Following a cookie break (sorry guys, you had to be here for that), the last few case studies and panels of the day at Realtime NY 11 took place. For coverage of the morning panels and some afternoon workshops, check out Part 1, Part 2, and Part 3 of my coverage.
Case Study: The Realtime Brand from 30,000 Feet
Allison Ausband, Delta Airlines (@DeltaAssist)
According to Ausband, you wouldn’t believe that kind of tweets you get from 30,000 feet. Being on Twitter allows Delta to be with their customers wherever they are. The solutions that they offer on their customer service Twitter feed are public and transparent, and they may answer questions that people haven’t asked yet.
The @DeltaAssist account used to be 4 people who would direct questions to appropriate answers on the Delta website. As it became clear that customers were looking to save time, the team was increased to 12 members, and these individuals were empowered to answer questions directly. The social media team sits in the same “lab” as the crisis communications and marketing teams, so the process of responding to customers has been streamlined.
Case Study: Show me the Money (Part 2): Realtime Coupons, Bottom-line Conversions
Jason Harty, Pretzel Crisps (@jasonharty)
Harty discussed what he called “just-in-time social media marketing”. Despite being a “humble pretzel cracker”, Pretzel Crisps seeks to be a catalyst of online conversation. He believes that brand affinity = brand engagement. The more that people keep their product on top of mind, the more likely they will make a purchase decisions.
To engage social followers, Pretzel Crisps uses Buddy Media “sapplets” (social applets) on Facebook and “social sampling” on Twitter. Harty discussed how classic Facebook engagement slowly increased their fan count until they implemented a fan-gated coupon code on their fan page. That is, if a user was a fan of the Facebook page, they would see a coupon. After switching from a $1 off coupon to a Buy One, Get One coupon, the fan count doubled in only 36 hours.
In their social sampling practices, the Pretzel Crisps team listens for specific phrasing (“I’m hungry”, “I’m having a party”, “It’s my birthday”) and then gives users free samples of Pretzel Crisps for them and their friends. The team has performed 265 intercepts, offered 4,200 bags of product, and earned over 2.5 million impressions.
Influence: Can you Earn It, Buy It, Measure It?
Moderator: David Armano, Edelman Digital (@armando)
Gilad Lotan, SocialFlow (@gilgul)
Derek Rey, Ad.ly (@d_rey)
Duleepa “Dups” Wijayawardhana, Empire Avenue (@dups)
Kevin Winterfield, IBM (@kmwinterfield)
There’s more to influence than just numbers. Influence has become a sort of Holy Grail in the space, and in turn has become a controversial subject. Is it possible for social influence to be a game? What value does social influence bring to the conversation? Dups believes that he is the wrong person to ask: in his opinion, life is a game, and we use game mechanics in everything we do in life. Lotan thinks that quite valuable things can be gamed (e.g., page rank), but he things that a game cannot totally explain influence. Do game mechanisms explain, for example, who you trust at a specific moment?
Winterfield believes that influence should be based on expertise, while Rey defines influence by how much one can drive action. Lotan believes that influence and action must be tracked over time. He is concerned that many quantitative algorithms ignore network effects.
Winterfield believes that influence is earned through high trust and credibility. IBM measures influence in four quantitative categories which are then checked through qualitative factors:
Dups views the market factors on Empire Avenue as a qualitative identifier of influence.
The panelists agree that influence may be fleeting, and that timing is an important part of the influence equation. Rey explained that Ad.ly continues to track their influence measurements continuously so that they are highlighting influencers at appropriate times. Lotan recognizes that it’s difficult to determine over what time period one should measure influence: if the time period is too small, then trends may not be apparent, but if the time period is too large, then the scoring may be affective by outliers who statute of limitations has passed. Armano noted that even Klout is changing its measurement to recognize temporality through the introduction of +K and the 4-6 day boost it provides to an influencer.