EsheleD Marketing & Technology

25Dec/110

Case Study: Universal Technical Institute Increases Application Requests Eight-fold With Website Optimizer

UTI revved up their conversion rate with Google Website Optimizer. By conducting rigorous testing, making design improvements to their landing pages, and increasing AdWords investment to capitalize on their higher conversion rates, UTI cut their cost-per-application request in half and increased their request volume by more than 700% in in marketing campaigns that used the new landing pages. Read more to learn how they did it!

Since 1965, Universal Technical Institute has helped automotive enthusiasts turn their passions into careers. Offering programs in Automotive, Diesel, Collision, and Motorcycle Repair, as well as courses to become a Marine or NASCAR technician, UTI has trained more than 140,000 professionals in the automotive industry. Today, UTI operates 11 campuses, and currently has more than 15,000 students that are working with their hands to learn career skills that will last a lifetime.
The Goal: Improve Conversion Rates Through Website Optimization
UTI has been an AdWords advertiser since 2004. After years of fine-tuning their search campaigns, and successful investments in Display, Mobile, and YouTube, UTI was looking for their next big avenue of growth. After doing some research - they realized the opportunity was right in front of them, on their own site! With the help of Website Optimizer, UTI tested different calls-to-action, images, videos, buttons, and button placements. After many different landing page combinations, they ultimately settled on a winner.
The Results: A Great-looking Landing Page and a 300% Increase in Conversion Rate
UTI's new landing page was a dramatic departure from their previous landing page. Most notable is their use of “gradual engagement” - asking for user’s information in several steps rather than showing an intimidating 15-field sign-up form.
Original landing page (click to see full size)
New landing page (click to see full size)
“Prospective Students responded extremely well to our new landing pages,” said Loring Kohrt, Online Marketing Manager for UTI, “not only did we triple our conversion rate, we reduced our exit rate and saw a boost in site engagement metrics across the board.”
More Good News: The Effect on Advertising
The jump in conversion volume was substantial and immediate, but UTI didn’t stop there. Realizing that their increased conversion rates had dramatically lowered their cost-per-lead, UTI decided that they had room to bid much more aggressively on Google’s Search and Display Networks. By doing so, they increased their average position, and brought in eight times the conversions at about half the CPA!
Improving conversion rate amplified all other marketing activities
Reflecting on his experience, Mr. Kohrt said “We are tremendously happy with the results of our redesign. We were able to dramatically improve the profitability of our advertising, increase our enrollments, and deliver a better experience to potential students.”

Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

New Year, New Asynchronous Tags for GWO

Happy New Year fellow testers! We’ve just made a change to the Website Optimizer tags that will help your pages load faster, improve data accuracy, and eliminate tracking errors when tags are not fully loaded.
The next time you go to create an experiment, you’ll be given the new tags, which use asynchronous JavaScript. Asynchronous tagging, which is already a standard for Google Analytics, allows the Website Optimizer tracking and conversion scripts to run in parallel to your site loading. For Website Optimizer veterans, these new tags appear in Step 2 of the setup wizard -- be sure to read the instructions as the installation is slightly different.
Here’s five things you should know about the new tags:
  1. The traditional (non-async) snippets will continue to work if you want to use them. And there’s no need to retag experiments.
  2. We’ve combined the Control Script and Tracking Script into one snippet of code. Be sure not to double tag your pages.
  3. Async tags go immediately after the opening <head> tag on your test and conversion pages, not at the bottom of the page like the old tags.
  4. If you customize your Control Scripts (for experiments with cross-domain experiments and the like), the Control Script still uses urchin.js-style methods.
  5. Articles in the Help Center have been updated to reflect the new tags though we still have articles with the traditional tag for your reference.
Here’s to winning experiments in the new year.


Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

Welcome, Google Apps users!

