Please email marissa@modernmediapartners.com if you have additional updates, or a social network that you feel should be on the list.

Twitter Hits 500 Million Registered Users, Pinterest Grows Rapidly in Europe

Twitter has surpassed the 500 million user mark, according to unofficial Twitter statistician Twopcharts.com.  Approximately 12.3 new accounts are being registered per second, and Twopcharts predicts that Twitter will reach the next milestone – 600 million users – in 103 days. However, that leaves two big questions: how many of these accounts are active, and how many are fake accounts?

Since May 2011, Pinterest has consistently seen growth rates in the double, triple, and quadruple digits in European countries, as reported by comScore.  In terms of unique visitors, the UK was the leading EU market with 245,000 unique visitors in January 2012.  Germany had 67,000 unique visitors in January 2012, while Spain had 62,000.

Germany took the lead as Pinterest’s fastest growing EU market with a growth rate of 2956% from May 2011 to January 2012, followed by Spain (a rise of 1348%), and Italy (794%) within that same time period.

While visitors from France and the UK decreased slightly between December 2011 and January 2012, engagement on Pinterest rose by 1740% and 20%, respectively.

That’s what Teradata wanted to demonstrate with its “Social Experience” program, which it ran during its recent PARTNERS  2011 Teradata User Group Conference & Expo on October 2-6 in San Diego.  The program, which saw a 20% participation rate from conference attendees, was so successful that it has led to several custom implementations of similar programs for Teradata customers,  and will be repeated next week at Teradata’s Aprimo Marketing Summit.  (Marketing software company Aprimo is owned by Teradata, a big data analytics company.)

Campaign Goals

According to Aprimo VP Product Marketing Deb Woods, Teradata wanted to create an experience that:

• let customers and partners connect with each other at the conference,
• introduced them to local vendors and services, and
• showcase how realtime and customer data could be combined to create tangible marketing value and deliver real returns.

How the Realtime Campaign Worked

Attendees were invited to opt-in to the Social Experience through a prototype Facebook app, the Aprimo Offer Exchange, to receive targeted offers based on their social preferences. The campaign captured registrant data, matched exclusive offers from San Diego businesses, event sponsors and exhibitors to registrants based on interests and real-time data from social channels–Twitter feeds, Facebook “likes,” and LinkedIn connections.

It delivered the offers according to registrants’ desired methods of communication (Facebook, email, text messages). There were approximately 100 offers from 40 merchants and exhibitors.

Teradata used the realtime data to target the offers based on influence, sentiment and the content that users were sharing.  The company analyzed 3,200 tweets and 6,200 Facebook posts to determine which attendees posted most often, and whose posts were forwarded or retweeted by others. Graph analysis helped visualize the path that retweeted messages followed, indicating the virality of the message and influence of its originator. Sentiment analysis tools helped the team identify positive and negative postings. Analysts also tracked tokenization to understand how users navigated through web sites and which had higher click-through rates.

Based on the analysis across both the structured and unstructed data, the program personalized the attendee experience (program agenda, session schedules, conference e-mails and surveys) and delivered targeted offers and discounts  from exhibitors and local merchants via the Aprimo Offer Exchange for exclusive offers and discounts.

For example, someone tweeting that they were headed to an early-morning session might receive a Starbucks coupon.  Someone posting about analytics-related content might receive a link to a relevant white paper.

Teradata shares more detail on the technologies used to power the program in this blog post, but the campaign ran on a mashup of proprietary Aprimo marketing automation software, a custom-built Facebook app, Teradata data warehouse analysis, and data and business analytics tools.

Campaign Results

• The Social Experience program in San Diego had 727 registrants (20% participation rate) of a total of 3,500 attendees–a 20% participation rate.
• The program sent a total of 18,932 personalized offers via SMS, email and the offer exchange Facebook mobile app.  Of those, 6,612 were viewed–which translates into a 35% open rate.
• Of approximately 100 offers, 89 were acted upon.
• And Teradata walked away with 12 leads for customers who were interested in buying the prototype application that powered the campaign, with two of those customers likely to implement a custom solution for their own client-base.

Teradata’s marketing team is especially pleased with the participation rate because many of its are not very active on social media.  The company has clients in all industries–some, such as retail, are very advanced users of social media, but others, such as its pharmaceutical industry clients, are less active adopters.  In some cases, Deb Woods told me, they found that attendees didn’t even have a Facebook account, so the Teradata onsite staff had to first help people who wanted to participate set up their Facebook.  In other cases, attendees chose not to participate because they restrict their Facebook activity strictly to personal use.

