Manhattan Marketing Maven

A Snapshot of Social Media Metrics

Eight out of ten agencies are monitoring social media for clients. Though there are too many tools, too few benchmarks and too few ways to integrate data from multiple sources according to the 4A’s inaugural “Agency Social Media Monitoring and Reporting Survey.” 

With “significant variation in social media monitoring, reporting and associated business practices from one agency to another” … more than half of respondents felt they were doing a good job for clients, even though many clients are still not convinced of the value of social media. Agencies, of all shapes and sizes, are looking for better ways to understand what’s being said about brands across a growing array of social media platforms. 

Google Analytics, used by 70 percent of responding agencies, leads the pack in terms of monitoring and reporting tools. Hootsuite and Radian6 with a 59 share and a 36 share, respectively, are the next most used tools. One third of responding agencies spend between $1000 and $4999 per month licensing tool sets. No other tool got more than 20% of responses in an arena where 52% of agencies believe the current tools available are insufficient and seven in ten have trouble evaluating and differentiating between tools. 

The outputs, most often measured by these tools, and presented using PowerPoint or Keynote, are engagement, connections/fans/followers and brand mentions. Agencies are clearly grappling with defining social media success. Just 18 months ago brands and their agencies were racing to accumulate fans and followers. Today we’re touting ill-defined engagement scores and counting posts, shares and comments as we attempt to explain to clients how these social activities impact brand awareness, preference and purchase or deliver incremental media value. 

Bottom line: Social media monitoring and reporting is still in it’s infancy. We’re not sure what to measure or how in order to quantify the value of social media to clients looking for marked increases in branding, sales metrics or ROI.

Facebook Goes Direct

From the moment, the first brand on Facebook collected 10,000 fans, marketers have been asking about the relationship between “likes” and sales. Initially Facebook found a dozen ways to deflect the conversation. 

Facebook doesn’t really sell, they’d say. Its about relationships, buzz and engagement. Facebook is a tool for generating comment, creating warm fuzzies and enrolling fans as advocates for products and services and creating some sense of loyalty. 

But the guys paying the bills kept asking. How many fans, followers or likers are buying my products? How many are referring friends? How many are buying repeatedly? Finally Facebook created ad units to address these ROI concerns. 

The latest innovation, launched in May called Facebook Exchange (FBX), is inventory in the News Feed (the primo real estate on Facebook). FBX cannot be bought directly from Facebook. Units must be purchased only through an FBX qualified company. Unlike right hand side (RHS) Facebook ads, which are sold on a cost-per-thousand (CPM) or cost-per click (CPC) basis, FBX inventory is priced as a cost-per-insertion, limited to one each day. 

These units, a 154x154 pixel image and several lines of copy, were created to enable precision targeting, links to outside websites where marketers actually sell stuff and pixel tracking to enable retargeting on Facebook and across the web. Consumers can like, share and comment on FBX ads, so a measure of virility is baked in. 

Adroll, a Facebook retargeting partner and FBX qualified company, looked at 547 advertisers across industry verticals spending between $1000/month and $20,000/month to generate over one billion impressions using FBX and using ad inventory on the right hand side of everyone’s page. They zeroed-in on retargeting, their specialty, and compared results against traditional display and retargeting inventory from online ad exchanges like Google’s DoubleClick, Yahoo’s Right Medias Exchange and AppNexus. 

The research validates FBX alone and in combination with RHS as a potential direct marketing tool. News Feed retargeting had a click-thru-rate (CTR) 49% greater than RHS and 21% better that standard web retargeting. News Feed cost-per-clicks (CPCs) was half of RHS ads and 1/5^th of web retargeting. Conversion fared less well. FBX ads yielded 9% fewer clicks than the web and 16% less than RHS, which are generally considered anemic by direct marketers. 

FBX plus RHS drove more clicks at lower costs, even though combined their reach is significantly less than standard web targeting. Finding the right inventory mix will be our next challenge. In the first seven days of adding FBX to RHS, Adroll discovered a 62% increase in total clicks with an average decrease in CPCs of 30 percent. In the campaigns studied FBX was just 0.5% of the impressions but drove 14.7% of the clicks – promising results. 

Clearly Facebook is trying to create the right inventory to help brands link social activity with sales and loyalty. FBX is a step in the right direction worthy of further testing.

Run Better Social Media Contests

Contests are a staple of social media. They generate leads, opt-ins, sharing, comments and virility. The conventional wisdom is that simple contests spawn cost efficient customer engagement that makes people happy. 

