Yesterday, Wikibrands held it’s first themed “Tweet Chat” on the subject of Measurement and Analytics for social media. The conversation moved quickly, but there were some excellent insights that came out of the discussion. Below I have pulled the highlights from the chat and some key insights from leading experts in the field. Thank you to everyone who participated. The full list of questions can be found here.
1. The Zeitgeist-What have been the biggest trends in social media analytics over the last year?
@SeanMoffitt Trends – acquisitions of little guys by big guys, the impact of mobile, data visualization, video measurement, others? #wikichat -
@kamichat @wikibrands The biggest trend is that people aren’t as interested in number of friends as they are in the number of biz results #wikichat
@AlterianJames Trends-understanding monitoring can deliver valuable insight & is not just customer service responding to issues & fails #wikichat
@HowieSJ @seanmoffitt segmentation / opt-in is an interesting trend for #wikichat – many apps I c r about user choice via opt-in & segmentation
@AlterianJames Trends-more great examples and case studies that prove the ROI of social media – in both B2B & B2C #wikichat
2. The Needs - What are most clients looking for in the way of information from you? What should they be asking that you’re not hearing enough of?
@TaraCoomans clients are mostly looking for a direct tie of social media to sales #wikichat -
@WalkerLucas Clients should be looking to get a more accurate sample size opposed to seeing everything #wikichat
TaraCoomans I wish clients took a longer view of social media metrics and better understood its long-term opportunities #wikichat
@Lkinoshita [agree!] @TaraCoomans clients are mostly looking for a direct tie of social media to sales #wikichat
@socialdave Clients are also looking to understand how social media influences traditional media. Identifying the influencers is key. #wikichat
@colinlewissmith a lot are interested in BASIC base lining as a starting point. Analytics to follow once some initial learning has happened #wikichat
3. The Right Stuff – we came across 46 different core metrics in Wikibrands on how the best companies measured their performance with customers on the web, what are the best ones? What are the most frequent metrics used? Do you have a magic formula (and dare to share..) or philosophy?
@WalkerLucas There are a high number of variables that have large amounts of variance when measuring performance on the web #wikichat
@TaraCoomans @wikibrands metrics are somewhat dependent on the goals identified. Agreed? #wikichat
@SeanMoffitt In our book, we described about 46 different things the top companies measured- i’m really interested in what is being overlooked? #wikichat
@WalkerLucas Best practices are to isolate the variables that have the greatest impact on performance for your goals #wikichat
@SeanMoffitt @taracoomans I see the trail now – metrics tied to goals tied to function that runs the measurement operations – yes/no? #wikichat
@dragosilinca Metrics depend on goals, so can’t really generalize. I think generalization does more harm than good #wikichat
@kamichat @wikibrands Magic formula: Just because something can be measured doesn’t mean it should. Be intentional #wikichat
@Lkinoshita @wikibrands visitor-to- lead conversion rate is most important metric, if you can boost that you can drive revenue to bottom line. #wikichat
@socialdave Measurement: goal should be to narrow down to the metrics that matter to you and your campaigns. 46 may be overwhelming. #wikichat #metrics
@TaraCoomans Agree RT @doverd4s: Important ROI metrics incl rate of innovation & reduction of inventory/discounts b/c of listening to clients #wikichat
@AlterianJames Benchmarking by looking at the historical data can let you see where you have come from & where to go #wikichat
@WalkerLucas It all comes back to sample sizes. Do you need to poll everyone to see who will win an election or can you take a sample? #wikichat
@dragosilinca @wikibrands the shortage is more of goals and commitment. If the goal is important, resources will be dispatched #wikichat
@dragosilinca Measurement Usability starts with defining the metrics you need, and ignoring the rest. See how everything starts to fit together? #wikichat
@SeanMoffitt Shouldn’t there be a measurement school that business people can go to ask the right questions first? @cnwgroup @alterianjames #wikichat
@dragosilinca @SeanMoffitt there are starting to be courses and programs for execs, check this out http://rotmancrm.com/ #wikichat
4. Usability “Is there anybody out there?” – There is a well-founded feeling out there that it’s NOT a shortage of data that companies have access to, it’s a shortage of people that can use it quickly and effectively to deliver business insight? True/not true? Ways to overcome? Examples of success? Better centralized group or decentralized owners in an enterprise?
