I returned this weekend from a week’s holiday and for the first time in 2 years, I didn’t take my laptop with me on vacation. (The magic of Wordpress’ scheduled posts kept Get Elastic alive while I was gone). With 250+ blog posts chilling in my RSS reader, I couldn’t wait to catch up on what I missed in the world of retail, marketing and tech geekery.

One of the events that happened while I was away was SMX Advanced in Seattle (Search Marketing Expo). Fortunately there’s always a ton of liveblogging coverage, as often breaking news from search engines get announced at these events, like support for the canonical URL tag. Because search engines are constantly working on improving their own tools and minimizing search engine spam, the “rules” and best practices for SEO (search engine optimization) also change. It’s important for SEO professionals, marketers and webmasters must stay on top of these changes as not to give outdated advice, and for bloggers to update old posts that may contain outdated advice.

While catching up I learned 2 important things about how Google follows links on a website:

1. How Google Handles the Nofollow Attribute

In 2007, the SEO world was a-buzz with a new trick - PageRank sculpting. The idea was you could control the flow of PageRank between pages of your site by plugging up “leaks” to pages like Contact and Privacy, so more PageRank would be applied to your product and category pages. (If you’re not familiar with the PageRank concept, please refer to this video explanation).

I recommended Stephan Spencer’s concept of PageRank sculpting for retailers in late 2007 as a “Killer SEO Trick Only 1% of Online Retailers Use” and referenced the practice in 9 Privacy Policy Usability Tips, Tips for SEO Friendly Affiliate Programs and Dodging Duplicate Content Filters While Assisting Affiliates.

What we understand now is that Google no longer treats the nofollow attribute the same, and the “trick” doesn’t have the same benefit as it had before. The nofollow attribute will still prevent PageRank from passing to nofollowed links, but there is no boost to links without the attribute - the juice just “evaporates.” If you’ve used the technique before, there’s no harm, there’s just no benefit anymore. The disappointing thing is that if you have a large number of links on one page (including links in comments on blog posts), they still dilute the link value of more important links on the page.

This is a perfect example why any internal SEO expert or SEO consultant you may be working with reads blogs, attends conferences (or at least keeps up with the event coverage) and stays on top of the industry, otherwise you may get advice that is either a waste of time or at worst, get you banned from search engines. It’s also important for bloggers like me to update old posts to reflect new information.

2. How Google Handles Javascript and Flash

Equally if not more important, your web developers should understand how Google and other search engines handle Flash, Flex, AJAX and Javascript. Google annnounced advances in searchability of Javascript and Flash at its own Google I/O event, and Vanessa Fox’s explanation is a must read for all of your Web developers. Whether you’re working with internal or outsourced devs, send them this article today.

This is the final installment of a 4 part series on mobile commerce design and usability…

Part 1: Home Pages and Navigation
Part 2: Search and Category Pages
Part 3: Product Pages and Cart Summary
Part 4: Forms and Checkout

This series is based on a review of 10 mobile ecommerce sites: Best Buy, Target, Sephora, Moosejaw (old and new design), Barnes and Noble, Amazon, Sears2Go, Ralph Lauren and Tickets.com. (Links point to mobile versions of each site)

Forms and Checkout Process

As I mentioned in Part 3: Product Pages and Cart Summary, Best Buy, Sephora, Target and Moosejaw’s new site don’t support mobile checkout. Sites mentioned in this post are Barnes and Noble, Amazon, Sears2Go, Moosejaw’s old site and Ralph Lauren.

I typically recommend ecommerce sites to include the following in their checkout processes to maximize conversion:

• Allow guest checkout option
• Provide security assurances and trust seals
• Link to privacy policy
• Ask for an email address in the first step of checkout
• Don’t ask for more information than necessary in form fields
• Mark required fields with an asterisk (*)
• Allow customer to copy billing and shipping addresses (use tickbox for shipping address “same as billing address”)
• Provide a telephone number for assistance at each step of checkout
• Store contents of shopping carts in cookies whenever possible, if customer abandons cart the items will still be there in a new session
• Use as few steps as possible and show a progress indicator
• Use personalization/tagging to identify customers with a promo code, and hide the coupon code field from others

When “checking out” mobile checkout processes, I noticed that none of them showed the progress indicator, linked to a privacy policy or had a “click to call customer service” link (perhaps because you can’t talk on your handset and input information at the same time like you can on a computer). I was unable to tests personalization/tagging capabilities.