Realtime Marketing:  Best Practices Recommendations from Teradata

The Teradata Social Experience is a great case study in how marketers can combine big data, customer preference and realtime signals to deliver highly targeted offers.  Teradata has a major “Socialization of Data” initiative designed to drive conversation around this topic and encourage its customers to identify and implement programs to drive value at the intersection of structured and unstructured data.

Aprimo CMO Lisa Arthur says that “social networks aren’t simply new channels” but “gateways to volumes of insight. Customers are sending daily – if not hourly – buying signals to companies on laptops, tablets, smart phones and other devices. The Socialization of Data shows companies how to interpret and respond to those signals moving forward.”

But Teradata is far more conservative about how it is stepping into this space when compared to fast-moving realtime marketing start-ups such as LocalResponse, especially when it comes to the issue of the opt-in.

Deb Woods told me that too often, marketers make two mistakes in how they communicate with customers:  1. not getting their permission, and 2. not sending the right offer, at the right time, in the right channel.  By combining realtime data with a traditional data warehouse, an integrated solution such as this program demonstrated can address both of these issues.  ”If you don’t have a pre-existing relationship with me, and if I haven’t agreed to receive communications from you, I will feel like you’re stalking me,” she said.  Companies who don’t follow these guidelines risk brand degradation.

What do you think?  Would you be willing to experiment with a realtime marketing campaign that’s based on a soft opt-in following the guidelines outlined by LocalResponse?  Or would you restrict your campaigns only to programs such as Teradata’s, which require a strict opt-in policy?

How many of Twitter's 500 million users are fake accounts paid for by scams? Nobody knows.

Someone may be using your name and picture to create fake Twitter accounts–at 5c each.

The same day that major news outlets were announcing that Twitter had reached 500 million users, I had a surprise waiting for me in my mailbox.

“Fernando, I am following up on the post that I made to the SocialMediaMarketing group on LinkedIn requesting someone to create email and Twitter accounts for me. I am interested initially in having someone create 100 email/Twitter accounts. If all goes well, we would create more. Here is a description of what we want:  Specs for Creating Yahoo Email & Twitter Accounts – downloaded from Google Docs (PDF, originally posted as a Google Doc here)

Here is the spreadsheet into which all of the information would be recorded:  Yahoo Email & Twitter Accounts Data Entry Template (PDF, originally posted as a Google Doc here)

Please let me know if you are interested in doing the project. We are interested in paying per successfully completed account (obviously, we will verify the accounts; we wouldn’t pay for those which are not valid or are suspended). Please let us know how much you would charge. We would be excited to work with you.”

Looking at the description document I immediately realized that this person, representing SNSAnalytics,  wanted to fill Twitter with even more fake users. For what? I don’t know, but I have my suspicions.

What surprised me the most was the level of detail that was being requested:

• full names, taken from the Census.gov website,
• postal codes,
• geographical distribution,
• age group and
• the use of a “real human face” as a profile picture for the Twitter account.

The legal concerns–“No copyright pictures should be used to avoid possible legal consequences”–made me laugh. When I accessed the documents at least 10 other users were reading them and that is also worrying.

I followed up on the mail in order to know what the payment would be for creating 100 fake Yahoo accounts and 100 fake Twitter accounts. I didn’t have to wait long to receive a reply: 5c – 10c for each confirmed fake account.

What does this mean?

For a long time I had suspected that there were US-based agencies out there that resorted to these crowdturfing practices to boost campaign numbers for their clients, but I had never had the documented proof of this until now.

These practices hurt everyone in the industry: They hurt the reputation of digital agencies as a whole, they hurt all professionals in the digital space, they hurt the clients that are presented with fake results.  Above all, they hurt the users and the platform.

Five Hundred Million Twitter Users? Think Again

I am always very skeptical about the number of users that every social network claims to have, and this specific case shows beyond any doubt that Twitter’s 500 million users milestone is not real. Until there are better ways to prevent scams like this—IP monitoring when creating a new account would be a simple solution that Twitter’s could implement—we will continue to see these practices where a company may use our names and our pictures for their own shady purposes.

About the author: Fernando Fonseca is the founder of SoMeOps, an agency that works in the digital space, based in Seattle.  Fernando doesn’t have a Klout score, likes craft beer and is a Google+ enthusiast.

Through Taap.it, fans will be able to see detailed pictures and descriptions of available food items, place their order, and have items delivered to their seats.

The app also goes beyond food and beverage service, encouraging fan interaction by allowing users to enter contests, vote for player of the game, rate player performances, and “much more.” According to Andrew Calagna, Director of Event Marketing for Taap.it, the app aims “to be the first to have a truely interactive experience with the fans,”  and is designed to “revolutionize” the way fans experience sporting events.

