CrapHammer

Rethinking Digital and Social Marketing for the Arts

Or the five hardest questions I've ever been asked.

Rethinking Digital and Social Marketing for the Arts

The presentation that I gave today to the International Society for the Performing Arts.

Measuring Twitter - You Gotta Be Kidding Me

Here's the scenario.  You launch a twitter account around a product launch.

It takes off like wildfire. But how do you go about showcasing the reach and re-tweet activity?  Laughable, right?   There are so many great tools out there.  How hard can this be?

Boy was I surprised.  Not that there aren't an insane number of tools.  Just that none of them appear to be 100% up to this task. So I set out on a quest to answer this question.

My Task:  Be able to showcase unique tweet and retweet activity (engagement) by day, week and month as well as the ability to tally up final reach.  So I want to know the WHO, WHEN and the total reach.

Here is the final tally.

The Fails = 7
The Partial Wins = 1
The Wins = 0

(Click to enlarge all screenshots below.)

#TheFails

RetweetRank (link) Score: F

I have no idea what scenario would enable this tool to be helpful or effective.  Though it's so awesome to know that my twitter account is in the 97.87 percentile.  Huh?  But hey, at least it's free.  Unlike the next one.

Tweetreach (link) Score = C-

Okay.  This gives us a total reach number.  And this was part of the ask.  But where did this number come from?  Over what time period?  What content was retweeted over the others?  What percentage of users participated versus not participating?  It's possible that the pro account would provide this but I'm hesitant to try it as the pricing escalates pretty quickly into the hundreds a month.

TweetStats (link) Score = C+

I feel really bad placing this service in the #fail category.  This tool really impressed me.  It's well done, free and an interesting way to explore your twitter activity in a profile.  I was able to see who I reply to more than others, when I tweet more during the week and even explore the clouds of my tweet content.  But I was unable to pull up any evaluations on what tweets were most retweeted or what my total reach might be in a time period.

TweetEffect (link) Score = F

WARNING!  Avoid this service like the plague.  It tries to show you what tweets you made that led to people following or not following you.  The graph at the top is so lacking in any insight that I literally laughted out loud when it loaded.  Again, looking at the list of what tweets "caused" people to leave may cause the stats addicts out there (like me) to lose way too much sleep.  Avoid at all costs!

Twitalyzer (link) Score = C [updated]

At this point I think I've given 30 tools access to my Twitter account and the design of this site made me really question whether I should grant access once again.  But I soldier on!  This service touts hundreds of reports. Um.  Really?  I'd say there are a few dozen reports.  Sadly, not one of them could give me the data I was searching for.  Their retweeters tab just said "Unfortunately we don't seem to know any of @passitalong's retweeters! Believe it or not we don't know every person who uses Twitter."  Sigh.  Fail.

UPDATE:  Per Jeff's comment at the bottom of this post, I went through each of the reports he outlines.  I have to say that some of the data points we would be looking for are here.  But it always remains a graph with the inability to drill down and see the data points.  And we would have to run a high number of reports to then pull each number into another spreadsheet (again, with the actual mentions being counted not being visible in the report we're looking at).  As many of the numbers are there, however, I updated the scoring for Twitalyzer to a C.

Twinfluence (link) Score = F

Someone needs to help all these services come up with a name that is not a variation of Tweet or Twitter.  That aside, this one was so bad it almost became my favorite.  I will have you all know I am "ranked in the of all twitterers!"  What?  Beat that!!  I think they left out a percentile number and we know how valuable that measure is!  oy.  Fail.  Next.

TwitterCounter [link] Score = B-

Okay.  TwitterCounter deserves an honorable mention.  Their solution is clean, effective and fast to repsond.  More importantly, they are proposing a paid service that will provide what I am looking for.  See image below.  But this remains vaporware at present.  It does not currenlty exist.

The one thing potentially missing from the paid service is the ability to identify unique retweets/twitters.  But if they manage to roll out this new service, I would move them up into partial win or higher.