At least one mobile site followed each of the rest of the best practices:

Guest Checkout Option

Nobody wants to be forced to register before checkout on the Web, how much more on a mobile device!

Barnes and Noble states explicitly that registration is not required, and hints that registration has a benefit — faster checkout next time:

Sears2Go also explains the benefit of registration and allows guest checkout, but puts the guest checkout option first and includes a bold button with the label Guest Checkout. Sears understands customers don’t necessarily read text, and Guest Checkout is self-explanatory.

Showing guest checkout first can improve conversion, as the customer is less likely to assume he/she needs an account or will be forced to create one to proceed. This approach assumes the customer wants the fastest checkout possible, which customers appreciate.

Ralph Lauren combines the sign in page with the first step of checkout (billing and shipping information). This approach can reduce the total number of steps.

Amazon’s simple log in is consistent with its Web checkout. It’s actually a required registration — but it appears so easy it doesn’t feel like a forced reg, your account creates itself as you go through the checkout process.

The downside to customer log in is it relies on email and password combinations which returning shoppers often forget (you have to remember at least your email address to request a password hint). Many simply create a new account - making it harder to reconcile customer information (one loyal customers looks like 5 flaky ones). This is an issue for both online and mobile channels.

Security Assurances

Despite the public’s fear of mobile payment security (as discussed in our Multichannel 2.0 webinar), none of these mobile sites had security assurances except for Amazon’s link “Why using a credit card is safe”

Ask For Email in First Step

The earlier in the checkout process you ask for an email address, the easier it is to follow up with a customer if the cart is abandoned with a triggered recovery email. Though the idea of chasing after a customer for changing his or her mind is controversial, many retailers have managed to save sales this way.

Of course, Amazon’s first step is the email address. The rest ask for an email address after 8-10 required fields on the Billing/Shipping page. Because some folks are paranoid about giving away any contact information, B&N includes “Your email address is required so that we can contact you about your order” and Sears2Go uses “Tell us where to send your confirmation email.” It might help to include a one-liner about your email privacy policy (i.e. you won’t share the address or send promotional emails.) Unfortunately Ralph Lauren has no assurances, and even pre-checks the email list permission box.

Perceived Difficulty of Forms

The longer the form, the more tedious it appears and the less the customer wants to bother with it. But most customers recognize asterisks (*) separate the required fields from the optional, so it’s key to include them.

It’s not a good idea to provide unconventional instructions like *Bold fields are required (which fields are bold below?). Never assume your customer is going to notice.

Allow Quick Copy of Billing Address to Shipping

This is standard for new customers, while returning customers and account holders should be able to select from existing shipping addresses and billing methods, like Amazon:

The only thing that’s confusing on a mobile screen is which button corresponds to which address when you have more than 1 or 2 to choose from. A little triangle marker pointing to the address on each button would help. An incorrect shipping address is a major problem.

A side note: Moosejaw’s old site (the new site has no checkout) had a time-saving feature where you enter your phone number in the first step, and based on your area code your city, state and country are pre-filled in the Billing Address step.

Estimated Arrival Times

I’ve mentioned before it’s a good idea to show actual dates of estimated arrival rather than “X business days.” It requires less thinking for the customer. Of the 4 sites, only Sears2Go provided dates rather than days:

Edit Cart in Process

Many retailers like Amazon remove navigation when a customer enters the checkout process in hopes to keep the customer in the funnel, but this can be a problem if the customer wants to make a last minute edit or add more products to the cart. Barnes and Noble allows cart editing, and Sears2Go allows an escape back Home or to the Shopping Cart summary. Even Amazon has a “Shopping Cart” link in it’s mobile checkout.

