Extra! Extra!

When we consult with companies to redesign their web presence, there are often key goals we work towards  including increasing leads, increasing sales, improving support, etc.  When considering all of these goals, an overarching pillar should not be overlooked: with content re-writes, changing URLs, and updated navigation – ensure that past achievements in organic SEO are not lost. 

When redesigning a site, to maintain (and ideally improve) your organic SEO traffic, make sure to keep an eye on the follow:

1. Links:  Most SEO experts agree that quality inbound links are the most influential variable in SEO and possibly the most difficult variable to influence.  Realizing substantial inbound links with quality linking text is tough work.  Make a list (probably a big list) of all inbound links. There are many tools for doing this and the best tools are free (Google Web Master Tools & Yahoo Site Explorer). We recommend using both – Yahoo will typically provide a more comprehensive list and Google will provide a more qualified list of links. What you should do:  For your most popular links, try to keep the same URL pointing to similar content on your new site.  For less popular links, implement a 301 redirect to point to the most similar (relevant) content on your new site.  Here are some link stats to be aware of:

a. Unique Links:  Track the number of unique links into the site (all pages).  If you want to be an advanced search manager, you can track the Google PageRank of each page linking to your site.– One inbound link with a PageRank of 8 is worth more than 8 inbound links from pages with a Page Rank of 1.

b. Linked Domains:  Make a comprehensive list of the unique domains (and links per domain) of sites that are linked to yours. 

c. Linked Pages:  The pages that have the highest amount of inbound links are typically your most valuable real estate.  Of course, you’ll want to pay close attention to those URLs and the corresponding content.

d. Internal Links:  Internal links are important for multiple reasons.  They help search engine crawlers find all pages on your site.  Strong internal links help demonstrate relevance to search engines – it’s intuitive that the pages that you link to the most are the most relevant.  What you should do:  Make it a goal that you will increase the amount of quality internal links to your top performing pages.  You can do this with persistent/global navigation, content navigation, quick links, site maps, etc. 

2.  Landing Pages:  We love landing pages because they typically convert visitors at higher ratios.  But obviously they need to serve SEO purposes as well.  Tips:

a. Focus on your top landing pages (that convert)

i. Top Search Landing Pages:  Within your analytics application, make a list of the top (>1% of  traffic) landing pages for organic search, especially for non-branded keyphrases; filter out the company name(s), acquired company names, popular product names, etc. 

 ii.  Look at the keyphrase density and styling of top landing pages:  Look at how well you are targeting specific phrases.  We use Ranks.nl to do this, but there are a lot of informative tools on the market.  Essentially you want the layout and styles of a page to help emphasize relevance of specific phrases.  Look at how well you are using titles, headers, links, bold/italic fonts, and more to focus on your keyphrases. 

iii. Search Engine “Friendly”:  It’s worth using a validation tool (i.e. W3C, SoreSite) to look at how well your landing pages comply with XHTML and CSS standards.  Essentially you want to make sure that your new site is either achieves the same level or higher of compliance (fewer errors and fewer warnings). 

3. Organic Keyphrases:  Keep track of which keyphrases are working now.  You shouldn’t have to worry to much about branded phrases.  Pay close attention to the non-branded phrases which are drawing a strong amount of traffic and a good amount of conversion (> 2%). What to do:  If you have non-branded phrases that are showing good conversion rates  - maintain that with similar URLs, similar landing pages, and similar keyphrase targeting.  This does not prohibit a content rewrite.  Just be mindful of the new site’s affect on keyphrase targeting. 

4. Update Sitemap / Redirects:  Both Google and Yahoo have some decent tools for informing search engines which pages have new URLs.  But your best bet is to create redirects (301/302) for all your high profile URLs and create new site maps (HTML & XML) for the search engines to reference.  It used to take months to see index updates in search engines, this now takes days or weeks. 

Benchmarks, benchmarks, benchmarks:  Keep track of monthly data for all the variables discussed in this post.  If SEO traffic (or corresponding conversions drop), look to these variables to see a significant variation. 