Consumers are comfortable with the give-to-win proposition. They will give up bits of personal information, post or tweet, upload a photo or a comment, vote, take a survey or refer somebody in return for a chance to win. Few calculate the odds. 

Many contests and giveaways are designed for lead generation. Recently Unbounce released a study, based on 3 million visitors to 100 websites using their platform between March 15 and April 15^th, which promises to improve contest performance and lead conversion. For example, contests with prizes of $500 or more generated 700% more subscribers than landing pages without contests. 

Here’s their POV. 

Set an end date. People don’t respond to open ended contests because they figure they have plenty of time to do it later. By specifying a start and an end date, you build-in urgency. Consumers understand that the clock is ticking. 

Use the word “Giveaway.” “Promotion” sounds like it might or might not be a sale. “Sweepstakes” feels too chancy. “Giveaway” resulted in 27% more conversions than sweepstakes and 50% more than promotion because it connotes a freebie. 

Like Gating Sucks. This trick has been played out. Widgets that asked for email addresses drove 80% more subscriptions than the request to “like” a page. The like gate has become a signal for consumers to bug out. 

Bottom Right, Brand Names and Photos Win. The contest sign up box in the bottom right corner of the page drew 125% more subscriptions than bottom left. Similarly when the promotion box had a picture, almost any picture, conversion increased by 22 percent. When a brand name is attached to the giveaway details conversions saw a 28% spike. 

We have trained consumers where to look and what to look for. Don’t swim against this tide. 

Signal Ease of Entry. Site visitors are lazy. They want all the goodies but don’t want to work for it. When you direct them to “enter in seconds” versus a big honking “enter” button, opt-ins increased by 33 percent. Tell them how easy it is. 

Optimizing giveaways and contests is a matter of tweaking and re-tweaking. These simple tips suggest, as Michael Aagaad suggests, “value plus relevance equals more conversions.” This is true for contest rules and prizing strategy as well as the subtleties of constructing effective landing pages.

New Visual Engagement Insights

A picture’s worth a thousand words as regular users of Facebook, Instagram, Tumblr, Twitter and Pinterest know all too well. A new visual study from Curalate offers unusual insights into which images resonate and engage us best. 

Curalate, a visual marketing and analytics firm, looked at 500,000 images on Pinterest across 30 visual dimensions. You can get their infographic here. The goal was to identify the visual elements that drove greater re-pins.   

According to CEO Apu Gupta humans interacted with pictures before words because sight is our most developed sense. He speculated, “We initially started painting pictures on walls which were more emotional than text and are a minimal commitment.” Pictures, he says, offer so much context that minimal copy is needed to explain them or to orient the viewer. 

Here’s a quick summary of the findings: 

• Red/orange images get 2X the re-pins of blue images
• Multi-colors are re-pinned 3.23X more than pictures with just 1 dominant color
• Medium lit images get 20X re-pins over darker images
• Images with less than 30% whitespace are re-pinned the most
• Images that are 50% saturated get 10X more re-pins
• Vertical images with an aspect ratio of 2.3-4.5 are re-pinned 60X more often
• Pictures with less than 10% background get 2-4X more re-pins
• Pictures with smooth textures get re-pinned 17X more than rough textures
• Brand images without faces get 23% more re-pins than those with faces 

To some extent, these findings are an insight into our reptilian brains. They are also a reflection of how we have been visually trained by media since childhood. The lack of interest in seeing people is counterintuitive especially when you consider that the underlying psychology revealed in this study suggests that warm, clear, proportional, colorful and familiar images are the things that draw us in, capture our imaginations and motivate us to share. This is a big step toward a visual formula for increased engagement.

Tumblr's Advertising Challenge

The cool kids have abandoned Facebook and headed to Tumblr.  Twenty-nine million unique visitors signed in four days each week to gain access to 44 billion posts on 102 million blogs. The average user logged-in for 154 minutes and looked at 30 pages per visit. One in eight used a mobile device to tumble. 

Tumblr users skew male, young (18-34), single, childless and rich (1/3 have household income greater than $100K). One in five is Hispanic and one in six is in the Pacific Time zone. 

Now that they’ve drawn a crowd, Tumblr is trying to figure out how to make a buck. They have a “Radar” feature, which highlights curated posts and is supposedly seen by everyone, although I can’t figure out which posts in my feed are the “Radar” ones. I checked my dashboard and couldn’t find an ad even though they sell access at a minimum of $25,000 a pop. Supposedly, a bunch of well known brands have used the platform. But I’ve never seen hide nor hare of them, nor do I follow any of them. Maybe they appear in the 50-odd content categories that range from Actors, Cute and Gaming to Poetry, Street Style, and TV. 