Alternative Payments / Pay by Phone

I didn’t notice any alternative payment options like PayPal mobile, but Barnes and Noble interestingly provided a Pay By Phone option. This isn’t a bill-to-phone-bill option (as discussed in the Multichannel 2.0 webinar), rather you can “Call 1-800-843-2665 (1-800-THE-BOOK)” to complete your order. You will be given a confirmation number to cite when you call after you have completed your order (on your phone and by email).

Perhaps some customers perceive this as more secure than entering a credit card over an unsecured mobile network, although placing an order by telephone carries its own security risks. It might also be a headache for Barnes and Noble if customers complete orders and don’t follow through with the phone call authorization.

Final Thoughts

With only a handful of reference sites, this series is not intended to be a best practices guide, rather a collection of recommendations based on observations. Please keep in mind I was viewing these sites with the iPhone and experiences on other devices may vary.

Most usability guidelines for the WWW carry over to Mobile Web, while common usability problems are exacerbated by mobile devices’ smaller screens and keyboards (or lack thereof), flakiness of mobile Internet connections and lack of standards between mobile designs, browsers and operating systems.

My recommendation is to have a mobile friendly site if the mobile channel is part of your retail strategy (whether for transactions or just customer service). When designing for the mobile Web, make sure you check out competitors’ sites in your industry, test your site on many different devices and involve real users in your testing.

I would love to hear your thoughts on your experience on mobile commerce sites, on this blog series or about your own efforts with mobile commerce in the comments.

Welcome to PPC Myth week! Today is the first installment of a 3 part series challenging common misconceptions about search marketing and analytics.

Myth #1: Organic search more qualified traffic than paid

I was surprised to see in print one of the most respected search marketing gurus state “Organic searchers who click on your pages are highly qualified visitors to your site. They are much more likely to make a purchase than some other kinds of visitors you receive.”

In fairness, the guru went on to explain that banner ad clickers are less qualified than searchers actively looking for a product in a search engine. Nevertheless — to claim that organic searchers are highly qualified is false. It also implies that organic search converts better than paid search, comparison engines, email traffic, affiliate leads and so on. This just ain’t so.

1. SOME organic traffic is better “qualified” than others.

Remember, in this context “qualified” means more likely to purchase. If you look through your organic search referring keywords, you’ll find a number of non-transactional terms, and transactional terms that are not necessarily close to purchase or even relevant to what you offer.

Examples from the 2010 Olympic Store:

• Non-transactional: “vancouver 2010 schedules”
• Transactional, not relevant to our offer: “how do i get tickets for the 2010 winter olympics”
• Transactional, too general: “business card holders” (may like our offering but is likely in research/comparison mode)
• Qualified: “vancouver 2010 sterling silver heart charm bracelet”

Also, organic conversion can vary by search engine. It’s possible for your market, traffic from Yahoo, AOL or MSN sends you more shoppers and Google sends you more information hunters.

2. SEO vs. PPC - it depends on the keywords.

PPC traffic “quality” also depends on which keywords get clicked - especially if you’re using the broad match type. In fact, broad match can trigger some really un-qualified traffic. If you were only bidding on a certain number of close-to-purchase keywords with the exact match type - you *could* argue PPC is more qualified than SEO if your conversion rates also confirm so.

3. Other channels - it depends…

Comparison engine traffic is *typically* closer to purchase since visitors have already evaluated your offer against competitors and the product against other alternatives, comparison engine traffic should convert better in theory. Your results may vary.

Similarly, email and affiliate referrals have been exposed to your brand and offer before clicking through - you’d expect better results for these channels than search. Again, your results may vary.

Type in traffic (no search engine or other site referred the visit) indicates brand awareness, and perhaps preference. Repeat customers, brick-and-mortar customers or people responding to offline advertising may convert higher than SEO/PPC traffic that’s also clicking on several other results to compare. But direct traffic can also indicate you should filter out your own staff’s IP address or you have missed Javascript tags on some pages (causing a null reference).

So what’s the point of this rant? I don’t want anyone making decisions to invest more into SEO than other channels because they heard that organic search is the most qualified traffic. I don’t want you to set the wrong expectations on organic search, and set goals like “increase organic visits” or “increase conversion for organic visitors.”