Being aware of these variables when creating new navigation, content, layouts, and styles will help you maintain your SEO standings.  Keep in mind, that even if you maintain your quality links, improve your keyphrase targeting,  and make everything search engine friendly.

As an avid member of the Web Analytics Association for the last 5 years, I’ve sat through many conversations about how important it is for web teams to have someone dedicated to web analytics.  The role started as a data sage with expertise in one or more of the major web analytics platforms.  The role has been a little slow to catch on with only bigger teams having dedicated web analytics members.  

Today, the tools are vastly different and as a result, web analytics, while still critical is a smaller piece of the data arsenal.  The Digital Strategists looks for answers, not necessarily in the web analytics, but wherever the data is.  When we discuss the necessary skill sets of the web analysts of tomorrow, we look for someone with a significant knowledge of each of the following tool types.

It’s important to remember that the best web analysts demonstrate a significant ability to analyze data and develop insight. The source of the data is almost irrelevant as long as the data and the insight are accurate and actionable.

Video is fantastic for user engagement.  We frequently see video positively affecting important click-paths and/or funnel entries.  Video also often positively affects contributing factors such as Repeat Visitors (%), Time On Site, and Page View Per Session. 

With all of the comedic bold statements about the future of Flash between Steve Jobs and Shantanu Narayen, as well as the recent failed attempt of YouTube to convert its massive video platform to HTML5, the focus seems to be on whether or not Flash is going to have such as significant hold on the future of the media on the web.  But it makes me wonder, what about other media platforms.  Currently Flash Video is supported on 96%+ of the PCs followed by Windows Media (~69%), and QuickTime (~61%).  These 3 video formats dominate the way we all watch video on the web.

HTML5 has not officially declared which video formats will be supported, but so far the video formats discussed are very interesting.  The initial supported video format in HTML5 was the open source OGG format.  Other non-common formats such as Theos were added at later dates. Undoubtedly, over the next 12-24 months, HTML5 video will become more advanced and start to be able to scale to sites such as Hulu and YouTube, but should Adobe be the only organization afraid about losing market share?  The open HTML5 video formats will likely cause Apple (QuickTime) and Microsoft (Windows Media) to lose some share of the market as well.   If a significant segment of the video market is dominated by open source video formats, perhaps organizations such as Adobe and Apple should start focusing on having superior video creation tools.

Periodically we find that some designers find Sharepoint 2007 to be a limiting platform.  This often stems from the amount of time it takes to build out free-formed design elements and/or integrating media components such as Flash.  To set the record straight, it is not that Sharepoint can’t do free-formed, table-less designs – it’s that most Sharepoint engineers have been trained to do rectilinear pages and it therefore takes more time to do free formed designs. 

The newly released Sharepoint 2010 makes good strides to eliminate some of the hurtles in designing unique public facing websites.  As Elisabeth Olson described at Mix10, SharePoint 2010 allows engineers to implement “pixel perfect” designs.  SharePoint 2010 gives designers more freedom with enhancements in the following areas:

1. Complete layout management with CSS, XHTML, and Div’s – Tabless layouts! - in other words you don’t have to fight SharePoint to implement non-rectangular designs.  Some Webparts do still require rectangluar Tables, but the designers can create nice liquid graphics.  There are also new capabilities in terms of placing Widgets outside of rectangular Zones.
2. Enhanced reusable controls with elements such as ribbons, dialogs, and status bars.
3. New functions for Blogs and Wikis allow engineers to implement this functionality within SharePoint 2010 as opposed to 2007 which favored integration with outside systems.
4. Better browser support for IE 7, IE 8, Firefox, Safari, and some mobile browsers.
5. Less effort is required to create unique Theme’s for simple designs.
6. Enhanced development features in SharePoint Designer enable engineers to make code updates and preview layout changes simultaneously significantly decreasing the amount of time it takes to implement unique designs. 
7. Better functions for pre-rending elements of a page for faster loading design elements. 
8. Completely new Theme functions for managing design elements (i.e. fonts), custom CSS can be used for generating themes. 
9. My favorite – improved integration with JavaScript.  A completely new set of JavaScript libraries that grab dynamic data from SharePoint (i.e. ActiveX, JQuery) allows designers to use cool JavaScript functions within a site. Client OM allows engineers to easily get data out of SharePoint in JavaScript without web services.  jQuery now comes as a built-in JavaScript library.  In fact, SharePoint 2010 uses jQuery all over the new interface
10. The Silverlight Web Part is completely updated allowing multiple dynamic data feeds for rich media designs
11. Designers and Web Developers now have at their disposal an arsenal of methods to access SharePoint data and present it in the most creative ways such as using the new Client Object Model right within a Silverlight animation.
12. External Content Types allow the user experience team to more easily include dynamic data outside of SharePoint (i.e. product pricing information). 
13. With Business Connectivity Services (BCS) an engineer can easily mashup data from existing backend systems and easily display, sort, filter and report on dynamic date right from SharePoint 2010 pages.
14. SharePoint 2010 now has more social features such as content tagging and the ability for users to “Like” posts by their peers and colleagues on various content

