People are wrong in their criticisms of the Apple social media policy that was allegedly leaked recently, and here is why:

A company’s social media policy should support the unique qualities that make the company successful. In fact, the elements of a successful social media policy must exist in concert with the unique culture and business context of any organization.

In Apple’s case, secrecy has been a critical key to success. While many social media pundits claim that Apple should be more open, very few of those people are running billion-dollar corporations, and the notion that all companies should apply the same level of “open-ness” is, at best, over-simplified.

Comparisons to other technology companies abound, but that just makes no sense. Those companies and Apple take completely different approaches to differentiation, which has led them to create very different cultures. They also rely on different business processes to create growth and value. Many are well-run and highly successful, but for very different reasons.

And those differences are the keys to understanding why they use different social media policies.

The chart above shows the business factors that companies should consider when developing an effective social media policy (which I published in The Social Media Management Handbook earlier this year).

Google announced today that searches by logged-in users will be securely sent over SSL encryption. Therefore, Google will no longer send the query terms in the referrer data to analytics tools that analysts use to understand the keywords sending traffic to their site. If you’re not using Google Webmaster Tools, you will no longer know all of the keywords bringing people to your website from searchers who are logged into Google when they search. This is potentially a big problem for folks relying solely on enterprise analytics solutions.

While Google says they are focused on protecting user privacy, the change clearly forces everyone to use Webmaster Tools and decreases the value of paid analytics solutions because they will no longer be able to collect as much data as they did before.

If you have not verified your website through Webmaster Tools, do it now.

It is only a matter of time before Google incorporates Webmaster Tools directly into Analytics. Until then, marketers will have to pull a portion of their search performance reports from Webmaster Tools.

Now that Google, Twitter and Facebook implemented SSL this year, privacy advocates will likely expect the same from Bing and Yahoo.

View the original blog post from Google.

Bit.ly recently published data indicating that links shortened in bit.ly and shared on Facebook, Twitter and YouTube receive the vast majority of their clicks within 3 hours of being shared. While clicks can be a critical call-to-action on social media, and the bit.ly data help us to understand the dynamics of viral sharing, viral distribution is only one important goal of content that is created or distributed through social media. Specifically, a balanced social media marketing plan must also include provisions for evergreen content: feature articles that deliver significant long-term impact — especially in Search Engine Optimization (SEO).

Evergreen Versus Real-Time Content

Evergreen content sticks around and continues to provide value after the 3-6 hours of initial sharing. For example, think about search engine results. In general, we see blog posts, wikis, reviews, forum threads, and videos with the longest staying power in search results. Perishable media such as Facebook and Twitter updates are important in search results, but evergreen content is equally as important.

According to Lijit, 20% of referrals are driven by social media, and search still delivers twice the amount of traffic versus social. Ignoring the role of search in your social strategy will simply lead to all of your hard work being lost to time.

As marketers increase their focus on real-time marketing, the concept of a timeless article is increasingly derided by some in journalism and public relations because its very definition denotes something that is not immediately newsworthy. Rather than appreciate the long-term benefits of evergreen content, they treat it as filler for a slow news day.

However, marketers must resist the urge to chase sharing statistics at the cost of search engine performance. Yes, social media affect search engine results, but they are not the primary determinant of search engine performance, so don’t let the spirit of real-time take your long-term marketing performance off track.

Caveat on the Bit.ly Data

While bit.ly offers nice analytics for link sharing, their raîson d’être is to shorten URLs for use where space and attention are limited. Few people use bit.ly in longer-form social media because companies typically own those properties and are able to use their own more robust analytics platform to track internal links.

The previously referenced Lijit study also showed that StumbleUpon drives as much inbound traffic as Facebook, yet this major social media player is not included in the bit.ly analysis. That is because the toolbar design of StumbleUpon makes sharing shortened links difficult due to bit.ly’s inherent URL redirection.

Therefore, while insightful, the bit.ly data does not adequately represent all social media sharing and clicking that occurs on the web so consider the source when citing the report.