A common SEO problem for ecommerce sites is CMS (content management systems) that create different URLs for a product that lives under multiple categories. The main reason this is bad for SEO is search engines only allocate so much bandwidth to crawling your site. If most or all of your product pages have duplicates, you’re less likely to get your site fully crawled and indexed — meaning lost organic search opportunity.

The above shows 6 copies of the product page for Abercrombie’s Clarissa skirt are currently indexed in Google. Half of the links lead to a 404 page, the rest look like this:

http://www.abercrombie.ca/webapp/wcs/stores/servlet/ProductDisplay?storeId=11306&catalogId=10901&productId=482314&langId=-1&categoryId=12280&parentCategoryId=12203&colorSequence=01

http://www.abercrombie.ca/webapp/wcs/stores/servlet/ProductDisplay?langId=-1&storeId=11306&productId=482314&parentCategoryId=12203&catalogId=10901&categoryId=12280

http://www.abercrombie.ca/webapp/wcs/stores/servlet/product_11306_10901_482314_-1_12280_12203

The best practice is to use a global alias for the product page URL for the UI to look up and render instead of the category-specific URL. If you wish to maintain breadcrumb trails, use a session ID or cookie to track which categories the customer clicked to locate the product. If the visitor lands on the page without browsing through a category menu (search engine referral, affiliate link, PPC ad, email, site search etc), default to a parent category.

Keep in mind that session IDs can get crawled and indexed by search engines, creating even worse duplicate content issues (and security issues with some ecommerce platforms). To block search engines from crawling URLs with session IDs, use this syntax in your robots.txt file:

User-agent: *
Disallow: /*?

Other benefits of reducing duplicate content is it prevents Page Rank dilution and may simplify your web analytics (product page views and conversions aren’t spread over multiple URLs).

The “Big 3″ search engines (Google, Yahoo and MSN) broke big news last week with the announcement they will support a Canonical URL tag to help webmasters (and search engines) better manage duplicate content issues. Duplicate content refers to identical or near-identical blocks of text on more than one page in a search engine’s index.

Examples of duplicate content on ecommerce sites:

• A product is listed under multiple categories, each with its own URL
• The search engine crawls a site and is issued a session ID. It indexes links with the session ID
• A blogger copies a product link with a session ID or navigation tracking parameter like
  http://www.site.com/B00DK/ref=acc_glance_sw_ai_549_1_img and unwittingly pastes the link as-is in a blog post
• An affiliate link like http://www.amazon.com/Logitech-Cordless-Laser-Mouse/?affid=1234 gets crawled and indexed
• Content is duplicated across sub-domains or sub-folders like canada.yoursite.com or yoursite.com/uk/
• The search engine crawls your print friendly version

Duplicate content problems include:

• When multiple copies exist, search engines want to choose one version to show in search results and filter the rest. They may not show the version you want (print friendly or worse, an affiliate URL so you’re paying commissions on organic search conversions!)
• Your SEO suffers because PageRank is diluted across several copies of a page (what is PageRank?)
• Your site might not get fully crawled by the search engine as search engines will only give you so much attention in a given session

Duplicate content can also occur across domains, like multi-stores with country-specific domain extensions like yoursite.co.uk or if many retailers are using stock manufacturer descriptions. The Canonical URL Tag does not remedy this situation.

Up until February, 2009, webmasters dealt with duplicate content by “sculpting PageRank” with rel=”nofollow” attributes, rel=”noindex” or using 301 permanent redirects. Now you can specify which is the original version of your content with the tag and rel=”canonical” attribute in a page’s head section, like:

<link rel="canonical" href="http://estore.com/womens/sweaters/esprit/B3H4H5"/>

Which *should* nudge search engines in the right direction when choosing which URL to display.

I say *should* because search engines consider this a hint rather than a directive - much like my hairdresser, you can give your suggestion but it’s going to do whatever the heck it wants. Nevertheless, I believe this will help a lot of online retailers’ SEO efforts and reduce the headache of duplicate content.

Bonus if you got the “canon” and “shot” pun in the title, yes I know it’s kinda lame.