#ThePartialWins

Radian6 [link] Score=B+

(Note:  Fake data in above image, but this shows the types of reports I used)

Radian6 remains one of my favorite tools.  I can see activity, trend it over time and even compute total reach achieved.  But what I can't easily do is pull up everything associated with a single twitter account.  I have to do a keyword search.  So what I did was do a search for anyone mentioning my twitter account.  But if your account is not, let's say, a unique word, then you're in for a lot of fiddling. 

I can't limit it to just looking at one source, because I want to see all mentions across the twitterverse, etc.

That said, I was able to pull unique mentions.  And I was able to pull a total reach number.  But I couldn't easily pull unique mentions by day, week or month.  This means I would have to run every report for every day and then tally it in Excel.  Not exactly fun.

Also, while I could pull unique mentions (choose : Unique Source Count as shown in image above) I can't easily tell what was a retweet or what was a mention without compounding significantly the number of reports I would have to run.

#TheWins

I got nothing.

But I know what I want.  I want PostRank for Twitter.  But more on PostRank later.

Know of a killer tool that would have made our life easier?  Ping me or comment below.

Slicing Radian 6 for Key Properties

I'm often asked for a list of key properties or spaces where conversations are occurring.  I'm trying to avoid using the word "influencer" here as anyone who reads my blog likely knows my views of such.

But let's face it.  A week doesn't go by in our industry that we aren't asked to find "influencers."

Here is a common request.

"Find the mommy bloggers that are best aligned to our brand or a new campaign we are running."

I was recently chatting with the awesome folks from Radian 6 and thought I would run this kind of request past them as I had traditionally wasted a lot of time trying to answer it.  I got the typical response.  Search for phrases most likely to identify these types of people/sites and then use the tools to narrow down to a set of key places.

Note to anyone trying this at home.  It won't work, or at least not well.  Which made me realize that I should likely share with the world how I now go about answering these types of requests. (Note:  Using Radian 6.)

Let's assume that the brand/campaign is a new type of cell phone.

The Wrong Way

We've been asked to find Mommy Bloggers so the first approach is generally to think of what types of conversations Mommy Bloggers are having that might be relevant to this campaign.

How young is too young to get a cellphone?

Impossible to keep in touch with my kids.

My phone always loses its signal at soccer practice.

Then we start to wheedle down to a list of phrases, includes and excludes that would narrow down to the conversations of interest.  This takes quite a bit of time configuring and reconfiguring your search parameters and slicing and dicing the data.  And we are faced with an immediate dilemma.  How do we specify that the conversation is taking place on a Mommy Blog?

Social Media Monitoring tools (at least the ones I work with) focus on the actual conversations (blog posts, forum entries, comments, tweets) not the source.  So we need to come up with phrases that only Mommy Bloggers would say when speaking about things pertinent to our campaign or brand.

Hopefully you are saying to yourself, "What??!!"  What would a Mommy Blogger say on her blog that wouldn't also appear on a parenting website?  Well, as a Mommy Blog IS a parenting website, the answer is "nothing".  So this simple sounding task quickly becomes a prolonged exercise in frustration. 

Note that it can be done and you will find results, but it can take you days.  And frankly, it's faster (if less comprehensive) to manually visit each of the top Mommy Blogs you know of and just do a search using their sites' search feature for conversations you are fishing for.

A Better Way

When the request (find the mommy bloggers that...) assumes that we are looking for the hottest and most relevant nodes of conversation within a "defined" slice of the Internet, then it is quite straight forward to just limit our listening tools  to that slice of the Internet. 

For example, we are looking for Mommy Bloggers in the opening objective.  So, let's limit our tool to only be sniffing the Mommy Blogs.  Simple and straight forward, no?  All great ideas are.  Or so I tell myself.  ;)

In the screenshot below, I'm loading in a list of all the Canadian Mommy Blogs we know about. I'm limiting Radian 6 to look only at the conversations from a defined list of sources that match the keywords we are looking for.

So now I can search for the conversations likely to be of value to our client and not worry about whether the conversation is happening on a mommy blog.

