[Featuring a Guest Appearance from Reed Smith – Read on!]
Yesterday I published some twitter analytics related to the Texas Hospital Association, in recognition of their 2012 Health Care Social Media Summit (#HCSMTX for short) that’s kicking off tomorrow here in Austin. There are a couple of reasons I did so.
- I wanted to draw attention to what the THA is doing with their healthcare social media summit
- I wanted to inspire healthcare leaders and communicators about what’s possible in terms of mining social data for insights.
Yesterday’s analysis (Texas Hospitals – Getting Serious About Social Media) was of the most simple kind:
- Which accounts have the most followers?
- Which accounts, since they were created, average the most tweets per day?
But it isn’t really those lists that are interesting to me. What’s interesting is how they came to be.
You might have noticed from yesterday’s post that there are 661 facilities listed on THA’s Texas Hospital Directory. One way to collect their “digital footprint map” (i.e., their web site, twitter account, youtube channel, etc) would be to go to each hospital’s page and search around to see if they have links to those channels. Another way would be to go to each social channel (youtube, twitter, Facebook, etc) and search for the name of the hospital. Using that method, a skilled internet researcher could probably build a list in 40-50 hours.
The really exciting thing about this is that we were able to collect the same data in about 2 hours. Our internal R&D team (in this case, Andy Boothe) has built a suite of tools that allowed us to:
- Automatically strip the URLs linked to the THA directory automatically and dump them into an excel spreadsheet
- Scan each of those home pages for any links to twitter.com, Facebook.com, youtube.com, slideshare.net, etc and rip them all down into an excel spreadsheet
- Take those twitter handles and pull down core information from twitter: Name, Bio, Location, # of Followers, # of Tweets, etc.
To do that manually might be worth it. But when you can give yourself a data source like that in 2 hours? That’s a no-brainer. Is the data set perfect? No. If any of those 661 hospitals has a twitter account that isn’t linked to their home page, our link-stripper didn’t find it. That’s why St. Davids‘ main twitter account – @stdavidshc – didn’t make the list (Note: I’ve manually added them to today’s analysis since it was brought to my attention). If we were doing this work for a client who was using these data-driven insights to drive key business decisions, we would have invested additional time to be sure that the data set was 100% clean. But for a demonstration project? It was “close enough for rock & roll.”
I put in an additional two hours last night to take the analysis a level deeper. Rather than just looking into the bio and statistics section of my Texas Hospitals twitter accounts, I pulled in their tweets. All of them. Ever. [In case you were wondering, the 65 accounts I’m tracking have issued a total of 42,182 tweets since their accounts were opened. Don’t you just love meaningless statistics?]
In order to produce a more manageable data set (remember, I’m on a two-hour time limit here!), I decided just to look at their tweets from the last month – from July 7 through August 6th. So without further ado:
Analysis of Texas Hospital Tweets collected from July 7 through August 6 – Highlights
Let’s start with the obvious: The number of tweets for the month There were a total of 1,736 tweets sent by 56 hospitals. Their frequency breaks down like this:
[15 hospitals had less than 10 tweets; 16 issued between 11-25 tweets, etc.]
As you can see, the spread is pretty wide here … the average number of tweets for the month was 31 (just about one per day), but the majority of the group were less active (in fact, 9 accounts we’re tracking did not tweet at all during the month in question).
