Tina Cary keeps a list of geospatial companies that tweet. Adena Schutzberg reviewed the tweets in the list for two days in September, the 12th and 13th, when the list included 216 companies. Last year she used tweets from September 9th and 10th, when the list included 130 companies. Schutzberg put this year’s 436 tweets (234 tweets last year) into the same 10 categories as last year to get a sense of what was being communicated. Some tweets fell into more than one category. Six tweets written in languages other than English were removed from the 2011 analysis for a total of 430 tweets analyzed.
In this article, Schutzberg and Cary share the categories and how many tweets each contained, and provide their observations on the data both from 2010 and 2011.
Categories
Below are the categories and the raw number/percentage of tweets in each. The first value is for the 2011 analysis, the second for 2010.
126/75 (29/32%) About Us - These tweets included links to white papers and articles about or referencing the company, company blog posts, meet our employee(s) stories, and links to company newsletters. For 2011, this category also included job openings and paper links.
4/40 (0.1/17%) Weather - In 2010, one organization tweeted the current weather 40 times during the two days. I presume these data were from an automated service and that the conditions noted were at the company headquarters. In 2011, none of the handful of weather posts were automated.
34/30 (8/13%) Presentations - These tweets included "we will give/just gave a presentation at an event,” "see us on TV," "join us for a chat" or "see us in a video." The ones related to events rarely had links to content, while all the others included a link. For 2011, these tweets also included “check in” type statements though none were for services like Foursquare or Gowalla. That is, they were not automated; instead they read on the order of “I am at this conference.”
6/20 (1/9%) Press Release - These tweets provided titles and links to company press releases or about products the company sells.
70/18 (16/8%) Retweets (RTs) - A retweet is essentially forwarding a tweet to that company's followers. The idea is that the message is important enough to share with a larger community. Few of the retweets included any commentary on the original tweet. In 2011, fewer RTs actually included “RT,” but might include “@xcompany -->” implying “x wrote” or use the term “via.” Very few used MT (modified tweet) or included any commentary.
50/17 (12/7%) Sign Up - These tweets included requests to sign up for events or demos, to vote in some sort of balloting or to provide feedback on websites or topics. The requests for feedback numbered just two in 2010. I recall no specific feedback requests in 2011, but there was one asking followers how they use the company’s software in one area. (put safe graphic here)
148/16 (34/7%) Other - These tweets covered topics such as "Happy Birthday," "Follow Friday" (a list of suggested tweeters to follow), "what I'm reading" and "FYI." One company offered URLs with no further explanation. These "URLs out of context" constituted the majority of the "Other" category. In 2011, “just URL” posts continue to be the majority of “other” tweets. Also added in 2011: OH tweets, that is, “overheard.” Two active tweeters included one commenting on his particular focus within the company (Microsoft tech) and a second who worked alone and offered details on each issue she faced during the day.
10/12 (2/5%) Pictures - These included links to satellite images of areas in the news and pictures of company staff. The Boulder wildfires were still raging during this two-day period in 2010. In 2011, the majority were twitpics, that is, pictures taken with a cell phone and sent to twitter. Most images were of people at conferences.
22/7 (5/3%) Good Article - These tweets referred to an article not about the company, but deemed "good." Some of these were about topics or technologies relevant to what the company offers. Many of these were also retweets.
8/5 (2/2%) Thanks - These tweets thanked those who had offered kind tweets or offered congratulations to the company for some achievement. Many were retweets.
Adena Schutzberg's Observations
Comments for 2010 are in plain text; those for 2011 are in italics.
One-way Street
There was one thing that was missing from nearly all the tweets in all the categories: the @reply. That is, almost none of these tweets were directed at anyone in particular. Instead, they were all sent out, in bulk, seemingly unprompted by a user's or potential user's or partner's query or comment. These tweets suggest that, at least over the two days studied, Twitter was used as a one-way communication tool, a broadcast mechanism. The most interactive action was asking followers to do something (see "Sign Up," above). Other than the two requests for feedback, there were no efforts to interact on Twitter itself.
In 2011 more tweets included an @reply, perhaps 15 or 20%, but most still had a broadcast quality about them.
All About Me!
As often happens in a one-sided conversation, the main pronoun used in these tweets was "we." That is, most of the content was about the organization itself, not about its partners, users, resellers, etc. Save for the two highlighted requests for input, there were no queries to followers of any kind. That suggests to me that those behind the content creation for these tweets were thinking not about the new social/interactive media, but about broadcast media: non-social, non-interactive, old media. In particular, they were posting information into the ether just like TV ads or print advertising.
In 2011, the percentage of “about” tweets was about the same as last year (29/32%). The largest contributor to this category was probably a GPS vendor who also repairs devices. The company sent out about nine slightly different tweets each day with a link to the company website and a different location that it served.
Weather?
I can only guess that the company spewing forth weather was in a bind. Someone, probably early on, may have set up an automated weather distribution and then could not seem to undo it. I know this happens, as someone did that with our All Points Blog and then contacted me about how to "turn it off." I had to remind them that they, not we, had turned it on in the first place!