There are some features that have not changed considerably.  The templates still rely on .Net 2.0 Master Templates, XSLT is still the favored platform for managing presentation, and only a few changes in menu controls.  Web Part Zones are still rectangular although this is not a significant limitation anymore.  SharePoint engineers still speak a unique language when discussing elements such as templates and widgets. 

Both the designers and the egnineers are exicted about these features and we are looking forward to seeing new sites taking SharePoint 2010 in new directions. Some of our favorite SharePoint site designs are below:

• AMD Gaming
• Ferrari
• Carlsberg (A good site and a good beer)
• Hawaiin Airlines
• EDP
• Cimail Solutions
• New Holland Agriculture
• Entre Voisons 

*Written by Mark Ryan and Oscar Medina

There is a lot of talk about Social Media and the new opportunities that it brings to organizations.  Large untapped communities, unprecedented demographic data that can be used for better targeting,  loyal communities spending more time online than any other marketing channel in history, and the list goes on and on.    Unless you’re a game developer or a trendy charity, it can be extremely difficult to harness the  power of these communities.  Honest relevant content becomes significantly more critical and it needs to be updated as often as the community wants it. 

I was at eMetrics Toronto last week and I had the opportunity to talk with web analysts about their Social Media campaigns.  As an analyst passionate about conversion opportunities, it only makes sense that I look at Social Media as an opportunity to get more leads, revenues, customers, cross-sell opportunities, etc.  I heard about a lot of innovative campaigns and organizations that were very successful creating large communities on sites such as Facebook and Twitter.  But little talk centered on ROI.  I believe that for most sites to achieve ROI they must attract social traffic from a social media site (i.e. MySpace) to their own site for conversion.  For example, if I am selling trendy shoelaces, I would start/moderate a community around tennis shoes as well as target advertise on other shoe communities such as facebook.com/adidas/or myspace.com/nike/.  But ultimately, my ROI comes when the social community navigates to my ecommerce site and buys shoe laces. 

This got me curious about some of the behavioral data around traffic from social networking sites.  I started to study traffic on 21 different sites.   I created a focus group of 21 sites across a wide variety of industries comprising everything from brokerages to tattoo parlors to national banks to local yoga studios to international network equipment manufacturers, encompassing both B2B and B2C sites, Fortune 500 sites as well as small organizations. I looked at the traffic sources for these sites and segmented traffic from Facebook, MySpace, Twitter (and various tweets), Delicious, LinkedIn, and Yelp.  I then compared basic data points such as repeat traffic, time on site, and page views per session to see if there are trends in the general quality of social media traffic. 

Below are some general trends that I found:

1. Traffic from Social Media sites typically accounts for less than 1% of all site traffic.  I found this in both my research group as well as casual conversations with many web analysts at eMetrics.

2. While the average page views per session was 12.8, Social traffic only averaged 3.8 page views per session.  Social traffic viewed 70% less pages in a visit than average visitors. 