Google Apps recently launched an improvement that made dozens of exciting Google services available to Google Apps users for the first time. As part of this launch, Google Website Optimizer is now available to our Google Apps users for free with their Apps accounts.
Google Apps is Google’s suite of cloud-based messaging and collaboration apps, including Gmail, calendar, documents, spreadsheets, and more, specifically optimized for use in organizations. These services, which run entirely in the cloud, are used by more than 30 million users in small and large businesses, educational institutions, government agencies, and non-profit organizations around the world. You can learn more about how Google Apps can lower IT costs and improve productivity and collaboration at your organization at google.com/apps.
For those users who have a Google Apps account, if your administrator has already transitioned your organization to the new infrastructure, you can get started using Google Website Optimizer at google.com/websiteoptimizer with your existing Apps account.
For more details, read the complete post on the Google Enterprise blog and follow all the updates on other newly available services for Google Apps users.

Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

Update your Website Optimizer scripts to secure your site

Earlier this week we notified affected Website Optimizer users of a potential security issue with the Website Optimizer Control Script. If a website or browser has already been compromised by a separate attack, a hacker might also be able to execute malicious code by exploiting a bug in the Website Optimizer Control Script.
We have not seen any evidence indicating that sites using Website Optimizer have been targeted through this bug, but wanted to proactively reach out to site owners. While the probability of this attack is very low, we are urging Website Optimizer users to take action by updating their Control Scripts. We have taken action, so all new experiments created after December 3 are not susceptible.
Any experiments you are currently running need to be updated to fix the issue on your site. Additionally, if you have any Website Optimizer scripts from paused or stopped experiments created before December 3, you should remove or update that code as well.
There are two ways to update your code:
  1. Stop current experiments, remove the old scripts, and create a new experiment.
  2. Update the code on your site directly. We strongly recommend creating a new experiment as it is the simpler method.
Instructions for both methods are available here at the Website Optimizer Help Center.
We’re committed to keeping Website Optimizer secure, and we will proactively work to prevent any future vulnerabilities.


Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

New Google Website Optimizer Online Video Training Released!

A new online video training course is now available from lynda.com to help you get the most out of using Google Website Optimizer. It’s called Google Website Optimizer: Essential Training, and it's created by David Booth of WebShare, one of our Website Optimizer Certified Partners and Seminars for Success Leaders.
The course itself is available both online and on DVD and contains almost 4 full hours of step by step instruction around how to identify test pages, plan and implement your experiments, and read and interpret the reports. Advanced topics like sample size calculation, Google Analytics integrations and dynamic page testing are also covered among a wide variety of conversion optimization and testing topics.
I sat down with David and asked him to share with us what he likes best about this new course, and here’s what he had to say:
A video course is great for a few reasons: First, it lets you accomplish things you can’t necessarily do in other formats. As an example, in this course we have start to finish screen captured demonstrations of how to create a new test, tag your pages with Javascript, and then launch both A/B and Multivariate tests. Someone watching this course can actually see it happening and not have to rely on instructions or a help file.
I also like the variety of learning styles that can be accommodated by what the lynda.com team has put together. If you’re a visual learner, you’ll love this - there’s a nice mix of presentation styles, screen captures and visual aides throughout the whole course. If you learn by doing, you’ll enjoy the step by step, narrated demonstrations you can follow along with.
And it’s all broken up into videos that are just a few minutes long. If you want to block off a morning and watch it straight through you can, but if you want to catch a couple of videos when you’ve got 15 minutes of down time you can get through it that way too. Or if you just have a question on a specific topic, such as when to use A/B vs Multivariate, you can jump directly to that video and have your answer a few minutes later.
Take a look at the trailer below:
Well, we like things that make it even easier to use Google Website Optimizer in even more effective ways, so thanks Dave. You can find the course at lynda.com!


Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

Join Website Optimizer at eMetrics & Conversion Conference in Washington DC

In less than two weeks, Conversion Conference East will take place in Washington DC. Google Website Optimizer will be there along with Google Analytics.
Conversion rate optimization is still a young discipline in the world of interactive marketing. The event was founded by Tim Ash, president of Site Tuners (a Website Optimizer Certified Partner). What's exciting about Conversion Conference is that the entire program is dedicated to the discipline. Experts in the field, as well as those who are just learning the art and science of conversion optimization, now have a forum to share best practices, network, and learn from each other.
If you are anywhere near Washington DC October 4-5, we hope you'll join us there. We have a coupon for $250 off for Website Optimizer blog readers. Enter the code CCE631 when registering at the Conversion Conference site.
Here’s a taste of some of the presentations at Conversion Conference:

Monday, October 4th
The Four Pillars of Building Instant Trust Online (Tim Ash)
The Power of Split Testing (Brooks Bell and Lance Loveday)
Pay Per Click Landing Page Continuity (Lauren Vaccarello and William Leake)

Tuesday, October 5th
Getting Started with Google Website Optimizer (I'll be giving this one)
Multivariate Testing (Eric Hansen and Chris Duskin)
The Science of Pursuasion (John Whalen)
Be sure to stop by the Google booth and say "Hi" and grab a squishy HiPPO. Looking forward to seeing you.

Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

Creating Targeted Tests using DoubleClick Ad Planner and Website Optimizer

This guest post was contributed by Daniel Waisberg, the Founder and Editor of Online Behavior, a Marketing Measurement & Optimization website. Daniel looks at how you can use DoubleClick Ad Planner to find ideas for testing.

Testing is probably the most effective way to optimize websites. Through testing we can understand what our customers like, which ultimately will help us create a better customer experience for our audience. But "our audience" is usually not a unique type of person; it is important use techniques such as Test Segmentation to understand the differences in the tastes of each cluster of customers.

However, where can you get ideas for tests? How do you choose, for example, if you should use an image of a man, a woman, a couple, a baby or a family? Most of us do not have the privilege of testing the YouTube homepage: traffic is limited for most sites, so it is important to run tests that have a high chance of making a difference. We have to focus our efforts on our best guesses. In this post, we will show a way to use DoubleClick Ad Planner to research for testing ideas that will be tailor made to the segments you are trying to target in your website.

Finding Your Audience on Ad Planner

In a recent blog post on the DoubleClick Advertiser Blog, the DoubleClick Ad Planner mission is described as:

...to provide the deepest, most accurate insight into online audiences possible. This insight helps display advertisers select the best sites for their media plans and drive results for their campaigns.

However, I believe this description is missing an important part, which is not less important to advertisers: to understand your audience tastes and which kind of websites they like. The DoubleClick Ad Planner provides important insights into how to design your campaign landing pages and your website at all.

So, let's suppose I am working to optimize the eMetrics Summit website for the San Francisco conference in 2011. The Summit targets marketing managers, web analysts and business intelligence experts that are trying to understand how to increase the return on online investments. Here is how to find the tastes and preferences of this audience:

  1. Sign in to DoubleClick Ad Planner and create a new Media plan;
  2. Go to Research tab, choose the Research by Audience secondary tab;
  3. Choose among the various segmentation options in order to narrow the audience and the websites they visit. Below are the segments chosen for eMetrics San Francisco audience:
  • Geography: chose country USA and refined it to include only West Coast states. That's the main target for this show since eMetrics also hosts a Washington DC conference
  • Demographics: included both males and females, between 25 and 44 years old, with at least a bachelor degree, with a household income above $75K. I think this segment is very close to the audience of the conference (but I have no inside information)
  • Online Activity: chose a large website that the audience is likely to visit: Google Analytics
  • Interests: chose everything under 'Business' and 'Computers & Electronics'
  • Ranking Method: chose the ranking method to be 'Best Match' since we are not doing this analysis in order to find a place to advertise (in which case we might sort the websites by reach), but to find a place that our target likes to visit

Below is a screenshot from DoubleClick Ad Planner showing all the segmenting options and the audience created above. We can call the list created below "Website Testing Inspiration"

Click for full-size image


Getting Ideas for Your Test

Once we find the "Website Testing Inspiration" table, which shows the websites where our targeted audience is surfing around, we have the raw material necessary to get ideas for our testing efforts. Continuing our example above, we can visit the websites in the Top 10 websites that match our audience and start analyzing them.