Michael Maoz recently wrote that CIOs these days shake their heads at the fact that marketing, sales and services leaders are able to obtain funding for social media projects without a business case, instead of being held accountable for the same level of quantitative rigor as other IT-enabled investments. While it is true that most social media investments still travel with no business case, anyone who wants to change that fact needs to undertand a bit of history:

One challenge is that most communications professionals and social media consultants don’t have much experience in organizational change. They’ve never led cross-functional change programs. They’ve never built a business case that had to stand up to the CFO’s rigor. So they just don’t know how to do those things.

And, in the corporate communications arena, they never had to measure business impact from their efforts. Clippings were all they ever counted.

But that is all changing as marketing, sales and customer service leaders begin to ask for real dollars for social media.

However, the one critical factor that is changing the slowest is that CIOs are simply not getting in the game. CIOs and their teams are simply not at the table when cross-functional social media efforts are launched. And, ultimately, the CIO has to change that. CIOs need to start reaching out to their VPs of Communications and Marketing, and start figuring out how enterprise IT will enable the business goals that social media supports.

The bottom line is that CIOs can not sit back and wait for other functional leaders to bring them a business case or a well-defined social application architecture. CIOs need to get out of their foxholes, and go be the smartest person in the room about how the organization should use technology to solve challenges in marketing, communications, sales and service.

Dr. Brent Coker of the University of Melbourne recently published findings indicating that web users tend to trust web sites 20% more today versus 2007, but are 30% less loyal to ecommerce sites versus 2007.

1. Why He Believes Trust Increased

Dr. Coker said the increase in online consumer trust is largely linked to the visual appeal of websites. “As aesthetically orientated humans, we’re psychologically hardwired to trust beautiful people, and the same goes for websites. With websites becoming increasingly attractive and including more trimmings, this creates a greater feeling of trustworthiness and professionalism in online consumers.”

Anyone interested in web credibility should also visit the Web Credibility Project at Stanford University.

2. Why He Believes Loyalty Decreased

“The biggest source of frustration is the inability to find relevant information on a website. The best way to stop defection to other websites, and increase loyalty, is to be interesting. Being pretty, but with nothing to say, is not enough.”

The research found that if a website has poor navigation or access to information, or is slow (i.e. more than two seconds to download), web surfers are more likely to opt against purchasing and navigate to an alternate website. (No surprises there.)

However, it is interesting that, in the last five years, the frequency of referring others to websites has increased by 32%. Largely due to social utilities, such as Facebook and Twitter.

Overall, “… we are more trusting of attractive websites, less tolerant of websites that have irrelevant information, and more likely to introduce ourselves to websites that are new.” Dr. Coker says.

“Shopping offline is very different to shopping online. Offline we shop in a large room, with clear signage, and often a sales assistant. Online, however, what we want to buy is buried somewhere, and we’re left to find it on our own.”

3. Converseon Thoughts

First, consumer trust is harmed where brands blur advertising with authentic conversation. On the other hand, when brands engage in helpful and sincere dialog with consumers, trust and engagement grow.

Specifically, we quantify engagement growth through performance scorecards that we develop for our clients, wherein we correlate content attributes with digital and social performance metrics.

Second, with more brands spending on social, we are seeing an increase in poorly conceived and executed social campaigns that don’t respect the norms of social platforms and conversations they seek to engage. As a result, many brands simply fail to achieve meaningful engagement within their targeted communities.

For many brands, the poor conception and execution results from inadequate and inconsistent performance metrics — where performance feedback in inadequate, or teams are allowed to spend without accountability for business outcomes.

Dr. Coker’s Methodology

Dr. Coker developed a formula to track patterns and trends in online behaviors and purchasing, called Webreep (shown in the image below, which you can click to enlarge). The formula, called ‘Webreep’, creates a score for 130 industries based on seven dimensions of quality: visual appeal, trustworthiness, ease of use, search quality, information quality, information relevancy and load speed. Webreep started mapping the internet in 2007.

He will present his paper at the 2011 World Congress in Computer Science, Computer Engineering, and Applied Computing in Las Vegas.

For information about Converseon performance measurement solutions, please contact us.

