable since he, like me, had just gone trick or treating with his kids the night before.

Yes, it was November 1st and already some radio stations were switching over to the increasingly popular "All Holiday" music format. It used to be that around a week before Christmas, certain stations would switch to All Holiday music. Then it became the day after Thanksgiving, and then mid-November. And from my limited Internet research, it appears that radio station WNIC in Detroit last year may have set a new standard, by being the first station to go All Holiday on November 1st.

Now it appears that others are following suit. Thank goodness for Halloween! If it weren't for Monster Mash, Werewolves of London, and other Halloween auditory staples, we'd almost certainly be hearing Rudolph the Red Nosed Reindeer in September.

So why do stations do this? To paraphrase the great poet Ray Davies, it's just giving the people what they want. And apparently, there are plenty of people who have the Christmas spirit even while the jack o'lanterns are still on the front porch. WNIC did a poll this year to ask when their listeners wanted to start hearing holiday music, and 59% voted for the beginning of November or earlier. Maybe Santa Claus will soon be coming to town in October after all.

There's a good lesson here. It's important not to make assumptions based on some preconceived notions of when people want certain things. When you actually ask people, or watch them behave, you may be surprised at what you find. We see this often in the multivariable tests we run for our clients. For example, prior to working with us, one web site had habitually switched off their pictures with winter imagery (kids running through the snow, etc.) on the first day of spring. However, when they actually tested winter images versus spring images, they found that the winter ones actually continued to perform best through the end of March and well into April.

So if you haven't already, it's a good time to start testing your website to figure out when your visitors are starting to think about the holidays. And it's not just when to switch over to holiday imagery that matters. It's also how to change your website to capitalize on the different behavior your visitors may exhibit during the holiday shopping season. Holidays mean more gift-giving, so emphasizing return policies can be beneficial. Also, since people will be rushing to do a lot of shopping for a lot of people, downplaying the bells and whistles can also be a good strategy. Don't force people to sit through that expensive video demo your marketing department is so proud of when many stressed-out people just want to buy their presents and be done with it.

The good news is that if you haven't already started testing for the holiday rush, it's really not too late. In our Customer Spotlight this month, we highlight the work we did during last year's holiday season for Dale & Thomas Popcorn, the nation's leading gourmet popcorn maker. The bulk of our testing took place in November, and we still were able to improve conversion rates versus the control by 13.4% during the balance of the year.

So make sure you start testing now to figure out what is the optimal way to deck the virtual halls. The holiday season brings the perfect storm online - higher traffic and higher conversion rates. Make sure you are not left out in the cold.

What is confidence and what determines confidence?
Confidence is the degree of certainty for an observed measurement. It can be best explained in the context of a comparison. Let’s take the example of two versions of a homepage being tested (Version A and Version B), and the measurement of success is the conversion rate for each page. As the testing process begins to get enough visitors to each page, let’s assume the observed rate of conversion for Version B is higher than that that for Version A. The marketing experimenter would like to be confident that Version B is indeed converting higher than Version A and will perform a statistical test to determine this confidence level, or degree of certainty. If this analysis tells us that the degree of certainty is 98%, then the experimenter is 98% certain that the observed conversion rate of Version B is in fact greater than that of Version A. However, there is still a 2% probability that Version B is performing equal to or worse than Version A.

Typically, 95% confidence is used as the metric for evaluating the significance between two measurements such as conversion rates. Achieving this level of confidence is based on two things: the margin of difference between the conversion rates and the number of visitors who see each version (the sample size). A large margin of difference between the conversion rates of the two page versions will require a lower number of visitors to each page and hence a shorter testing period. However, when the margin of difference between the two conversion rates is small, the sample size required to achieve 95% confidence will be much greater, increasing the testing period.

What is an interaction?
An interaction is the interdependency that exists between two or more variables. When testing web pages through a multivariable approach, the influence of each variable on the success metric, such as conversion rates, can be quantified. Multivariable testing also allows the experimenter to detect whether or not multiple variables interact with one another. For example, assume there are two areas tested on a page, A and B, which represent the header copy and introduction copy respectively. Both areas contain a copy version using the word "Free" within the text. The multivariable approach will determine how much influence the header and introduction copies have on conversion rates AND will also investigate any interdependency or interaction between the two sections of the page and what effect that interaction has on conversion.