@AlterianJames @wikibrands Combination of tool AND skilled user will deliver best results. Unless the vendor offers in-house insight services… #wikichat
@TaraCoomans I think for most companies defining which metrics are important then having right people to implement are biggest challenge #wikichat
5. Witch Doctory - What is the biggest myth about social analytics that should be debunked and exploded before somebody gets hurt?
@WalkerLucas By far the biggest myth is that there is no standard deviation. Nothing will ever be 100%. #wikichat -
@dragosilinca @wikibrands biggest myths are that social can solve any problem or replace any other marketing channel. #wikichat
@WalkerLucas @SeanMoffitt You have to work with what you have, even if a few people are speaking about your brand, how many people do they rep? #wikichat
@kennynorton I would say the biggest myth.. would be the work it takes. Alot of managers/organizations underestimate the time commitment #wikichat
@socialdave Biggest Myth is that Social Media now outweighs Traditional Media when if comes to influence. It’s but part of the equation #wikichat
@SeanMoffitt One big myth I’d love to pop-measuring share of category – sure it’s what we used to do, but its like measuring the fastest turtle #wikichat
@WalkerLucas The biggest myth is the # of variables. There are many that matter, more that don’t and the ones that do are different every time #wikichat
@WalkerLucas The other big myth is that you can jump right into SM never having used it before. I’ve been on FB for 5 years, Twitter for 2.5 #wikichat
@kcheevers RT @WalkerLucas:myth is that you can jump right into SM never having used it: True, but have to get clients to start somewhere. #wikichat
@kennynorton Another huge myth is that it’s Quantity over Quality.. Quality/Brand ambassadors win every time. #Wikichat
6. The rightful owners to the measurement throne? What functional areas are best in dealing with monitoring/measuring the social web?
@dragosilinca I think the owner should be the CMO #wikichat
@kennynorton @dragosilinca so both roles? Do you think they should be the Community Manager as well? #WikiChat
@SeanMoffitt Agree the CMO should own (do in 70% of cos. we studied),but man are they poorly informed/experienced @kamichat @dragosilinca #wikichat
@dragosilinca @kennynorton I think the CMO should own it, as social media is more than community management, it’s customer service, PR, etc. #wikichat
@WalkerLucas The CMO should only own it if they know the space. If they’ve never tweeted then they don’t know what they’re talking about #wikichat
@kennynorton @dragosilinca I agree 100%. They need to adapt, still hard for some to accept. #OldMoney #Wikichat
@socialdave @wikibrands Ownership of socia media measurement shouldn’t fall with any one group. Marketing, PR, IR each have a stake in this. #wikichat
@40deuce @dragosilinca @kennynorton I do all those things… I just don’t have the fancy title #wikichat -
@dragosilinca @socialdave sure all have a stake, but you know what they say about design by committee #wikichat
@SeanMoffitt @socialdave i’ll be provocative, if everybody was in charge of social media, isn’t that a utopian world, it would never get done #wikichat
@socialdave @dragosilinca @KennyNorton Which is why I strongly believe that empowering your organization is the best approach #wikichat
@johnlovett @wikibrands analytics should belong to the CFO. Just like financial data, good metrics are unequivocal and drive biz decisions #wikichat
@SeanMoffitt @johnlovett good provocation – but if you gave metrics to the CFO – wouldn’t they make it into a soulless muckup? #wikichat
7. Standards and Benchmarking – we’ve seen various companies suggest a Facebook like is $1.35 and a Twitter follow is $0.75. Is this a good thing that we’re trying to find the quantitative, tracakable proof? Are we getting to a point of maturation where the soicial web can be…gulp…predictable? where we can determine success one brand vs. another? one marketing spend vs. another?
@Lkinoshita @wikibrands suggesting Facebook ‘like’ is worth $1.35 is just selling snake oil. The gullible will eat it up, but worthless IMO #wikichat
@ennynorton @amberturnau I’ve seen more Facebook and Twitter mentions in Traditional Media in the last 6 months, than in my entire life. #Wikichat
@SeanMoffitt Don’t care if a Facebook like is worth $1.35, particularly when 75% of that audience never comes back again, or am I wrong? #wikichat
@MyPoKos #wikichat data shows FB twice as expensive and half as effective as online banner ad. and how many FB mobile d/l-s use App every day? -
@amberturnau I wonder if Facebook/Twitter call to action in trad. mktg is here to stay or simply a fad. Thoughts? @KennyNorton @seanmoffitt #wikichat
8. Sentiment-What do you think about sentiment analysis – can we access sentiment efficiently? How accurate can it be? How much can be automated? Will this always be an intensive human input thing? Are we as businesses lazy in wanting sentiment served up to us as an algorithm?
