Patricia Hader

I recently completed Market Motive’s “Conversion Optimization Master Certification” program led by Bryan Eisenberg (@thegrok).  As part of the program, I volunteered to perform a comprehensive conversion optimization analysis and provide testing recommendations (100+ A/B test recommendations) for The New York Public Library.  

Furthermore, I also had to run an A/B test.  Together with Johannes Neuer (@johannesneuer) from The New York Public Library we selected the newsletter sign up page for the experiment.

We actually ran two tests, but here’s the overview of the second test:

What we tested: Simplification of the newsletter sign-up page

• Remove newsletter options
• Shorten Privacy messaging
• Show thumbnail with link of newsletter sample
• Keep first test changes:

– Remove unnecessary information fields (first/last name)
– Stronger, benefit oriented headline
– Add unique value proposition points
– Clicking on “privacy policy” should open up new window

Control

Variation

 Did you guess it right?

…because “continuous improvement comes from continuous testing and optimization, not one-time research” (from Greg Burningham, President of MarketingExperiments). 

What other reasons do you need?  How else do you think you’ll improve if not through trial and error?  Bottom line, you won’t know what works and what doesn’t, unless you test.

Some buzz these days on real-time bidding in display advertising. 

New York Times: Instant Ads Set the Pace on the Web by Stephanie Clifford
ClickZ: Real-Time Bidding: What It Is and Why It Matters by Mike Baker
ClickZ: A 10-Step Guide to Evaluating Demand-Side Platforms by Mike Baker

Bottom line: 
With real-time bidding and optimization you can get the right ad, in front of the right person, at the right time, place, and price. 

Article highlights:

“While companies have been plugging real-time bidding for a couple of years, industry heavyweights are now behind it…Advertisers say real-time bidding cuts down on wasted money. You can use less media, because you’re using more selected, or more efficient, media.”

 “Until the arrival of real-time bidding, said Mr. Mohan of Google, “the technology hasn’t really been there to deliver on the promise of precise optimization, delivering the right message to the right audience at the right time” in the display world.”

 “…match users to targeted ads based on the attributes of each ad impression.”

 “Cost efficiency… Your ad spend can go further, with less budget spent on poorly-targeted impressions.”

“…an advertiser has to receive an ad impression, analyze it, decide what to bid on it, and decide automatically what ad to show in less than a quarter of a second to avoid slowing down the page-load time. It also lets companies funnel what they know about a Web user into the ads they show that person.”

 “Actionable insights… By learning what works for your campaign at the impression level, you have a great opportunity to identify trends and discover new insights about your best — and worst — consumers, context, and creative.”

This is heaven for marketers. 

You take your converter insights from your site and build segments based on the predictive attributes of the converter.  Your media optimization agency takes the segments, finds and targets them out on the web.  Their technology will make the decision on what ad to serve, when to serve and to whom to serve, and at what price based on the target’s likelihood to convert.  You can pay $10 for segment A, which converts at 4% and $3 for segment B, which converts at 0.5%.  Or, you can reject the impression, because there is no likelihood to convert.  You get the point.  Ultimately, you drive conversions with less impressions, because you’re only showing your ad to qualified audiences.  To sum it up — cost efficient and targeted.

Anyone doing this?  Any success?

It’s stunning that only 25% of interactive marketers use behavioral targeting.  Consumers are not one-size-fits-all and expect an experience relevant to their needs. 

For instance, this evening I searched for a flight to Munich, Germany.  After reviewing the flights results page, I decided to click on “Vacation Packages” in the top navigation and got targeted with an ad about Carribean resorts.  I’m not interested in Carribean resorts.  The ad is not relevant to my needs – which was a trip to Germany.   On top of that, I’m targeted with an ad for New York vacations.  Oh my…  My experience on the site could have been more customized through behavioral targeting.   They could have made some money on me with a tour package to Munich or a vacation package to the Alps.  Lost opportunity  (speak sales).   I’m going elsewhere…

Deliver the right offer to the right consumer at the right time!

It so happens that I received Coremetrics’ “A Marketing Imperative: Profile, Personalize, and Profit” whitepaper in my inbox today.  It highlights why delivering a personalized experience can increase sales and optimize marketing.  Here are the highlights:

Despite widespread recognition of the power of personalization, only a handful of companies do it truly effectively.