Thankfully the organization behind the weather tweets either shut down the automated feed or was removed from the list! I suspect peer pressure of some kind helped solve the problem. I compare this oddity to early websites which very often included a widget with the local time and weather. Those disappeared in most websites’ second iterations.
Most Entertaining
Reading the 200-plus tweets was a bit taxing, but one organization's tweets stood out because they were clear, varied and offered something of immediate value. Which organization was that? NASA. I'll agree NASA may not be on everyone's list of geospatial companies and it does not "sell" anything in the way that many on this list do. Still, the tone of the tweets and the clarity was very appealing. Come on, tell me you wouldn't want to check on this: "Chat now with summer intern Heather Arneson about what she's doing to keep planes on time even in bad weather. http://cot.ag/bv016y"I might even follow NASA after I finish writing this article!
NASA continues its solid tweeting tradition in 2011. This year I select Morten Nielsen, @dotMorten of Esri as the most entertaining tweeter. I’d not heard of him before this study. His bio: “Silverlight MVP. ESRI Silverlight API Dev Lead. GIS, APIs .NET, Silverlight, WPF & JavaScript.” He covered all things Microsoft, many of which I didn’t really understand but did it in what seemed a very personal (“here’s how I feel”) sort of way. In short, this feed does not read like a PR effort. Perhaps that’s why he has nearly 1,500 followers.
Biggest Waste of Tweets
A tweet with just a shortened URL in it and nothing else, such as http://bit.ly/9SS309sld (that's an intentionally fake URL), was perhaps the least appealing corporate use of the service. First off, there was no telling, in most cases, the full URL of the target, so it was possible you were headed for a virus-laden site or worse. Second, would you click on a link if you had no idea whether you were heading to a video, a "great" article or a press release? And, how would you know if it was of interest? If the organization can't take a few minutes to use the rest of the 140 characters to explain why you should click through, I think its Twitter ID should be revoked.
This situation continues unabated. There continue to be tweets of “just a link.” However, with new tools in Twitter’s app and on the Twitter homepage to decipher many of the shortened URLs to their full URLs, readers of these feeds are more likely to have a sense of source. Still, with built in shorteners in both Twitter apps and on the Twitter homepage, there are now plenty of characters for tweet writers to explain why a reader should click through. Most, however, still do not.
Identity Crisis
The best news about these accounts was that most had complete bios on Twitter with the full company URL and a description of the organization and the nature of the account. The most common adjective used in the bio of the account: "official," as in "the official account of" this company. A few had an individual's picture and name associated with them. The only oddity in a quick spot check of the bios: one large company had a picture of an individual (a man) but no indication of who he was!
In 2011, the list seems to include far more consultants and small company owners. Their tweets seem to be more tied to their day-to-day work and less about marketing their organizations.
Tina Cary's Observations
Average Tweets per Day
The 234 tweets from 130 companies averaged out to 1.8 tweets per company. Since the time period was two days, this seems to indicate that on average, geospatial companies tweet no more than once a day. This may mean companies are trying to keep the time investment low while they explore what works for them.
In 2011, the 436 tweets from 216 companies averages 2 tweets per company, no change from the one tweet per day per company in 2010.
Broadcast
The overall impression made by the categories is that companies are primarily using Twitter as a broadcast medium, though the Retweets, Other, Good Article and Thanks categories did add up to 20%, or one-fifth, of the tweets, suggesting the companies are trying to "be social."
In 2011,About Us decreased from 32 to 29 percent, Press Releases from 9 to 1 percent and Presentations from 13 to 8 percent, or a total decrease from 54 to 38 percent in categories that are essentially broadcasting. Retweets increased from 8 to 16 percent, Sign Up from 7 to 12 percent, and Good Article from 3 to 5 percent. That is, categories that have a more social aspect had an aggregate increase from 18 to 33 percent. (The Thanks category, which I called social in 2010, remained unchanged at 2 percent.) While this suggests progress in understanding that Twitter is social, the fact that the Other category (which increased from 7 to 34 percent) continues to be dominated by links unaccompanied by explanation of why or to whom they would be interesting suggests that significant opportunity remains for companies to increase the value they deliver to their followers.
Not Enough Retweets
As a marketing consultant, the fact that few retweets had commentary makes me sad, for I see that as an opportunity to add value for one's followers. Sharing the reason why the company believes that a tweet is worth retweeting would give the company more personality/identity.
I continue to view curation as valuable – why does a company think its customers and prospects would find a link worth clicking? Whether originating or retweeting a link, providing such information shows respect for that most precious resource, a follower’s time. It also conveys that the company has employees who stay informed, and that the company supports the geocommunity by enabling its employees to share information.
Follow Up
This article is our first follow-up on the original article. We are exploring dividing the companies on the list into three categories based on when they joined Twitter, to see if the patterns we see vary with experience.
Our Conclusion
Our goal here was to review changes in how geospatial companies are using Twitter in 2011 compared to 2010. We hope it will help them and others think about what's possible beyond simply having a presence on the social network.