3. Social traffic spent 17.2% less time on site than average visitors.

4. The bounce rate of Social traffic was 9.9% higher than average visitors.

5. Visitors from Social sites were 7.7% less likely to return. 

Considering the amount of time it takes to maintain a social media campaign, these numbers looked a little grim. But remember, I picked these sites at random. 

As I looked through the 304 traffic sources on 21 sites accounting for approximately 27 million visits I found some successful components.  Check out site #3

The visitors from social media sites to Site #3 are viewing more pages per session, staying twice as long as other traffic and are approximately 50% more likely to return.  I was impressed and curious; what was Site #3 doing differently?  This B2B company with a small B2C offering wasn’t selling products that are popular in social communities such as games, retail items, or charitable contributions.  Looking at the Social Campaigns for Site #3, it was obvious why they have such great Social Traffic – they are taking Social Networking seriously.  While Extractable comes across many companies that have some of the lower members of the marketing teams maintaining the Social Communities, Site #3 had LinkedIn contributions from the CEO an average of 4 times per month.  The company blog has 7 different contributors on a regular basis.  Their Facebook community gets to see relevant industry research several times a week.  Site #3 is clearly making Social Media a priority and is seeing the benefits of it.  Site #3 is not a large organization with a lot of resources to throw around; they are a small company with good focus.

For each of the sites researched, conversion came in different formats.  Some companies are selling products on their own site (like Site #3), some were selling products on distributor sites, some were doing lead generation, and so on.  Therefore it was difficult to produce a straight comparison of ROI between the sites.  But it is obvious from the data above that conversion rates on traffic from Social Sites was either lower than average or non-existent.  Determining the costs of Social Campaigns is challenging as the majority of resources are usually internal content writers.  Generating an ROI analysis for Social Marketing is as easy SEO ROI reports.  The analytics applications allow analysts to generate conversion reports isolated to traffic that originated from Social Media sites.  From there, an organization needs to calculate the amount of time and people spending time on content for Social campaigns.  The following is an example from one of the sites. 

NOTE: As a precursor to this blog post, I asked a group of Web Analysts on LinkedIn if they saw traffic from Social Networking sites converting at different ratios and I got some interesting responses:  www.linkedin.com/groupAnswers

One of the strongest Project Managers that I have ever worked with, Nick Giordano, is also the lead guitarist for the band Monsters are not Myths.   Yesterday, Nick walked me through some of the analytics that he monitors on his band’s fans base (groupie analytics?) on both Facebook and MySpace. 

Facebook Insights has come a long way in the last 12 months. After creating a page and gaining fans, administrators can see a large variety of reports. Most of the reports focus on two main areas (1) How successful a page is at creating interaction within a community and (2) How successful Facebook advertisements are at getting Facebook users to act (convert).  The most impressive piece of data that you can find is your Post Quality.  This variable measures engagement by tracking each of your posts (text, video, photos, etc) and how many of your fans comment, like, or write on your wall.  While the report does not necessarily tell you which demographics in your fan base react more than others, it gives you an excellent idea of how strong your content is.  This would be an awesome feature for any site.  While Facebook puts a little too much emphasis on their simple demographics data (literally just age and gender), the export feature and the graphs on data such as media consumption give administrators an excellent understanding of their ability to create an interactive community.  Facebook Insights would be more useful if they offered the ability to integrate with offsite analytics applications such as SiteCatalyst, Webtrends, or Google Analytics to show downstream analytics – one can only dream.  With some work, an analyst could start to track trends between the applications – for instance, when Facebook Post Quality goes up, Facebook traffic to our site (and conversion) increases.

While the MySpace Dashboards are not as user friendly, the data offered is still very impressive.  The dashboards provide similar demographics to Facebook with age, gender, and location being the primary tracking variables.  MySpace gives a little more insight into interactions by showing posts, comments, views, and kudos over days, weeks, and totals (Facebook focuses on the last 7 days).  What impresses me the most about the MySpace dashboards is their unique variables suited for the goals of the administrator.  For example, Nick’s band uses the MySpace music player to let fans play the band’s songs and their dashboard shows how popular different songs are on MySpace.  It measures the song popularity by plays, dedications, iLikes and trending data.  For Monsters are not Myths, “Old Soul”  tops their MySpace charts. 