According to Forrester Research, only about 25 percent of interactive marketers use behavioral targeting (a term synonymous with personalization).  That means that 75 percent of interactive marketers don’t use personalization, or use it in limited and ad hoc instances.

If budget-conscious consumers and companies grow more discriminating, it becomes more important than ever for retailers and B2B marketers to zero in on customers and prospects with content that is relevant and customized to a customer’s needs.

And studies 8 by Aberdeen of companies that have deployed personalization solutions revealed that:

• 88% of best-in-class companies say they’ll realize return on investment (ROI) from personalization
• 76% agree economic gains from personalization outweigh investment costs

Personalization should not be viewed as a one-time or ad hoc effort. It is best treated as a core business strategy and should be supported by:

• Well-defined and detailed performance goals
• A structured performance monitoring system
• A dynamic roadmap for continuous optimization
• Automation across all stages of the personalization lifecycle

“In today’s homogenized world, customers want, expect, and desire a more customized and personalized experience,” the CMO Council said…

Forrester says that “almost twice as many marketers say they’ll be using [behavioral targeting] in the next year, which would make its adoption rival that of other established forms of online targeting, such as geographic, demographic, and contextual.”

Personalization means delivering product recommendations, offers, and content based on what you know about a given customer. Done right, it means developing dynamic profiles of individual customers and demographic segments by integrating data from disparate sources.

Understanding customer behavior in real time and matching marketing efforts to what you know about a customer at a given moment gives you the best chance of delivering content and recommendations that match your customer’s interests.

Test, Monitor, Analyze, and Adapt

As dozens of studies have shown, smartly executed personalization systems almost invariably generate highly attractive ROI, enhance online conversion rates, boost average order value, and strengthen customer loyalty and retention.

Anyone still not sold on personalizing the user experience?

You’re tracking engagements on your website, but are your analytics giving you the insights you need?

Avinash Kaushik has some interesting thoughts on measuring Engagement in his latest book, Web Analytics 2.0.  Here are the highlights:

Do ‘time on site’ and ‘repeat visits’ really measure engagement?
Fact is, measuring engagement with only quantitative data (web analytics) is a challenge.  Your web analytics tell you the “what”, but not the “why”. 

For instance, your site visitor spent 15 minutes on your site, but your analytics won’t tell you whether or not your visitor spent happy or frustrating minutes with your site.   Also, your analytics won’t tell you whether or not your visitor ran off to fix a sandwich for 10 minutes, meaning, he really just looked at your site for 5 minutes.  Further, your visitor came back 2 mores times (which your tool will count as a returning visitor) , but was he returning for updates or trying to find your contact information for the third time?  Unless he closed a sale (which assumes a happy experience), you really don’t know if your customer had a positive or negative engagement with your site.  You get the point…

Avinash notes that web analytics is “limited in that it can measure the degree of engagement, but not the kind of engagement.”

Based on this definition, recency and frequency of site visits really reveal just the degree of engagement.  So, five visits in a one-week period (the degree of engagement) doesn’t tell you if the visits were negative or positive, which Avinash refers to as the “kind of engagement”. 

In summary -

Degree of Engagement
- Web analytics measure the degree of engagement.
- It’s important to measure the degree of engagement along with outcomes.  Outcomes will tell you how successful the engagement was.  For instance, did the customer end up buying from you (outcome), as a result of coming back to your site a few times?

Kind of Engagement
- Surveys and  primary research measure the kind of engagement someone had on your site.
- Most likely someone who is making repeat purchases from you is happy with your products and service.  They keep coming back for more.

In order to measure engagement on your site, you should look at both - degree and kind of engagement – web analytics and qualitative data.

What are your thoughts on this and how are you measuring online engagement?

It can be quite challenging to get your management’s buy-in on measurement strategy, analytics resources, and the need for optimization.  Though it’s almost 2010, we still have Avinash Kaushik evangelizing web analytics and Bryan Eisenberg advocating for the need to test and optimize (and rightfully so!), not to mention the many analytics and optimization sessions offered at Marketing Conferences. 