“We are working with a few teams, like New York Islanders, to help them make their patrons’ experience more enjoyable by using our app” wrote Matthew James, Assistant Director of Public Relations for Taap.it.

Taap.it’s new features will be available to the public for the upcoming baseball season.

Taap.it currently serves as a “Craiglist” for mobile with local classified advertisements, and enables users to “sell anything, socially, locally and in realtime” from their mobile phone (The Next Web). The app has created partnerships with over 5,000 businesses in New York City since its launch in May 2011.

The ability to order food without having to wait on line at games will surely be appreciated by fans. But will they readily use the same platform to participate in more interactive, game-centric activities, like rating player performances?

LocalResponse has started policing its clients' realtime marketing campaigns.

LocalResponse monitors realtime content and allows marketers to automatically message Twitter users based on that content.  Last Wednesday, the company launched the pro version of its realtime advertising platform, but the weeks up to the launch were surrounded by controversy around campaigns that the company powered for some of its beta customers.

Most notorious among these was a Toyota Camry Super Bowl promotion in which a series of verified accounts sent unsolicited @ messages to users who were tweeting with Super Bowl-related hashtags.  According to LocalResponse co-founder and VP product Michael Muse, the company will no longer allow clients to run that type of campaign.

I met with Michael last Friday to get his take on the difference between realtime marketing and spam, and to see if he could convince me that there is a way to run a campaign based on sending automated Twitter messages–without users feeling like they were being spammed.  Michael believes that, if done correctly, realtime messaging is not only possible, but also enormously beneficial for both the marketer and the user.

LocalResponse Co-Founder Michael Muse is developing guidelines and best practices for realtime marketing campaigns

You Need a Clear Signal

The key, according to Michael, is that “you don’t want users to get a message for something that user didn’t check into.”

The company is implementing a new set of rules and developing a set of best practices for using its platform.  In order to send an @ message to a Twitter user, that user must have specifically “checked in with the brand,” either via a location-based app, with the check-in shared to Twitter, or by mentioning the brand on Twitter in a positive context. (LocalResponse has proprietary natural language processing technology that filters out tweets with negative sentiment. So someone who tweets “Can’t wait to order the Burger Sandwich at Brand X” might get a coupon for a side of fries, but someone who tweets “Brand X food is terrible” would not receive the offer.)

Muse says that their experience with the 50+ campaigns they have run so far is that, when people mention a brand in a positive way, they are delighted to know that the brand is listening, and to get a relevant response.

For marketers who want to reach consumers who do not meet these new criteria, LocalResponse is offering a separate product called Intent Retargeting, which can deliver mobile banner ads based on various realtime signals.

A Share Means I Want You to Care

Not all users are created equal, and Michael cautioned against using your own interpretation of what works for you to judge the expectations of other users. For instance, Walgreens recently ran a campaign using LocalResponse to deliver coupons to customers who checked in at its stores. The campaign drew heavy criticism from some bloggers, who described it as spam.

But under the new LocalResponse guidelines, this Walgreens campaign would still be legal. And, according to Michael, the reason the campaign worked is because it was targeting users who checked in at the pharmacy and then chose to share that information:  ”If you choose to share a check-in, then you indicate that you want someone to care about that piece of content.” Personally, I may not want to hear from Walgreens–but then I’m also not very likely to check in when I visit one of their locations or to publish that check-in to my Twitter feed.

Do I Know You?

Michael and the LocalResponse team have developed other rules that their clients must follow. The entity that responds to the user must be the verified entity that the user mentions. In other words, no tweets from Burger King to people who check in at McDonalds–only McDonalds can respond to tweets that mention McDonalds. The company also insists that clients be transparent about why they’re responding to users and offer something of value to the recipient in the tweet (“thanks for checking in with us – here’s your coupon!”). The rules provide guidelines about creating content that does not feel “too cheesy”–no hashtags or generic marketing slogans, for example.

A Little Goes a Long Way

LocalResponse sets strict guidelines around the volume of messages a user can receive:  no more than one a day from any vendor using the LocalResponse platform, and there are limits on the number or messages a user can receive from any given merchant. The landing page included with the marketer’s tweet always includes an opt-out so that users can chose not to receive any messages powered by LocalResponse, or they can opt out of messages from a given vendor.

Most brands are still learning the right way to engage with Twitter users. “We have to say no every day,” Michael told me.

Keep the People in the Picture

I wish I had a sign like this in my bathroom.

The company strongly encourages brands to have trained social media professionals monitoring campaigns and prepared to engage with users who respond or comment. The PRO Dashboard, released last week, allows marketers to see realtime analytics and to pause a campaign to make adjustments if it is not getting the anticipated response. The LocalResponse team monitors running campaigns, too, and will step in if there is a problem.