The top 10 most active accounts for the month were:
- @texashealth – 138 Tweets
- @VeteransHealth* – 105
- @BaylorHealth – 95
- @CTVHCSMenu – 92
- @MethodistHosp – 88
- @DentonRegional – 76
- @LCMedicalCenter – 75
- @NorthHillsHosp – 67
- @BuildUrBounce – 64
- @MedCArlington – 63
- @TheMedCtrPlano – 61
The number and regularity with which an account tweets is much more interesting and important than its number of followers. But perhaps what’s even more important is how well the account-manager knows his or her audience – and how often that audience is acknowledged through the account. We have a couple of simple measurements that point in that direction: Use of relevant hashtags (to make it easy for their audience to find their tweets) and the use of @ mentions (meaning that they are interacting directly and publicly with their audience measurements and other important accounts. Here’s how our Texas hospitals did by those standards last month:
On average, the accounts in this group used hashtags in 45% of their tweets – which feels about right to me. Any less than that and you’re probably not doing everything you can to understand and connect with your audience. More is fine – as long as you’re staying relevant and not posting content that doesn’t have anything to do with the hashtag in question. Here are the accounts that used hashtags most frequently:
- @NorthTexasVA – 88.89% of tweets contained at least one hashtag
- @MedCArlington – 74.60%
- @VeteransHealth – 74.29%
- @TheMedCtrPlano – 73.77%
- @PlazaMedical – 73.68%
- @mclewisville – 72.97%
- @NorthHillsHosp – 70.15%
- @LCMedicalCenter – 68.00%
- @SanAntonioVAMC – 64.52%
- @DentonRegional – 64.47%
- @uthsct – 61.54%
@Mentions are great – because it means that the account is either interacting directly with, or calling attention to, a member of their audience or another important, relevant entity whom they feel their audience should know. In this author’s opinion, the number one factor in running a successful corporate twitter account is how successful the account manager is in engaging with others online. If you’re not doing so with your twitter account, you’re a clanging gong or a sounding cymbal – not someone that many people will be interested in talking to and getting to know. For the last month, the average for the group was that 31% of tweets contained at least one @mention. Here are the accounts that used @mentions most frequently in their tweets:
- @dellchildrens – 88% of tweets contained at least one @ mention
- @MethodistHosp – 83%
- @shrinershosp – 77%
- @setonfamily – 76%
- @StDavidsHC – 75%
- @NorthCypressMC – 67%
- @wecareforwomen – 67%
- @conroehealth – 60%
- @kindredhealth – 58%
- @valleymedcenter – 58%
- @riohealth – 52%
- @BaptistHealthSA – 50%
Another really interesting thing to look at in a tweet is what our Texas hospitals are linking to. Because twitter is an extreme version of short-form communication, it’s frequently used to transmit links to longer-form content. Our hospital accounts are no exception – 77% of the 1,736 tweets send this month contained a link. Not surprising. It’s also not surprising that a pretty significant percentage of those links led right back to their very own properties. It is pretty interesting to see, though, who else they linked to. Below are the most often-linked-to sites by Texas hospitals:
- facebook.com – 133 Links from 56 Texas Hospitals twitter accounts
- youtube.com – 60
- pinterest.com – 37
- wordpress.com – 24
- adam.com – 22
- twitter.com – 16
- cdc.gov – 12
- cnn.com – 11
- hcatodayblog.com – 10
- healthfinder.gov – 10
- yfrog.com – 8
- eventbrite.com – 7
- constantcontact.com – 6
- dailybloom.net – 6
- itriagehealth.com – 6
- newsradioklbj.com – 6
- nytimes.com – 6
- dallasnews.com – 5
- heart.org – 5
- msn.com – 5
- npr.org – 5
- tylerpaper.com – 5
Why should we care what Texas Hospitals are linking to? Because the link is the king of social currency. It’s what one person or organization passes to another that tells you what they value most. Who they trust most. Who produces the content, engagement or experience that they find most valuable.
I had an opportunity to talk to my friend Reed Smith this morning. Reed was one of the folks who helped, early on, to make the Texas Hospital Association Health Care Social Media Summit come to pass. We talked about the statistics that I was pulling, and Reed made a really important observation about them. I want to share it, verbatim, with you here:
“Many hospitals (not necessarily any mentioned here) still have fundamental gaps. What I mean by that is a lot of hospitals setup a Twitter account because of someone else. With this being the case we as an industry struggle and constantly ask “now what?” Now what do we do that the initial fun wore off. How to we engage with those who are there and grow the connectivity to those we want to reach.
I still get questions like “how many followers is considered good.” We need to be asking how many (as a percentage) of our followers are we engaged with. How many re-tweets do we get. How many responses do we get? How many people do WE re-tweet? Are we engaged in topical and geographical hashtags of value?
Looking at many on this list and really studying how they use there accounts is a great place to start.”
The HCSMTX Conference starts tomorrow here in Austin, and I’m guessing that it isn’t too late to find yourself a pass and go … but at a minimum, be sure to follow the twitter hashtag #HCSMTX. And we at WCG may still have a surprise our two up our sleeves. 😉 UPDATE: Be sure to click through to the 3rd and final post in this series: Bringing Data to Life
* National account used by Texas-based facility