I just finished participating in a Digiday panel on big data management in beautiful Deer Valley, UT where I warned about an over-infatuation with technology, to the exclusions of people.  Indeed, many conversations today make it seem as if humans are simply a tangental voyeur to the vast processing intelligence of our evolving algorithms.

That is far from the truth.

Machines and technology do some things wonderfully well.  They churn through vast amounts of data — searching for anomalies and patterns — and machines help to filter the signals from the noise. But the signals have to be interpreted by humans to spark the insights that can change a business. No machine has yet developed much of a creative streak.

We see this every day in our Conversation Mining technology.  Instead of using algorithms to interpret the meaning of human conversation on the web, we use machines for what they do best: to find and identify the obvious conversations, and look for patterns of meaning that go beyond what humans can generally perceive.

Then, we use people to interpret the nuances of slang, sarcasm and implicit sentiment in the tweets, mentions and status updates flowing through the online conversation.  That’s what we call human coding.

We then use that human coding to make the machine even smarter by letting the machine identify patterns based upon the coding that was created by our people.  It’s a constantly learning algorithm, informed by human intelligence.

But that human coding is not true insight… yet. The human-coded data must be analysed by people with the skills to turn data into insight, and, then, business decisions.

In the social media listening and intelligence space, there is a widespread misperception about “human coding” versus  ”human analysis.” They are two distinct capabilities, using different tools and skill sets, but working together to help leaders make smarter decisions.

The first “human layer” is the consumption and reporting of the records themselves — or, human coding.   Brands who consider doing this themselves rapidly face the insurmountable challenge of the vast volume of conversation to be coded —  and not just coded — but coded as general consumers would code them.

For many clients, understanding the conversation about their brand, their products, competitors and their category often requires analysis of tens of thousands of records per day.   Coding the deluge with internal brand resources simply does not scale because brands don’t have the army of humans and systems required.   And it’s probably not the best use of their time.

Even if they had the resources, would that be the best approach?   Brand analysts cannot help but see social conversations through the prism of their own experiences and knowledge of the brand.   Coding records with brand staff leads to biased results because the conversations are not consumed and perceived through the lense of an average customer.

Converseon tackles this human coding problem through highly scalable, proprietary systems that enable us to process tens of thousands of conversations — or more — daily, in very rapid fashion.   And, by putting in place inter-coder reliability processes (in short, three humans need to agree on a record for it to be considered accurate) we ensure highly scaled, unbiased, accurate human coding at a fraction of the time and cost that it would take teams of brand analysts.  Better yet, this human coding is done by general consumers absent of any inherent bias, so they interpret conversation just like your customers would.

The second level — Human Analysis — is the interpretation of the coded conversation for insights and integration into marketing, communication and other uses.  On this level, brand teams are often essential to translate what they see into action within the enterprise.  For most of our clients, this is a collaborative process.  Our research teams will identify what they see, applying our team of cultural anthropologists, social media strategist and more, while client internal teams access the coded data directly for their interpretation.

This is where the eureka moments emerge — where human intelligence and creativity are essential. As Henry Ford famously said, “If I asked my customers what they wanted, they would have told me a faster horse.” Only human creativity can deliver the breakthrough ideas that provide competitive advantage and change the world.

For the vast majority of brands with constrained resources, brand staff typically add the most value through analysis, rather than coding. But the muddied distinctions have sometimes put brand staff in the unenviable and impossible place of trying to do both (coding and analysis).

So, next time you hear about the “human element” in social media intelligence, remember the difference between human coding of the records themselves and the human analysis that interprets data into insight and action.

As Information Week recently cited, Converseon is about the humanization of social listening. As such, we give brands highly scalable human coding capabilities, as well as sophisticated analyses of the coded data in collaboration with client staff. If you would like to know more, drop us a line any time.

While studies have shown that positive online ratings correlate with higher sales, a study at the University of Maryland using data from BazaarVoice found that brands maximize their long-term benefits from online ratings when they attract a balance of positive and negative ratings from customers.

If you want to increase sales in the long run, you need to attract a balance of negative reviews, in addition to lots of positive reviews.