They’ve also rolled out a mobile ad unit, initially embraced by GE, Warner Brothers and ABC, that is served four times each day in their iOS and Android app. And they insist that ads appear as posts rather than as ads. I’ve never seen one of these puppies either. 

Buying Tumblr for brands is a challenge.  Blog content is highly visual and idiosyncratic. Like Facebook users only see content from those they follow. Unlike Facebook you can follow anyone without his or her blessing. So brands will have to develop significant followings to get substantial reach and or frequency against desirable segments. 

The content categories aren’t channels per se just convenient ways to find blogs to follow. And they don’t reflect anything by a tiny selection of the 102 million blogs. Brands can’t frame up appropriate messaging because there is no common experience. Each of the 29 million users follows a different set of bloggers for 29 million different reasons and nobody has crunched the numbers to determine what the patterns and affinities might be. 

The other consideration, beyond advertising competition from Facebook, Twitter and LinkedIn, is an undefined user experience and customer expectation set. Brands need to know why people use Tumblr and how either the people and their moods and behaviors differ from the other social networks. And we haven’t even started to talk about qualifying the audience or determining product and service use. It’s hard to imagine running contests, begging for Likes or distributing coupons to this crowd on this platform. So what’s a brand to do? 

As a focus group of one, for me Tumblr is a diversion; a time waster filled with startling images of people, places and things that I don’t see on Facebook. The 111 people I follow (fewer than the 133 average number of Facebook fans per user) post historical documents, travel shots, photography, cartoons, landscapes, portraits and very little copy. Posting short essays, I’m generally in the minority. And unlike Facebook, I don’t follow any brands and I don’t know or care how many followers I have. It’s a semi-private, self-constructed universe, where intrusive ads would be unwelcome. 

Tumblr faces the classic social media paradox. They’ve developed a sizeable audience but they can’t yet package it and sell it to advertisers. And if they do, will Tumblrs hang around and take in the ads or will they defect and be off to the next cool thing?

9 Ways to Assess Blogs

Reach, credibility and influence are the ultimate objectives of social media marketing. Reaching the right audience in ways that resonate, prompt viral distribution and drive action are the reason we get out of bed in the morning.

In the scramble for attention, credibility and ad dollars the social channels jockey to make their cases. One has volume and global scale. Another has real-time immediacy; a third touts its influence and opinion leadership. A fourth is the playground of the too-cool-for-school crowd. 

Enter Technorati, who replaced the annual state of the blogosphere study, with a new report titled “The 2013 Digital Influence Report” based on data drawn from its blog network with a reach of 130 million unique users per month and a survey that included 6000 influencers, 1200 random consumers and 150 marketers. 

Evidently 75 percent of digital ad dollars are flowing to display, search and video but attention and influence are being accumulated by influential bloggers who get a measly 11 percent of the pie. “Where brands are spending is not fully aligned with how and where consumers are seeing value being influenced,” the report concludes. 

Recognizing the practical difficulties of creating, measuring and buying blog networks, the report claims that “when making overall purchase decisions for consumers blogs trail only behind retail and brand sites.”  Furthermore “blogs were found to be the fifth-most trustworthy source overall for information on the Internet” trailing news sites, Facebook, retail sites and YouTube. 

Self interest aside, nobody disputes the role of blogs or their potential influence on consumers. The question is how to separate signals from noise and how to assess genuine influence from hyperbole. 

Consider ten criteria to assess the potential value of influential bloggers. 

1. What is a blogger’s base audience day-in and day-out?
2. Do they routinely generate significant buzz and virality?
3. Is the buzz about what they think and say or about the blogger as a personality/guru/speaker?
4. Can you immediately identify their insight or expertise?
5. Do they generate original content or endlessly re-tweet others’?
6. Have they written a book, done a TED talk or produced significant data?
7. Are they independent or part of a group?
8. Are they overtly or clandestinely supported or sponsored by someone?
9. Are they looking for freebies and handouts?

Dissecting Facebook's Newsfeed Re-Design

Facebook’s Newsfeed re-design, the first since this feature was introduced in 2006, is aimed at increasing engagement, expanding time spent on the social network, overcoming Facebook fatigue and creating more inventory and more interoperability for advertisers. 

As usual Zuckerberg & Company will benignly dictate how a billion people see, interact and use the platform now described as “a personalized newspaper.” The new design emphasizes larger pictures, cleaner lines and new content streams that will look and feel the same across devices. This recognizes the fact that today a huge number of users access Facebook on tablets and smartphones and that those numbers will continually grow. 