Done right, realtime marketing works. Michael says that LocalResponse customers have seen “click-through rates around 50%,” and one of the company’s office is papered with advertising insertion orders.

It’s a Hug Thing

The best thing about Twitter, says Michael, is that “you get feedback.” That means that people who hate what you’re doing are far more visible. But it also means that if you “learn to embrace these customers,” listen to what they are telling you and learn how to respond to the content they’re sharing in the right way, “they will love you.”

So what do you think–are you convinced? Should brands be using what customers tell them on Twitter and other realtime platforms as an invitation to engage using platforms like this? Or do you think brands should stick to more personalized realtime outreach, like in this Pretzel Crisps example?

… and stay tuned: tomorrow I will report on a conversation I had with a technology vendor who has very different views on the question of messaging customers who have not specifically opted in to hear from a vendor. UPDATE: that post, on Teradata’s Social Experience campaign, is now live.

The October 2011 online survey of nearly 1,000 U.S. residents revealed that 21% of employees – over 1 in 5 – are friends with their supervisor on Facebook.  Younger employees are significantly more likely to be Facebook friends with their supervisor: of those who have both a Facebook account and a direct supervisor, 26% of those ages 18-34 are friends with that boss, compared to only 10% of those ages 35 and over.

Frequent Facebook users were also more likely to be friends with their boss on the social network.  Of those who log into Facebook daily, 23% are friends with their supervisor on Facebook vs. only 12% of those who log into Facebook less than once per day.

The survey asked whether it was considered inappropriate to friend one’s supervisor on the social network.  Not surprisingly, the younger generation (18-34) and more frequent Facebook users were both less likely to feel this relationship was inappropriate (only 28% and 31% respectively), while those ages 55 and older (44%) and less frequent Facebook users (41%) were more likely to believe being Facebook friends with their supervisor was inappropriate.

Facebook friendships are initiated by either party. Among those who report being Facebook friends with their supervisor, 46% say they initiated the connection, while 38% claim the friend request came from their supervisor.

Some employees have chosen to end their connections with a supervisor on Facebook. Of those who are not currently friends with their supervisor on Facebook, nearly 6% say that they were connected at one time; this figure climbs to 10% among those ages 18-24.

Once the connection is established, it is largely used for social purposes. Of all Facebook users with an existing connection to their supervisor, 70% use it for social communication.  While some do use the social network to communicate in a professional context, it’s more often men (53%) than women (37%).

Does a Facebook connection with one’s supervisor affect job performance?  Nearly one-third of men believe that being friends with their supervisor on Facebook allows them to do their job at least somewhat more effectively; only 15% of women agree.  The vast majority of both genders believe that having a Facebook connection with their supervisor has no impact on their job performance.

Are you friends with your boss on Facebook?  Does your company encourage or discourage social network friendships between employees and supervisors?  Does it have a social media policy in place to define appropriate boundaries?

Download the full report from Russell Herder here: Making the Connection: How Facebook is Changing the Supervisory Relationship.

The online survey – from September 2011 – looked at Canadians’ preferred sources of product information:

• newspapers topped the list as a preferred source of product information for 86% of respondents
• TV was a top source for 83%
• company websites topped the list of digital sources for product info, listed by 68% of Canadians
• blogs were a top source for 29%
• Facebook for 21%
• only 15% preferred to look to Twitter for product research

However, despite a clear overall preference for traditional media sources, younger consumers were significantly more likely to turn to social media sources for product information:

• Almost 4 in 10 Canadians (38%) aged 18-34 listed blogs as a top research source when purchasing a product or service, compared to less than half that (16%) of Canadians aged 55 or older
• More than 1 in 4 Canadians (27%) under 34 years of age reported YouTube as one of their top research sources vs. only 15% of the boomer generation (adults over 55)’

Perhaps even more important, the younger generation was twice as likely to trust social media and company web sites as “credible news sources”:

• 18-34 year old Canadians were nearly twice as likely as those 35-54 to list social media sources (such as Facebook) as credible news sources (22% vs. 12%)
• 18-34 year-olds were also twice as likely to trust company websites as credible news sources vs. boomer Canadians (23% vs. 10%)

According to Carol Levine, chair of the CCPRF, “A significant portion of our younger generation sees blogs, YouTube, Facebook and company websites as credible sources of news. This suggests to us, that in their minds – and in contrast to older Canadians – the boundaries of credibility between news, “circle of trust” conversations and marketing are blurring.”

Do you agree?