@AlterianJames Sentiment – Let automated sentiment identify trends in your data, to explore further. #wikichat
@dragosilinca @wikibrands businesses aren’t lazy they just need to accept that automated sentiment won’t be 100% accurate and use as trends #wikichat
@AlterianJames Sentiment – A horror movie needs “scary” “sick” “frightening” vocab to do what it is supposed to. #wikichat
@SeanMoffitt One stat I read re: automated sentiment/nat. language processing was 60% accurate- great in baseball, great in social measurement? #wikichat
@WalkerLucas Sysomos has been documented by a 3rd party as being 84%-87% accurate in measuring sentiment. #wikichat
SeanMoffitt @kennynorton 87% on sentiment – that’s not bad, now we’re talking – how do you deal with language piece? #wikichat
@socialdave Sentiment analysis: any automation should be closely monitored, whatever tools being using. Sentiment may vary depending on topic. #wikichat
@BethHarte @wikibrands The problem with sentiment is that it always falls to PR people who are tasked with getting ink. They cancel out. #wikichat
9. Influence – What do you think about influencer rankings (i.e. Klout) – are influencers more important than the crowd? Have we been able to identify them well? Is their a blind spot we’re missing when we track influence? In a real life business environment – more efficient/effective to target your top 2% of stakeholders/potential fans or connect to the 100% crowd? Cues to find them in your own database?
BethHarte @wikibrands Influencers…my favorite subject. I have some thoughts on that for sure. #wikichat
@SeanMoffitt Influencers? – I’ll stake in the ground here,may be undemocratic, if I had to choose one vs. the other – I would choose the Is #wikichat
@BethHarte @wikibrands It’s my job to be influential. If I can’t influence someone it doesn’t matter if they are “influential/popular” #wikichat
@NadforGenGo: @bethharte My issue with influencers is when they are identified and lose focus almost like being famous. I don’t care what u ate for lunch.
@BethHarte @NadforGenGo Exactly. Typically so-called influencers (i.e. A-listers) don’t care about you. People who care will help & influence#wikichat
@NadforGenGo: @bethharte The question then is, how does a company identify true “influencers” and not popular people? How do you draw that line? #wikichat
@BethHarte @NadforGenGo You can’t go after popular. You need to recognize who has true influence by getting to know people. It’s WORK! #wikichat
@BethHarte: @NadforGenGo The more important question is “How am *I* influential?” Do I have a trusting relationship with the brand evangelist?#wikichat
@BethHarte @Lkinoshita @NadforGenGo Really believe that you can’t be an influencer without being popular? Read this: http://bit.ly/dYMitA#wikichat
@Lkinoshita @NadforGenGo True influencers are the first leaders. People perceive their behavior to be the correct behavior, then mimic #wikichat
@Lkinoshita Clues you’re an influencer: people adopt your language, terms. You get multiple layers of pass-along. Ideas stand test of time. #wikichat -4:32 PM Apr 5th, 2011
10. In the year 3000… What is the Future of Social Analytics? What are we not thinking about now – that we’ll all be thinking about a year from now? We’re big fans of infographics and visualization of data and how a picture is 1,000 words – any future here? Blue sky – what does this all look like in 2015?
@AlterianJames Year 3000: Hoverboards, Tools will differentiate by niche. Social insight will shape business functions #wikichat #wikichat
BethHarte @wikibrands I am hopeful that we’ll start to measure true customer relationships. Not NPV or CLV, but get a pulse on true feelings #wikichat
There were definitely a couple of clear take aways from the discussion and some subjects which I think were overlooked but worth exploring in more depth such as matching up the analytics with ROI, and potentially a more in depth look at the mobile market.
As a side note though, I hope a programing guru out there is trying to develop a more stream lined way to view the discussion post Tweet Chat as one can only capture a snap shot of the discussion and truly make use of the chaos. For a full look at the chat please click the link enclosed here.
Category: Events, Wikibrands Insights