Figure 1. % Conversion Dependent on Variable A	Figure 2. % Conversion Dependent on Variable B

From Figures 1 and 2, you can see that the individual effects from including the "Free" language in the header improves the conversion rate but the mention of "Free" has no real influence if the language is included in the intro copy.

Since the test was conducted in a multivariable fashion, interactions between A and B can be detected as shown from Figure 3. Identifying this interaction will determine whether or not reinforcement with the "Free" text improves conversions – in this case, having that language in both places is the suboptimal setting. To maximize the conversion rate, the optimal setting will have the header copy with the "Free" text and the intro copy would be without the "Free" text.

Figure 3. Interaction Plot with % Conversion Dependent on Variables A and B

If you would like to submit a question to Dr. Montero, please send it via e-mail to .

Founded in 2002, Dale & Thomas Popcorn is the leading gourmet popcorn marker. The company’s unrivaled selection of unique flavors and assortments has resulted in world-wide recognition and features in Today, CNN, US Magazine, and ESPN. Oprah Magazine selected Dale & Thomas Popcorn as one of her "Favorite Things."

As last year’s critical holiday season approached, Dale & Thomas Popcorn needed to capitalize on that great press coverage and the accompanying increases in online traffic by converting a significant percentage of these new web visitors into real customers. Dale & Thomas worked with Optimost to optimize their home page and maximize their online revenue.

A dedicated Optimost project team developed a test plan that identified 7 home page variables and 41 different values (versions of the variables). Based on this plan, the Optimost solution engine quickly identified and generated 1,866,240 possible permutations of the home page. Using its advanced methodology, Optimost was able to conduct multivariable tests on a subset of these possible permutations of the home page. After only 30 days of testing, Dale & Thomas increased their online sales by over 13%.

In addition to dramatically increasing their online sales, multivariable testing enabled the company to understand how each variable impacted behavior and ultimately online sales. Some of the key learnings from the test included the fact that changing images, as well as changing the layout and presentation of those images, made a significant improvement on conversion. They also learned that, counter-intuitively, changing the order button from green to red improved results! They were also able to benefit from a host of other changes which improved their sales, leading to a very happy holiday season!

For more details on this success story and other customer success stories, please see http://www.optimost.com/case_studies.phtml.

Every company that is serious about their on-line presence has invested in, or will be shortly investing in, a robust web analytics package. A web analytics solution gives you the foundation to understand what is happening on your web site and how your visitors are behaving. It is the first stop for finding trouble spots on your web site. This is why Optimost partners with most of the leading web analytics vendors – the analytics data tells you where your problems are, and web site optimization solves these problems. One of Optimost’s web analytics partners is WebSideStory.

WebSideStory's HBX Analytics is delivered as an on-demand hosted service that is easy to implement and enables customers to avoid expensive, up-front hardware costs and software license fees. The cornerstone of WebSideStory’s platform is it’s Active Marketing Suite – this is a set of solutions which enable marketers to collect powerful data about web site visitor behavior, online advertising, email campaigns, and most importantly, how each new dollar of marketing spend is impacting the top line. In addition, WebSideStory provides a full suite of professional services to assist customers with everything from training, consulting, and best practices.

For more information about WebSideStory’s solutions, please go to www.websidestory.com or contact Jon-Paul Shoff at .

Click here for a complete listing of Optimost Partners.

If you are an Optimost partner and you are interested in either having your company featured in our Partner spotlight or your upcoming events listed in our newsletter, please send an e-mail to .

Optimost recently partnered with Google in the launch of their new Website Optimizer feature. Optimost is the first and only multivariable testing platform to develop a partnership with Google. As part of the relationship, Optimost will be offering professional services to client’s using the Google self-service tool.

Google Tests New AdWords Feature
MediaPost, October 19, 2006

If you’re interested in learning more about how Optimost can help you implement the Google Website Optimizer solution, please contact us.