Takeaways

I have spent days configuring the first approach and would consider myself lucky to have 1 mommy blog in my top 10 influencer view.

Or I can spend 30 minutes on my "Better way" approach and find 10 hot mommy blogs of interest immediately in the top 10 influencer view.  And even better, I can remove a lot of bias by looking at trending phrases, top conversations, etc. versus just searching for the conversations I want to find.

Photo Credit: hownowdesign

What Spammers Know That We Don't

We are all working like mad to develop location based services and technologies. Phones that know where the nearest cool coffee shop is, discount coupons that apply to a store you just snubbed and walked past.  I believe it is considered to be a multi-billion dollar opportunity and the salvation of the mobile industry.

But frankly, we are sucking eggs and doing next to nothing while a few key individuals are proving you can mine for location based gold today with free or near-free resources.

Spammers Innovating Faster than Hi Tech:

The Real Estate, Online Pharmacies and Porn Spammers have once again proven they are leaders in adapting new technologies to their annoying purposes.

I tweeted the following:

And within minutes I received a follow from KatyFStewart.

She offers one link on her page.  To this site:

 

Should I be promoting her Twitter account?  Should I have clicked through on a link from a semi-naked nubile photo?  No.  Of course not.  But will I be alone?  Nope.  Guaranteed.  Not only that, I would bet that many a media person would cry to get even 1/10 of the clickthrough and conversion rates these people are achieving.

Not Just Porn:

Tweet the name of a city and you will undoubtedly get a sudden increase in followers.  Why?  Because the real estate industry is not about to be left behind by the porn industry.  Quite a few real estate agents appear to be using auto-subscribe services to add you as a friend in cities they do their work in.

These are services that listen for a set of keywords and once it finds a match with any twitter user in the world, it sends that person a follow request on your behalf.

Let me be clear. I hate auto-subscribe services.  It's like paying someone to get you more friends.

That said, it does make me wonder.  Do we really need fancy and expensive GPS and high tech solutions or could we just better monitor and respond to the contextual content people are providing in their tweets and status updates?

Brilliance:

What I love about [what] Ms. Stewart is doing (real name likely Amanda Hogsbottom) is that they are using simple, free and contextual tools to mine twitter and facebook.  Then they are doing simple (and free) IP sniffing from their website to provide an end to end contextually relevant and highly personalized (location based) solution.  Something the big brands are spending hundreds of millions of dollars to achieve with likely far less success.

A System for Measuring Change through Stories

I stumbled across a reference to the Most Significant Change process a week or so ago.  For the life of me, I can't recall exactly where.  A search on Google brought up a PDF that presents both the concept and the process for implementing MSC. It is designed as a participatory monitoring tool for social change campaigns that have complex goals, involve organizations with many layers and where conventional monitoring systems are proving less than effective at furthering the mission.

LINK to PDF.

What most impressed me about MSC was it's focus on stories of significant change and a system for evolving and growing the values of an organization around such.  The stories that most resonate with the organization are the same stories that slowly transform all levels of the organization.  A wonderful use of social phenomenon to shape and grow an organization (as opposed to market norms).

1. It is a good means of identifying unexpected changes.

2. It is a good way to clearly identify the values that prevail in an
organisation and to have a practical discussion about which of
those values are the most important. This happens when people
think through and discuss which of the SCs is the most significant.
This can happen at all levels of the organisation.

3. It is a participatory form of monitoring that requires no special
professional skills. Compared to other monitoring approaches, it is
easy to communicate across cultures. There is no need to explain
what an indicator is. Everyone can tell stories about events they
think were important.

4. It encourages analysis as well as data collection because people
have to explain why they believe one change is more important
than another.

5. It can build staff capacity in analysing data and conceptualising
impact.

6. It can deliver a rich picture of what is happening, rather than
an overly simplified picture where organisational, social and
economic developments are reduced to a single number.

7. It can be used to monitor and evaluate bottom-up initiatives that
do not have predefined outcomes against which to evaluate.

Again, here is the link to the PDF for those interested in learning more.