According to Jupiter Research, 73% of marketers aren’t doing any testing whatsoever, with 49% having no plans to do so. Before I share 4 compelling reasons why you should be testing, let’s look at some of the reasons why marketers aren’t testing:

• Lack of resources (in-house talent, budget for tool or consultants)
• Lack of tool (now with Google Website Optimizer this is less of an issue)
• Lack of buy-in from upper management
• Lack of ownership of the testing process
• IT bottlenecks
• Lack of understanding of the value of testing / don’t think it’s important
• Not sure what to test, how to prioritize testing opportunities or how to take action on results
• Lack of understanding of tools or testing methods

Despite these challenges, serious Web marketers should push to overcome them and get testing capabilities because testing is the only way to truly know what works and what doesn’t for your site based on the combination of your industry, customers and product mix.

Tools such as Omniture Insight and Webtrends Segments offer significant analytical and visualisation capabilities but leave much to be desired in terms of modelling and future trending. Google Analytics is only starting to address these issues scraping the surface with its Intelligence feature.

In comes SPSS, IBM’s acquired predictive analytics software and solutions firm. IBM is now in a position to merge Coremetrics’ web and mobile analytics collection platform with its powerful statistical modelling software. Such analytical capability is common practice in the offline CRM and Direct Marketing worlds but so lacking in the online world.

 

Web analysts spend significant amounts of time guesstimating which visitor segments are most important to their business and how to break down users’ behaviour in order to optimise the online customer experience. With SPSS’ decision tree algorithms analysts will no longer need to go through this trial and error process. Instead the software will tell them which visitor segments they should focus their attention on. They would then be able to approach the data in a much more constructive way gaining deeper insight faster.

The Social Marketing Hub was designed to make the social media management and tracking responsibilities manageable for large enterprise organizations. The platform does so by presenting users with the ability to create content — photos, blog posts, video, files — and automatically post it to all or select social sites with just a few clicks.

These destination sites are referred to as channels, and the Hub supports updates to all of the following: Facebook (), Twitter (), YouTube, WordPress (), Flickr () and Awareness-branded online communities. Of course, the Hub includes management for multiple accounts on each of the various channels.

The aggregate data covered in the new dashboard essentially highlights Facebook () users’ sharing behavior related to both content on and off the site with better data, graphs and charts.

Users can visualize top level info, as well as drill down to look at specific metrics around individual stories and comments “Liked” on the page. Graphs in the Insights dashboard can even be saved, printed, or upsized and viewed in full screen mode.

Domain administrators who associate their domain with a user ID, page or application can view detailed — yet anonymous — demographic data for each domain. These demographic visualizations cover gender, age, country, city and language breakdowns for a comprehensive look at audience makeup.

Application developers can get information about referral traffic to their applications, feedback on stream stories and other significant data on user engagement with the application.

The refreshed Insights dashboard appears to be designed to prove to publishers and application developers that sharing and the new Instant Personalization feature are being heavily adopted by users. It’s a smart upgrade that will better help business-oriented Facebook users measure progress within the social network and on their own pages.

Here are our predictions in the web analytics industry for 2010:

1. Multivariate testing and site optimization will cross the chasm and become an imperative for online businesses and marketing departments of any reasonable size. Next generation solutions will leverage proven approaches, ease of use, and expert services to enable this transition.
2. 2010 will be the year that integration of online with other enterprise data will take off. Consumers are demanding that companies they engage with take into account their cross-channel behavior during interaction and as we go into 2010, if consumers are not receiving this tailored, informed interaction, they will find a new company to buy from.
3. Interactive marketing will continue to gain adoption in 2010, as we’ll see email marketing, web analytics, and traditional campaign management vendors race to become the owners of the “hub” for interactive marketing, along with optimization and analytics.
4. 2010 will define the principals for social marketing and lay the foundation for the next decade of marketing. As people recognize that media is still media, and that social is about behavior, the social behavior theory will emerge. Social search will heat up among top search players, thus the nature of SEO will give way to SSO and enterprise social platforms will hit a tipping point.
5. Mobile applications continued to grow throughout 2009 and will explode in 2010 and 2011. In 2010, Apple will continue to grow as their exclusivity with AT&T expires next year and Android will accelerate exponentially as more droid phones will be brought into the market and developers will surge at the openness of the platform. This will give way to application analytics that will be generate significant buzz within the web analytics industry.

At this year’s Omniture Summit, Omniture customers were able to share insights from their companies’ data-driven journeys during a roundtable-format session. Most of the table discussions seemed to focus on aligning the implementation with the company’s goals through a clear measurement strategy. Here is a sampling of some key takeaways or discussion points that the client moderators shared with me:

Many companies don’t have a clearly defined web measurement strategy because they have too many cooks (departments) in the kitchen, they focus on soft goals that are difficult to measure, they are indecisive in separating what should be measured from all of the things that can be measured, and they don’t appreciate the full value of what web analytics can do for their organization beyond reporting.