Fact is, a lot of companies need desperate help, don’t get it yet, or don’t realize the missed opportunity.   And here you are, trying to sell them on everything you believe in…  So, how do you get your management on board?  There are obviously several ways to get them on board (e.g. “Stragegies for Embarrassing your Organization”, Avinash Kaushik, Web Analytics 2.0, p. 429), but another way to get their attention quickly may be to give them a reality check on where your organization stands. 

Use the [x+1] Digital Marketing Maturity Model to assess the maturity level of your organization.   The [x+1] Digital Marketing Maturity Model defines the capabilities required to convert prospects to customers through an integrated marketing experience.

 Pillars of Digital Performance-Based Marketing consist of
• Target Market
• Offer
• Creative
• Analytics
• Marketing Integration
• People / Skills

Are you a broadcast marketer or a value validator?   Do you know any organizations that are channel champions?  I can tell you that when I worked at AOL back in 2003, we were between value validators and channel champions.  Pretty impressive, given that it was 2003!  But, aside from smart marketers, we also had management support.

 
Another source to help you assess your digital marketing maturity would be Webtrends’ Digital Marketing Maturity Model (DM3).

The DM3 (beta) provides a framework for assessing and building digital marketing maturity over time in six core areas:
• Measurement strategy
• Analytics resources and domain expertise
• Data integration and visualization
• Data analysis and insight
• Adoption and governance
• Ongoing optimization

So, how mature is your digital marketing?  And, what are you going to do to take your organization to the next level?

“Testing is the biggest no-brainer, and the killer of most stupid ideas… Testing is great because you can get the most important person’s opinion: The Customer’s.”  –Avinash Kaushik, Occam’s Razor

Site Optimization: No More Guesses – ClickZ.

Oldie (January 2009), but goodie.   Obviously… I’m a big advocate for testing and optimization.   I don’t get why companies, in this day and age are still basing decisions and changes on someone’s gut-feel, opinion or guess (especially when you have free options available like Google Website Optimizer).   That someone may guess wrong and lead your performance to decline instead of improve.  So, why are companies gambling with potential success and missing out on improvement opportunities?   And, if you’re not optimizing, what is your excuse?

There Is No One-Size-Fits-All Conversion Rate | FutureNow’s GrokDotCom / Marketing Optimization Blog.
Amen! I’ve always been saying that there are no one-size-fits all average conversion rates. Your goal should be to increase your current rates and do so by optimizing your efforts. And you need to take a look at your rates individually (traffic source, etc.) versus in aggregate.

What’s the Best Way to Improve Conversions? eMarketer, October 29, 2009

I can’t say it any better than eMarketer — it’s through testing and analysis.  

According to an Econsultancy study, companies considered “A/B testing as the most valuable way to improve conversion, with more than one-half of companies saying it was highly valuable and another 42% saying it was quite valuable”, while agency-side respondents considered cart abandonment analysis as the  most valuable way to increase conversions. 

I agree with both.  While you can probably drive your biggest, immediate impact by determining why shoppers are abandoning your shopping cart (assuming you make immediate fixes), you should also test. 

Where do they drop off?  And why?  Are they surprised at the shipping costs, which aren’t revealed until the second page in the cart?  Is it a legal disclaimer that’s causing folks to abandon your cart?  Do you have too many steps in the process? 

You won’t know until you investigate.  And you won’t know what (and which change) works best, unless you test it.   Look at your web analytics to determine where the highest drop off points are.  Launch a shopping cart abandonment survey to get more insights (real customer feedback).  And if you’re making changes, test them to determine what will get you the highest conversion rate. 

I’ve run shopping cart abandonment surveys in the past, which uncovered issues we didn’t know existed (thanks to shoppers).   We also provided the “abandonner” with the option to be contacted by our customer service team, which was then able to reach out, provide immediate support (while the site issue was being addressed) and close the sale over the phone.   Of course we also ran tests to determine which changes would yield highest conversion rates.

And to quote Bryan Eisenberg in Always Be Testing:
“It verges on insanity to look at the success of a company such as Amazon.com and still hold out hope that avoiding testing will produce the results that are well within your grasp… Through testing, you empower customers to collectively decide what works best for them… Intelligent testing removes opinion, guesswork, and faulty assumption from the marketing equation. It gives you truly meaningful results upon which you can act.”