So, here are a few insights from the analysis above for the eMetrics San Francisco home:

  • First of all, looks like Jim Sterne chose the right color, blue is very prominent in all the websites;
  • Idea #1: it could be worth a try to add some geeky machines to the page, such as in the Pitney Bowes, Kaiser Permanente and Frys websites;
  • Idea #2: call these companies and have someone present at eMetrics and feature it at the conference homepage;
  • Idea #3: submit a post to both TechCrunch and Gizmodo, which would certainly be happy to feature interesting content about social media metrics. The posts would be useful in order to promote the conference and, in terms of testing, the eMetrics homepage could try featuring in a prominent place that the conference is being quoted in these websites (something like "In the news");
  • Idea #4: interesting to see that Stack Overflow is number 5 on the list, a website for "professional and enthusiast programmers". It looks like many technical people are inside this audience. Maybe it could be worthwhile to try showing a classification on the site targeting different types of people: "Programmers only talks", "Business Minded talks", "Marketers, Statisticians and liars"...
These are initial ideas that should be discussed and improved based on the website and the target being studied. As the analysis gets deeper, the insights will become more valuable.

Bonus: Instead of looking for your audience and which sites they visit, you can also look into your competitors' sites and understand which segments they are attracting that you are not. Read more about it on Avinash's post: Competitive Intelligence Analysis: Google / DoubleClick Ad Planner.

25Dec/110

Website Optimizer at eMetrics San Jose and Conversion Conference West next week

Two great industry conferences are happening next week in San Jose: eMetrics and the Conversion Conference. Veterans of the industry will be familiar with eMetrics, but you may not know about the Conversion Conference.
The Conversion Conference is a brand new event being chaired by Tim Ash. Tim is the president of Site Tuners (a Website Optimizer Certified Partner), and if you’ve been to any online marketing conference in the last few years, you’ve probably heard him speak about the importance of conversion. He’s helped assemble some of the best minds in conversion (like Bryan Eisenberg and Jakob Nielsen) for a really amazing event.
If you haven’t registered yet, Conversion Conference is offering a $250 late-bird discount. Just enter CCW646 when registering at the Conversion Conference site.
Here’s some of the happenings at the two conferences:
Tuesday, May 4th
  • 11am - Introduction to Website Optimizer session (I’m giving this session)
  • 1:50pm - What’s New with Google Analytics with Brett Crosby
  • 2pm - Web Analytics & Uncovering Problems with Eric Peterson & Brett Crosby
We’ll be hosting a book signing of the brand new book Performance Marketing with Google Analytics. All three authors, two Google Analytics Certified Partners, Justin Cutroni and Caleb Whitmore and a former member of the Google Analytics team, Sebastian Tonkin, will be on hand giving away and signing the new book on May 4th, starting at 3pm in the Cupertino room.
Wednesday, May 5th
  • 9am - Keynote: Leveraging Bleeding Edge Analytics Goodies with Avinash Kaushik
  • 12:15pm - Brown Bag Deep Dive with Google Analytics product manager, Jayanth Mysore
Also, don’t miss your chance to come face to face with the HiPPO at the Google booth. Hope to see you next week in San Jose.

Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

Website Optimizer system upgrade coming soon

​Within the next two weeks, Google Website Optimizer will undergo a system upgrade. During this upgrade, Website Optimizer will be in maintenance mode. Your experiments will continue to run and collect data, and you’ll be able to view your reports. However, you will not be able to create or modify experiments. This includes launching, pausing, or stopping an experiment.
We’ll notify you of the exact timing of this upgrade both here on the blog and with a message inside your Website Optimizer account. Thanks for bearing with us as we continue to make Website Optimizer better.
If you also use Google Analytics you should know that they’ll be undergoing a similar upgrade, which you can read about on the Google Analytics blog.
Update: 3/2/2010 at 8:55 AM
We've begun the upgrade and Website Optimizer is currently in maintenance mode. A reminder that you'll be unable to create or modify experiments during this time.
Update: 3/3/2010 at 8:10 AM
As of about 2AM PST this morning, Website Optimizer is out of maintenance mode and working normally. Thanks for bearing with us during the maintenance period.

Posted by Trevor Claiborne, Website Optimizer team

25Dec/110

​25 Google Website Optimizer Tips for Better Product Pages

Late last year, a fantastic post went around the Internet with 55 Google Website Optimzer Tips and Tricks. Today, we have a sequel for you. This is a guest post from Khalid Saleh, who is the president and cofounder of Invesp, an ecommerce conversion rate optimization company.