Let’s look at the key changes. 

Bigger Pictures. 50% of Facebook user posts are pictures. The Newsfeed has become a defacto photo sharing service and archive. It makes sense to improve the things your customers like best. 

Increasing the size of pictures, encouraging higher resolution, making it easier to upload albums (multiple pictures) and bumping up the size of thumbnails posted by friends and brands all cater to existing behavior.  Increasing image sizes will make the Newsfeed more magazine-like. In some cases it will do so while 2/3rds of screen is devoted to ads, which also will be bigger and clearer.  

More Discrete Content Streams. By separating out streams of activity by topic, Facebook hopes to create content segments and improve time spent on site by offering users a way to circumvent the Edgerank filtering algorithm. Separate streams featuring music, video, games and music could become ad-supported channels. Users will have the option to see all the content posted by friends, most of which is filtered out today, by selecting “all friends” or from brands that a user likes by selecting  “following.” 

Ideally this will give users more control and more incentive to stay longer and dig deeper into the content they care about most. It will also create opportunities to create inventory for heavily advertised categories where consumers frequently interact with each other. 

Consistent UI. Facebook will look the same on desktops, laptops, tablets and smartphones. This is the first step toward creating a 24/7-device agnostic seamless brand experience. This too, will encourage more frequent access and dramatically improve the mobile experience. Down the road Facebook will sell roadblock ad packages that work across devices to enable brands to make bigger, bolder splashes. 

The implications, of the Newsfeed redesign, for brands and advertisers are four-fold.

1. Brands, like consumers, will have to be more visual using more, better, higher resolution images to get their messages across. Ads will get bigger and more intrusive. 

2. Segmented channels and clusters of users will become available. Brands have to decide the tone, manner, content and frequency with which they will approach them. 

3. Users will have more control over access to their Newsfeeds. Brands will have to earn the right to greater access and sharing by focusing on content and relevance. 

4. Advertisers will need to understand when and how users access Facebook and plan or parse content and ads to optimize awareness, preference and action. 

Facebook's Emerging Ad Strategy

If you pay attention, you can begin to see Facebook’s ad sales strategy coming together. 

The announcement that Facebook is buying Atlas from Microsoft comes on the heels of Facebook’s announcement of big data alliances. The pieces of Facebook’s future advertising juggernaut are emerging. 

First, figure out how to better engage advertisers by cross-tabbing and de-duping their CRM databases with Facebook fan bases so they can plan ads and campaigns more effectively. Second align with big data suppliers to enable segmentation building, intense precise targeting and personalization. Third, use the Atlas infrastructure to build more ways to dice and slice Facebook audiences and to serve Facebook ads with better precision and metrics. The game plan directly addresses the biggest objections in adland. 

The Atlas buy feels like a play to accelerate building infrastructure rather than developing a new third-party ad-serving revenue stream. The value of Atlas, widely considered second tier behind Google’s Doubleclick platform, is a head start on proving the impact and value of Facebook ads to a growing number of skeptical advertisers. With Atlas in-hand, Facebook programmers and strategists can start modifying and adding onto an existing platform, which will give them something to brag about and sell more inventory sooner.  

The combination of these three moves will allow Facebook to answer many of the outstanding questions on targeting, impact, metrics and ROI by big name advertisers who have embraced social media advertising tepidly. 

Data-centricity will fuel better targeting, tighter message strategies and the ability to drive sales in real time. The mother lode of data emerging from the combination of sources, not to mention Facebook profiles, will yield a huge competitive edge in identifying likely prospects, building propensity-to-buy models, scheduling ads and forecasting or driving sales in real time. 

Add this information to growing opportunities for advertisers to track conversion events like registrations or checkouts and the possibly inadvertent slip exposing a new metric called “return value” and you might get the idea that at Facebook, social media is morphing into direct and database marketing.

Social CRM Prompts Skepticism

There’s a lot of talk about social CRM but much less action. The idea of mining social media to develop one-to-one relationships, curry favor with customers and prospects and develop loyalty is a seductive illusion. The reality is that very few companies have a clear understanding of what social CRM is, why they need it or why they should invest in the time, energy or technology to make it happen. 

A recent survey in CRM Magazine found that 41% of a thousand companies surveyed had no CRM program and that 47.5% are either evaluating social CRM or doing it informally, whatever that means. A mere 11% of those surveyed have a formal program in-place. We are looking at a slim segment of early adopters and a great mass of wait-and-see marketers. 