Without a clear measurement strategy, you end up with mountains of reports and data which aren’t necessarily helpful to the business.

The best web strategies are focused on driving insights and actionable intelligence to improve future performance, not just a scorecard on past performance.

The two essential elements in establishing a web measurement strategy are defining/articulating the business goals and determining the KPIs that measure or influence how successful the company is doing in meeting those goals.

Determining the business goals seems pretty straightforward, but is often difficult and contentious in practice, especially in larger organizations.

When interviewing stakeholders for your measurement strategy, interview at both management and individual contributor levels. Then circle back with the larger group to introduce findings, gain buy-in, and plan next steps.

Focus on the highest value KPIs (they’re called “Key” Performance Indicators for a reason).

In some cases there is no way of directly measuring performance, so proxy indicators need to be used.

KPIs should cover not just the top-line results (e.g. sales, leads, etc) but also the key levers or “supporting/contributing metrics” that impact those top-line KPIs. For example, if the top-line goal is increased sales, the supporting metrics may include things like conversion rate, average order size, leads, and retention rate.

Different parts of the organization may have different reporting needs in terms of which KPIs are important, the level of granularity required, and the area of the business they are interested in.

In a recent comparison for CBS Television on conversations around the show NCIS, Sentiment360 found as much as a 50% swing in sentiment, or lack there of, in machine vs. human analysis. Essentially, once you let humans analyze the data, the machine-produced results are crap.

Here’s a run-down of their comparison:

• Sentiment360 used an unnamed social media monitoring service to collect the data. (They’re tool agnostic and say they use several different ones depending upon the client need. Companies they report as part of their arsenal include Radian6 and ScoutLabs.)
• The results showed 50,000 conversations around the show in a given month with the search performed in March of this year.
• According to the service’s automated scoring, of the 50,000 conversations, 84 percent were neutral or passive mentions, 11 percent were positive and five percent were negative. NCIS is talked about a lot, and more positively than negatively, but the vast majority of the conversations don’t hold an identifiable opinion.
• Sentiment360 pulled a sample of 3,000 of those conversations and had their analysts go to work. So the results show just the human analysis of the sample, but the sample is six percent of the total data set, far better than most market research firms offer.

And here’s what Sentiment360’s analysis found. Some of these numbers astounded me:

• 23 percent of the entries were irrelevant. They mentioned NCIS or linked to the show, but contained no other qualifying information about the show or were spam sites.
• Once the irrelevant entries were removed, only 30 percent of the entries reviewed were found to be neutral or passive, a 54 percent difference than the machine analysis.
• Human analysis found that 63 percent of the online conversation around NCIS was positive, not 11 percent as the machine asserted.

As we jump into the brave new world of paid “twinfomercials,” a few points bear remembering when deciding how best to leverage this social media sea-change:

1. The Acceptance of Advertising. Twitter’s core demographics are familiar with paid online advertisements. They largely accept that the best things in life cannot remain free if providers fail to leverage the financial opportunities they bring about. They do, however, want their preferred platforms to be as discreet as possible in pursuit of profit. And, above all, they want full disclosure of how their engagement is capitalized.

Twitter’s approach to advertising has so far been one of the more transparent and open to feedback we’ve seen in the social media space. And by remaining transparent to its community, as Twitter has successfully done to date, the platform could avoid many of the same pitfalls that continue to plague just about every new update Facebook has let loose on its half-billion users.

2. Advertising Alone Won’t Do the Trick. Despite the eager hopes of corporate communicators and brand managers who might be looking for the path of least resistance, promoted tweets are no silver bullet. They, and whatever future developments roll out in the months ahead, are but one instrument in a symphony of Twitter engagement tools. Used in coordination with sustained, value-driven engagement, they will certainly help to amplify messaging. Made to stand alone, however, they will unquestionably fail in a marketplace that values continuous personal interaction.

3. Tailor Your Advertising to Twitter. Corporate marketers need to step up to the plate and take it upon themselves to create value-driven Twitter ads. This is not another opportunity to rehash a Facebook Mafia Wars advertisement or show users how they too can have a ripped six pack. Advertisers need to sit down and think about how content is used and shared on Twitter, and, importantly, what makes a Tweet (paid or not) successful.

Meteor has built an analytics solution that tracks how content is shared. It dynamically creates a social graph that uses first-party tracking tags in the URL of hosted content. When a person copies and pastes that URL into some vehicle, and someone clicks on it, Meteor captures that activity and creates a new node. If one of those people then copies and pastes the link, it creates another node.

This is a very simple mechanism that takes advantage of standard behavior. When an article, video, game, ad, or any other content is tracked this way, the content owner can understand how that content was disseminated from a root source (such as a publisher site, corporate site, or even a UGC sharing site) and where it ended up.

Custom variables is one of the most powerful features in Google Analytics. With them you can segment traffic by almost any attribute. We recently hosted a webinar on how to make the most effective use of custom variables. If you missed it, you can now watch this highly informative webinar on the Google Analytics YouTube channel.