Increasing e-commerce conversion rates is challenging. With thousands of possible pages, where should you start optimizing? Even if you can answer that question by picking a small set of pages to start with, what changes should you make?
The following tips will be focused on product page testing, but we will first set the stage with few setup items:

1. Segmenting your visitors. Nothing is more powerful than segmenting your visitors and examining the different trends these visitors follow on your site. At a minimum, you should segment visitors by source of traffic, keyword, type of visit (new vs. returning) and country.
2. Segmentation with Google Website Optimizer = Testing on Steroids! Not all customers are created equal, nor should you deal with all customers equally. Testing different versions of your page or funnel is powerful. Testing different versions based on the type of visitors is even more powerful. There are different ways of doing this, but we recommend starting with BTBuckets or with these instructions.
3. Setup your funnels. Funnels tell you what pages are causing visitors to exit the site. At a minimum, we recommend you setup the following funnels:
  • Home page to cart page funnel
  • General checkout process funnel including the cart page
  • General checkout process funnel excluding the cart page
  • Checkout process funnel for different payment types including the cart page
  • Checkout process funnel for different payment types excluding the cart page

4. Funnels & visitors. Beyond setting up funnels is segmenting your traffic and examining how it moves in funnels. But how do you do that in Google Analytics? We will let you in on a small secret:

  • Setup the different funnels in Google Analytics
  • Setup different profiles to represent the different segments
  • Now, you can examine each of the funnels for the different profiles
5. Returning visitors vs. new visitors. Comparing different segments of visitors can be confusing. You should expect visitors from organic search to interact differently with your site compared to visitors from paid search. We find that segmenting new visitors versus returning visitors can reveal major design problems in the site. Generally returning visitors will take longer on a site, view more pages, and have lower exit and bounce rates. If that is not the case, then you might have major design issues to deal with.
6. Product pages: A micro conversion. We like to run the product pages tests twice with different conversion goals:
  • Micro conversion goal: When a visitor adds an item to the cart, we count that as a micro conversion.
  • Macro conversion: When the visitor completes an order, we count a macro conversion.
7. Product pages: the image. Yes, you heard it before. An image is worth a thousand words. However, product images do require an investment, so test the correct solution for your website. Each of the options below will require considerable time, software, and resource commitment. So, test with 5 to 10 products before choosing the right solution for your website.
8. Product pages: Image of product in use. Test with showing the product in use vs. images with solid backgrounds. While in many cases showing the product in use helps the visitors see themselves use the particular product, this is not the case for all products. Apparel websites seem to benefit from this feature. Electronics or book websites, not so much!
9. Product pages: Location of the image. Where should you place the product image? Common wisdom is to place product images on right side of product pages. But common wisdom fails to predict how visitors will act on your website. So test placing the image on left versus the right and see which converts better.
10. Product pages: Product zooming. If you have the software installed on your website, then allowing visitors to zoom in and examine the product closely is an excellent feature to add.
11. Product pages: Multiple product images. Finally, consider having multiple images of the product. We usually recommend having 3-4 images per product to start with. But a word of caution here: hire a professional photographer for these images. Having several low quality images versus a single high quality image is something worth thinking about.
12. Product pages: your product descriptions. Most e-commerce stores take the product description from the manufacturer and place it on the site. So, you will find the same description on competing e-commerce stores. Pick your top 10 best selling products and:
  • Test the impact of having the standard manufacturer's description versus custom copy your team creates. What impact does this have on the micro conversion (visitors adding items to their cart)?
  • Test different versions of the custom copy you create: Not all copy created equal. You might want to test technical copy versus non-technical. Consider the different market segments and what words will have the most impact on them.
13. Product pages: Reviews. Reviews are great for increasing conversions. There are many studies that show product reviews are responsible for up to 30% uplift in conversions. However, if you are just starting out, allow customers to add reviews but do not display them right away. Wait for a product to collect enough reviews before sharing them with your visitors. Consider using services such ProductWiki to add images on your site.
14. Product pages: add to cart buttons. Yes, they do impact conversion. And there are many things you can test with these:
  • Test the location of the add to cart button
  • Test different designs of the add to cart button
  • Test the wording on the add to cart buttons
15. Product pages: don’t overdo it! What do you think of this?
Too many CTAs (call to actions) can stop a visitor in their tracks.
16. Product pages: cross-sells and upsells. We know that you want to sell more of your products. Cross-sells and upsells can help you do that, but they also can distract the visitors. So, test the placement of cross-sells and upsells on your product pages. Should they go in the left navigation or somewhere in the middle section of the product pages?
17. Product pages: pricing. Most of your visitors will do some sort of comparison shopping. So, test different prices and see what impact that will have on your conversion rates. We had a client who reported a 135% increase in conversion rate by lowering product prices by 10%. We do not like to focus on pricing as a mechanism to increase conversion rates but sometimes it works really well.
18. Poor man testing: pricing & Google Website Optimizer. It is usually difficult to switch back and forth between different pricing for items, but here is a general framework to help you:
  • Most ecommerce packages provide a way to maintain a single price for a product. Plan on a strategy for testing multiple prices:
  • Price A is the original price
  • Price B is the original price x 1.10 (10% increase)
  • Price C is the original price x 0.90 (10% decrease)
  • Price A (original is already saved in your database). You can either maintain Price B, C in the database or have application logic to switch between the different prices for a single item.
  • One option to implement this is to have all prices available on the page and use GWO to display one price and hide the other two prices. Another option will be to use GWO to pass a variable value to be used in a formula to change the price.
  • Implementing this should take less than a day and will provide a mechanism to test prices in Google Website Optimizer.
  • 19. Product pages: stock availability. This is a no brainer. No customer likes to add an item to their cart only to discover later that the item is out of stock. So, tie your front end e-commerce store to your inventory system. If that is too much, then consider a daily update.
    • Bad: No product availability on your site
    • Good: Always display product availability
    • Excellent: Display messages to encourage customers to buy a product if you have limited stock available (Only 5 left in stock -- order soon.)
    20. When can I have it? Tell visitors when the product will be in their hands. Do not tell visitors the items will arrive 3 to 10 days. That will simply kill your conversion and if it does not, it will irritate visitors.
    21. Product pages: Feature list.
    The first level of attracting customers is features. Features are the facts about a given product. Features of an outdoor grill might include:
    • Electric
    • Grease tray
    • Non-stick cooking surface
    • Temperature control
    • 300 square inches of cooking space
    • Attached working surface
    22. Product pages: The benefit list. Benefits start you on the road to closing a sale. The benefits of the above outdoor grill include:
    • Convenience
    • Electric
    • Non-stick cooking surface
    • Attached working surface
  • Easy to Use:
    • Electric
    • Non-stick cooking surface
    • Temperature control
    • Attached working surface
  • Saves Money
    • Electric
    • Temperature control
    23. Product pages - People buy benefits, they do not buy features. What results do your customers want from your barbeque? They want a grill that offers:
    • Tasty, Healthy Food Every time
    • Temperature control
    • Grease tray
  • Perfect for Easy Entertaining
    • 300 square inches of cooking space
    • Attached working area
    • Easy to clean
  • A Double Bonus: Environmentally Friendly While Saving You Money!
    • Electric versus charcoal emissions
    • Do not have to buy charcoal for the life of the grill
    24. Bundled pricing. Consider bundling shipping costs with product items so visitors will not have to worry about paying for shipping costs. If the shipping costs are too complicated to understand or might be on the high side, we recommend testing bundling them with the product price.
    25. Free shipping is a great way to increase sales. We have many clients who offer free shipping on orders above a certain size and usually see a nice uplift in both conversions and average order value by doing so.
    26. Bonus tips: If you are about to start conversion rate optimization, then these resources are a must:

    Conversion rate 101: a series of articles on conversion rate optimization.
    Personas 101: a complete guide to creating personas for your website
    30+ principles to a better landing page design: a must read for anyone interested in landing page design.
    Landing page templates: we optimized these templates using GWO to increase their conversion rates to the double digits. They are a good starting point.
    Thanks to Khalid for putting together this great list. Share your own favorite GWO tips in the comments.


    Posted by Trevor Claiborne, Website Optimizer team

    Page 1 of 3123