Most importantly, Nick can log into the bands Google Analytics account and see how traffic from Facebook and MySpace to their band’s main webpage (http://www.monstersarenotmyths.com/) interacts with the site.  For instance, he can see if visitors to the site that came from MySpace, Facebook, or Twitter were more likely to be repeat visitors, viewed more content, and/or signed up for newsletters and events.  To isolate traffic from URLs posted on a fan page versus random users posting URL’s on a social networking site, be sure to append tracking variables to the URLs you post.  For instance, if Nick was to post his tour dates the URL might look like this - http://www.monstersarenotmyths.com/v2/music_shows.html?source=facebook.  If you are using a URL shortener such as ow.ly, tr.im, or bit.ly, be sure to append the URL variables before you shorten them. 

http://www.Facebook.com/pages/Monsters-are-not-Myths/6306011877

http://www.myspace.com/monstersarenotmyths

I am feeling a little philosophical this morning and thinking about some of the behaviors of web analysts. 

It’s funny how most web analysts present the positive side of analytics.  They may produce reports which illustrate that 9.8% of the visitors converted to quality leads, that 749 users signed up for the newsletter, and that 56% of the customers came back for a second purchase.  The traffic reports often start with the keyphrases that get the best click-through-ratios and the best conversion rates.  Web analysts seem to be always looking on the bright side of the data.   The analysts are also often complaining that their web teams are not acting on the data they report.

I wonder if the analysts focused their reports on the 90.2%  of the visitors who didn’t convert, the 6,567 users who didn’t sign up for the newsletter, and the 44% of the visitors who never came back – would the web teams be more likely to act on the data? It might help site owners to make more objective decisions if all reporting tools showed both sides of the story.  For instance, traffic reports might show the 98 visitors that came from organic search results as well as the 7,528 visitors that searched for our phrases but decided not to click on our link.

While the buzz around micro-blogging is all about what is new and cool, marketing groups should realize that a successful Twitter imitative takes a significant amount of work and though you can’t thoroughly track your Twitter activities with your traditional web analytics suite, you should be tracking trends in your Twitter community. Many tweeters out there are simply posting 140 characters about the main course of each meal and/or retweeting every article they see online (or on other tweets).   Extraneous banter may create an active feed but it will not create a loyal fan base or ROI.  

A successful Twitter community takes continuous thought about what your followers want to read.  If your content is good, the analytics should prove it.  The following is a list of some variable that you can track on your Twitter initiative:

1. Traffic to your site: Track the amount of traffic to your site from Twitter (i.e. ow.ly,is.gd,bit.ly) as well as from Twitter aggregators (i.e. LinkedIn, Facebook).  Be sure that when you post articles to Twitter you’re embedding URL parameters to help track how much traffic is coming directly from Twitter.  This report will help you determine ROI for your Twitter initiative.  Typically the best ROI for a Twitter campaign is cost effective traffic that converts. 
2. Traffic from your site: Track how much traffic you are sending to Twitter and how much of it is coming back. This is often surprising to site owners how much a link to Twitter can influence abandonment.
3. Followers / New Sign Ups: Keep a count of how many new followers you tweets are getting on a daily or weekly basis.  Remember that followers are not always reading your tweets!  A follower is not necessarily an impression.  One common mistake is to assume that your followers are reading your every word – remember that approximately 60% of Twitters don’t come back to the site after their first month.
4. Tweets: In order to make sure your content is fresh, record the number of tweets you produce daily/weekly.  This may come in handy if you need to correlate it with the growth of your community.  If you content is good, you may find that your community grows proportionately with your tweets. 
5. Replys, Retweets, Lists:  How many replys or retweets are your post getting?  This is best indicator that your posts are getting read and appreciated.  Dive deeper and determine if it is a small group of followers that are retweeting often or a broad group that is retweeting when they see fit.
6. Competitors: Many of the Twitter tracking tools out there allow you to track the variables above on your tweets as well as your competitors.  While I don’t typically see a lot of value in watching the competition, this may give you a good indicator of how well you are interacting with your community (or how well you should).