Here’s why:

Research Findings

In The Value of Social Dynamics in Online Ratings Forums, Wendy Moe and Michael Trusov of the University of Maryland found the following:

Recommendations

Moe and Trusov say that, “… marketers should not necessarily encourage strictly positive word-of-mouth, but also encourage variance of opinions in the online discussion.” Rather than asking only the happiest customers to write reviews, ask for feedback from everyone, and ask for constructive comments.

Seek multiple points-of-view that balance the discussion while keeping the average rating high. If you have a great product or service, the community will help to balance unfair negative reviews with more positive reviews.

A few bad reviews add balance that can help boost future sales and product ratings.

Professor Wendy Moe can be reached at wmoe@rhsmith.umd.edu.

Anyone running affiliate marketing or influencer outreach programs requires a mixture of automated and human analyses to design and operate their program.  Automated algorithms are great when you need a quick decision in real-time, but when you are choosing 5, 10 or 50 influencers for long-term relationship development, you need to be sure that they are the true influencers in your category.  That requires human analysis.

While Klout can now include LinkedIn connections and activities into the calculation of Klout scores, brands should be very careful about using Klout scores for affiliate marketing or influencer outreach, for the following reasons:

1. The only way to identify influencers within a category is through a combination of automated and human analyses.
2. Klout can not measure influence within a category.  For example, Ariana Huffington has a high Klout score, but she is not relevant to most brands.  For example, if you sell baby diapers or desktop virtualization products, Ariana is not an influencer.  And Klout is not capable of telling you who influences the conversation around baby diapers or desktop virtualization products.

In recognition of the need for category-specific influence scoring, Klout recently launched a +K button, which lets Klout users tell Klout when someone else influences them, but the method has two weaknesses:

• +K is subject to significant self-selection bias. The inputs come only from Klout users who choose to contribute, and
• +K does not capture the extent to which one person passes along another person’s messages. Therefore, the scoring does not adequately allow a brand to choose influencers based upon the extent to which the influencer will drive messaging into the market.
• +K does not associate the influence with a category. For example, I might say that Seth Godin influences me, but does he influence my decisions about car purchases, laptop purchases or the foods that I consume? No.

Converseon’s monitoring dashboard, Conversation Miner, includes Klout scores for authors because we recognize the value of Klout in real-time decisions.

For our clients who are managing influence outreach or affiliate marketing, we offer more sophisticated research services to identify the top influencers in the categories that the brand is targeting.

Both are useful tools. The key is to choose the appropriate tool for each purpose.

UPDATE: Alex Braunstein posted a nice analysis of Klout’s usability on his blog.

This post was co-created by Chris Boudreaux and Jasper Snyder of Converseon.

Jasper Snyder oversees all Listening and Research efforts for Converseon clients. Jasper can be reached at via email at jsnyder@converseon.com or via Twitter.

Converseon recently launched a collaborative research project with comScore, the Advertising Research Foundation (ARF), Communispace, and Firefly Millward Brown to explore the roles of social media in the purchase process.  The project brings together a range of research techniques to understand how and when people turn to social media as they make purchase decisions.

Converseon will provide insights based on social intelligence for the project, helping the ARF to establish an expanded understanding of online conversations around purchase decisions for items in the CPG and other categories.  Converseon analysis will uncover where in the purchase process social media conversation is taking place, who is taking part in that discussion and where it’s happening.

Our research will include psychographic and demographic analysis, in addition to a range of other metrics based on Converseon’ unique technology + human methodology, giving the ARF the opportunity to cross-tab the analysis and extract the most meaningful information to support its research goals.

We all look forward to the results of the joint project.

This afternoon at the WOMMA School of WOM in Chicago, I had the privilege of sharing lessons from enabling the social workforce at IBM with my clients, Susan Emerick and Bill Chamberlain of IBM. You can view our slides here, including a few transcript notes at the bottom of the page.

If you’d like to know more, all of our contact information is listed in the slides, or feel free to contact me:
@cboudreaux
cboudreaux [at] converseon [dot] com