Part of the slow uptake is a growing skepticism about the genuine business value of social media. Part of the hesitation is uncertainty about access to reliable data, doubts about dealing with Facebook, Twitter and their ilk and confusion about the ROI of undertaking a labor intense social CRM effort.  

The leading business goals of social CRM are lead generation, community building, brand awareness, marketplace intelligence, increased web traffic and customer satisfaction. But hardly anyone can connect the dots between the available solutions set and achieving these business goals. To a certain extent, marketers 

The common objective is actionable customer insight. Conventional CRM wisdom is that the more data points; the better. The big software firms offer what appears to be the whole enchilada -- an online-offline aggregation of purchase history, contact, campaign and response history, web traffic, mobile usage, social activity and geographic location. 

The upstarts offer social data and social activity with the promise that these data sets can be married to others but more importantly they can drive real-time communication through pre-fabricated digital, social, mobile or location-based apps or messaging. 

Both sides will provide software-as-service (SaaS), secure cloud storage and consulting help. Most will provide ways to analyze the data or dice and slice on-demand. Unfortunately there are very few successful case studies and nobody is touting their business results from marrying up with either the start-ups or the established players. 

The key to unleashing social CRM is building a business case and positioning this emerging tactic in the context of an overall marketing effort. Making smart social CRM decisions will turn on the answers to these five critical questions, which will be unique for each brand: 

• What are your business objectives for social media?
• What function is social media playing in your marketing mix?
• Do you collect, aggregate and use offline customer data now?
• Can you aggregate, analyze, mine and report on social media data today?
• What will you do with what you learn?

 

Facebook Flirts with Big Data

Facebook has to build a better case for ad sales. It looks like alliances with big data suppliers like Epsilon, Acxiom and Datalogix are a key part of the strategy, according to Ad Age. The operative theory is that more precise targeting yields better results and higher customer satisfaction.

The good news is that this will extend classic direct marketing targeting and segmentation to Facebook and provide marketers with all kinds of inventive ways to target campaigns and allocate budgets. The bad news is that crunching big data and matching it anonymously to Facebook profiles will further muddy Facebook’s loosey-goosey reputation as protectors of customer privacy. 

In recent moves, Facebook allowed marketers to upload data from brand e-mail lists and loyalty programs and crosstab them against brand followers. This answers the burning question among marketers; “Are we hitting the same guys twice in two separate channels or are we addressing two distinct audience segments?” The answer affects message construction, offers, frequency and contact strategy. 

It may also help answer the chicken or egg dilemma; are customers Facebook fans because they love the brand and seek it out across online experiences? Or does Facebook create interest, awareness, preference and loyalty among people who haven’t already opted-in to the brand? 

One of my clients who participated in this “custom audiences” exercise found just 8% overlap between their weekly CRM e-mail list and their Facebook fans. They inferred that they have two disparate audiences with different expectations from the brand. This impacts content strategy and our editorial calendar plus it begs for further research. 

The exercise also triggered client paranoia that Facebook would glom onto our lists and use them for their own purposes as a quid pro quo. Facebook denies this and claims it keeps no record of uploaded matching data. But this worry will persist. 

The big data hoopla is not much more than extension of direct marketing techniques into the online space with the uncertain promise of real-time processing and communication. Traditionally a direct marketer goes to Acxiom or Epsilon or their competitors and appends data to their lists in search of more new or “best” customers. By understanding who owns a home valued above $250K, who has diabetes, who has kids under the age of 12, who has a credit score north of 675 or who owns a gun campaign narratives and targeted offers can be constructed. Data appending is a hedging tactic to increase the likely number of responders. 

Once these variables are appended to a list, by matching either e-mail or postal addresses, marketers create models, build segmented lists by anonymously selecting names that fit the criteria or use these variables to borrow names from list co-operatives. Appending these variables to Facebook profiles would give Facebook the opportunity to create standing sales channels of Moms, sports car enthusiasts, campers or fishermen, Vera Wang buyers, gluten free eaters or asthma sufferers that could be merchandised to brands again and again. Similarly, for a premium price, Facebook could build a targeted list to suit any particular product or campaign criteria. 

Marrying big data to Facebook is a potential bonanza for Facebook’s shareholders and brand marketers. The trick will be how this is sold to Facebook users. 

You can hear Zuckerberg falling back on “make the world better” as a rationale for serving up individualized experiences and personalized ad messages. At the same time you can image how creepy or annoying it might be to be bombarded by ads while you’re trying to keep up with friends and family. A backlash against big data and the continuing commercialization of Facebook could kill the goose that laid the golden egg or the skillful application of big data could propel Facebook toward global domination. 

Stay tuned. The game is just beginning.