Please email marissa@modernmediapartners.com if you have additional updates, or a social network that you feel should be on the list.

Google+ Still Male-Dominated, Popular with Students

An infographic recently released by Website-Monitoring.com looked at Google+ users by country, occupation, gender and relationship status.  The results?  Google+ is still a male-dominated social network (67% of users are male), but it is also home to a significant student population.

The most popular occupations for Google+ users:

• student (20%) – 1 in 5 Google+ users are currently students
• software engineer (2.65%)
• consultant (1.99%)
• manager (1.72%)
• photographer (1.59%)
• teacher (1.39%)

Google+ is still most popular in the US, which claims 31% of the social network’s user base.  India ranks 2nd, with 14% of Google+ users.  However, Bangalore edged out New York to be the most popular city in terms of Google+ users (3.86% vs. 3.63%). The other non-US cities in the top ten were London (3.34% of Google+ users), Hyderabad (2.97%), Moscow (1.71%) and Calcutta (1.6%).

Now that Google+ recently opened up to teens, it will be interesting to see whether the percentage of students grows even larger on Google’s social network.

The number of microblog users in China increased a whopping 296% in 2011, according to the report.  Social network user growth was even slower than the growth of general internet users, showing a decline in total penetration for social networks in China (eMarketer).

But according to eMarketer, these growth levels are unlikely to continue. “Regulators are rolling out real-name registration requirements to tighten control over popular social media services, which will likely deter new (and multiple) signups.”

How will social networking and microblogging growth continue to play out in countries where social media platforms are heavily regulated by the government?

NPR recently decided to try a new social media tactic to drive traffic to one of its local stations: customizing Facebook posts by location.  In a case study recently covered on NPR’s Digital Services blog, NPR’s 2.3 million Facebook fans became part of this geotargeted experiment.  Back in October 2011, NPR’s Facebook page began sharing relevant stories from member station KPLU in Seattle.  These posts — visible only to fans in the Seattle area — led to record traffic on KPLU’s website.

How did NPR weave the KPLU content into their regular news posts on Facebook, and how did Facebook fans interact with these targeted local stories?  The experiment is still ongoing, but NPR’s Eric Athas and Keith Hopper decided to take a look at how the targeting worked, and what effects it had on KPLU’s audience numbers and engagement after the first four months.

How did NPR geotarget posts effectively for their Facebook audience?

NPR has Facebook fans in thousands of cities, but decided to experiment with customizing Facebook posts for just one city.  The news media outlet worked closely with Seattle’s KPLU on a daily basis to choose the right story to post to NPR’s Facebook fans in the Seattle area.  Each story had to pass “the content coffee shop test:” meaning the content had to “raise curiosity” and have the right “splash of local flavor” that would “be talked about in a Seattle coffee shop.”

Seattle-based NPR fans on Facebook might have noticed an increased volume of Seattle-oriented stories, with links to KPLU.org, while NPR fans living outside of Seattle would never see these posts.  However, the number of stories increased only slightly – up to one per day – in an effort to maintain the “steady cadence” of regular posts from NPR, generally one per hour throughout the day.

How did the NPR posts affect KPLU’s audience numbers and engagement?

NPR chose to measure audience growth from the experiment through campaign tracking, by attaching a unique tag onto the end of each KPLU link posted to NPR’s Facebook page.  In the first four months, NPR’s Facebook page posted ~50 geotargeted links to KPLU.org (just a small piece of all KPLU content), and those posts accounted for 12% of all visits to KPLU’s site over that time period.  While NPR expected the targeted posts create a temporary rise in KPLU traffic, the level of response was more impressive than anticipated — KPLU traffic hit three milestones within that four month period:

• record traffic for a single day (January 19)
• second-highest traffic for a single month (October 2011)
• highest traffic for a single month (January 2012)

User engagement – how NPR Facebook fans reacted to these local stories vs. globally shared posts – was measured using Facebook Insights.  While the “flurry” of likes, shares and comments from NPR Facebook users following a post is obviously greater for global stories (seen by a much larger audience), in this case NPR looked at each post’s engagement rate, or the “number of likes, shares and comments on a Facebook post as a percentage of the number of unique people who viewed it.”

The results? NPR’s team found that “geofocused posts to the Seattle region usually had a much higher engagement rate than links shared to the global NPR Facebook audience.”  More Facebook fans (relative to the total audience viewing each post) were engaging with the local content.

Athas and Hopper believe that so far, “no other news organization has used Facebook to geofocus content in quite this way.”  The careful selection (and discreet volume) of “conversation-style” local posts led to a significant boost in traffic and – more surprisingly – to consistently higher engagement rates than global posts.  Will other major news organizations start following suit?