Determining ROI in Twitter can be difficult.  The most challenging step is determining the costs in creating a successful Twitter initiative.  Use of the website is free, the followers (if your copy is good) are free. Typically the largest cost is the time put into producing relevant tweets.  As you plan and implement the Twitter campaign, keep track of your time and associate a fair value for all time spent (Cost = time spent * value of time).  The return on your investments comes from the traffic you derive from Twitter that converts on your site. 

Below are some great tools for simplifying you data gathering on your Twitter campaigns.

1. Great for tracking article promotions – twiterurly
2. Great for tracking community development – twittercounter or twitteranalyzer
3. Learning about your community – tweeps and twitter.grader.com
4. Hashtag tracking – whatthehashtag

A lot of focus is put on using web analytics data to make incremental improvements in site performance.  But as audiences, brands, products, and technologies change – sometimes a site needs a complete redesign.  When it’s time for a complete site overhaul, it’s important to dive deep into the web analytics for the site early in the process.

Data and insight derived from web analytics should be a key driving force in the creation of wireframes, information architecture, and usability testing.  The following is a partial list of bits of data that help User Experience experts make informed decisions, when creating the layout for a new site that will more effectively achieve the intended goals:

• Navigation Preferences: By categorizing your navigation types into groups such as drop-down menus, primary navigation, content navigation, quick-links, content links, text links, graphic/button links, flash/media links, etc you can start to see how visitors prefer to navigate across the site.    This becomes more exciting when you are able to compare these numbers with the navigation preferences of other sites. 
• Visitor Tasks:  Content popularity and/or click-paths are essentially the visitor showing you what tasks they are performing successfully (conversion) or unsuccessfully (abandonment). 
• Vocabulary / Nomenclature / Taxonomy:  The visitors are telling you how they describe and categorize their needs in 3 important ways.  First, the choice of words and phrases they use on public search engines such as Google.  Second, their choice of words and phrases on your site search engine (i.e. Endeca).  Thirdly, they are exposing what words make them click within the navigation and content in the site (as well as what words they don’t click on). 
• User Research:  Often overlooked is the data that tells us what company the visitor is at when navigating the site as well as demographic information such as if they are using an expensive ISP or a more practical one.  Their technical aptitude can be loosely derived from their connection speed, browser, operating system, and plug-in preferences.  The language, region, country, state, and county of your visitors also helps in defining web marketing strategies.   
• Conformity / Variance: Analyzing the deviation in variables such as visitation (frequency / latency), click-through-ratios, and search trends shows degrees of loyalty and dependencies within your visitors during a sales cycle/product research.  

Generating a report of these variables and more helps the entire team review and evolve site layouts more intelligently.  Often in projects we find that if the wireframes and information architecture are data-driven, the revision of designs becomes an informed process rather than one of personal preferences.

Many of our clients ask us how to justify larger web marketing budgets to their executives.  One of the obvious places to start is an ROI analysis or a cost/benefit analysis.  If there is no straight forward conversion event (i.e. Lead Generation or eCommerce), it can be difficult to measure how effective web marketing is.  In some cases, web budgets need to be analyzed as a cost of doing business.

Marketing budgets are often divided up into channels/segments such as events, print, mail, television, radio, etc.  The sales team, often operating on a separate budget, actually spends budget on some channels with the same goals as marketing - such as inside sales.

When analyzing how the marketing budgets should grow (or shrink) an organization should analyze how the different channels stack up against each other for common variables. 

For example, as a touch-point (impression), ask your team to quantify how often each channel serves as a touch-point for potential and current customers.

Another example would be to put a cost to each impression.  Take the total budget per channel and divide it by the number of touch-points (impressions). 

From these numbers, as a channel for interaction with potential and current customers, the website and email are the most cost effective.  Adding conversion rates to this table may paint a different picture.  Analyzing budgets in this fashion shows that online channels are more cost effective at interacting with potential and current customers.