Archive for October, 2010
One of the biggest challenges we face with TweetReach is the 1500-tweet, 5-day restriction on Twitter search results.
We completely understand why this limitation exists; it’s both difficult and expensive for Twitter to keep billions of tweets accessible in their Search API. But we also know how hard and frustrating it is for you to explain to your clients that we can’t include older tweets in reports. It’s unfortunate, but that’s just the way it is – tweets more than a week old are simply not available for us to access for a TweetReach report, whether it’s a free report or a full report. After seven days, those tweets are gone.
So, what can you do?
Whenever you can, create your monitoring and measurement plan early. Set up a TweetReach Tracker before you start a campaign so that we get all the data you’ll need. A Tracker finds all tweets about a term in real time, as they are posted to Twitter, and then stores them on TweetReach servers for analysis. This allows us to track tweets over periods of weeks, even months, and there’s no 1500-tweet limit. You can then analyze the tweets whenever you want, and you don’t have to worry about them disappearing in a week. The Tracker can only find new tweets, though, so make sure you set one up before your campaign starts. Even the Tracker can’t go back and find old tweets.
And if you need help setting up your Tracker’s search query, let us know! We have lots of experience with disambiguation and data cleaning, so let us help you get exactly the data you want.
Capture the data while you can.
Our one-time snapshot reports are essentially a historical analysis of the most recent 1500 tweets about a term from the past week. Since tweets are gone from search results after a week, make sure you run snapshot reports while the data is still available. Even if you’re not sure if you’ll need it, wouldn’t you rather be prepared? So run a report right now. Get the data while it’s still there.
And in cases of unexpected or crisis situations where you weren’t able to set up a Tracker preemptively, then set one up as soon as you can. You’ll want this information later, and it’s better to have some tweets than no tweets.
Maybe someday this won’t be an issue. But for now, the best thing you can do is be prepared and proactive. Set up your Trackers early and run reports as soon as you can. And if you ever have a question or need help getting your queries right, just ask us. We’re here to help.
A Shooting on Campus
On Friday, September 28, 2010, a 19-year-old college sophomore took an AK-47 to the University of Texas at Austin campus, fired numerous shots, and then killed himself in the Perry-Castaneda Library. Fortunately no one else was injured in this incident, but it was a scary and sad day for thousands of students, faculty, staff, family members and Austin community members.
Even though it happened in Austin, Texas, the shooting was national news. And for us here at TweetReach, it hit pretty close to home. We started TweetReach in Austin, and lived there for five years before moving to San Francisco. And I recently earned my Ph.D. from the University of Texas at Austin, where, in addition to my coursework and research (which meant I spent long hours in the P-C library), I also taught undergraduate communication courses for four years. Many of my friends and colleagues still work there and were on campus on September 28.
For several hours on the morning of September 28, it was unclear how many shooters there were and how many people were injured, and the entire campus was on lockdown. I anxiously watched as some of my closest friends posted Twitter and Facebook updates about armed police officers in the hallways outside the classrooms where they were sequestered. It was a tense day.
Twitter and Real-Time News
But the day was made less tense with lots of updates from friends on campus and off, as well as plenty of information coming from local news sources. Social media have certainly impacted how we consume and create news. For good or bad, breaking news spreads incredibly quickly in new media channels like Twitter.
There’s really no need for me to spend much time writing about the impact of Twitter on news – just do a quick Google search and you’ll find all the history and analysis of real-time news you could ever want. This is not a new idea. For example, two years ago, ReadWriteWeb wrote a nice piece on how they use Twitter for journalism. What I do want to write about is the specific example of how Twitter was used the day of the UT shooting.
The Austin American-Statesman has been a paragon of how local news outlets can use Twitter in innovative ways. I know Robert Quigley and the other folks at the Statesman work hard to make sure the @statesman Twitter account is useful, accurate and timely. In the past year alone, the Statesman has informed and guided the Austin community through several significant local news events, including the November 2009 Ft. Hood shooting and the February 2010 Austin plane crash. The September 28 campus shooting was no exception. And this time, we have tweet data to analyze.
How @Statesman Kept Us Informed
We’ve been running a TweetReach Tracker for a few months that measures Twitter activity with and about the @statesman Twitter account. [Note: The Statesman is not a customer of ours, nor did they sponsor this research.] On September 28, @statesman regularly tweeted updates and important information about the shooting. Those tweets were retweeted and discussed repeatedly, resulting in a huge Twitter reach day for @statesman. Huge.
On a normal weekday, the @statesman Twitter account, which has more than 26,000 followers, reaches on average around 75,000 unique Twitter users through retweets, mentions, and replies. We’ve been tracking this activity since mid-July, and before 9/28 it peaked at a daily reach total of almost 180,000. On the day of the UT shooting, that number skyrocketed to a unique reach* of 1.45 million unique Twitter users. That’s more than eight times its previous peak reach day and 19 times its normal reach.
The overall exposure (total potential impressions generated*) on September 28 was nearly 3.7 million, up from a typical weekday average of 125K and a previous peak of 292K. 2,700 Twitter users posted more than 4,500 tweets that day. In fact, 56% of all tweets with “@statesman” in them for the entire month of September were generated that day. @Statesman tweets were retweeted by many other news outlets, including the @washingtonpost and @dallas_news. They were also retweeted by lots of Twitter notables, many of whom don’t even live in Austin (a surprising amount from the Bay Area, actually).
These numbers are impressive and show the power of Twitter. How often do Statesman stories reach almost 1.5 million people in one day? I’m guessing never (their daily print circulation is about 140,000). But what else does this information tell us?
What We Can Learn
These data demonstrate the potential impact of someone in the right place at the right time (remember the first Twitpic of the US Air plane that landed in the Hudson?). As a well-respected newspaper, the Statesman was already positioned as the go-to news source for Austin area current events. But over the past few years, they’ve built a loyal following on Twitter of people who are eager to contribute to the news process. People send photos and updates to the @statesman, essentially expanding the paper’s news staff. And by making their Twitter updates consistently relevant and timely, the @statesman is insuring that they’ll be retweeted when something like this happens, drastically expanding their circulation.
These data also reiterate the importance and potential pitfalls of real-time news. If no one cared what was happening on campus, then no one would retweet the @statesman’s updates. But people do care; we demand information in real time, especially when something scary is happening. And we’ve all heard what the 24-hour news cycle does to news – stories are rushed out before being fact-checked or properly edited, stories that aren’t actually news are published just to fill space, sources are faked to get information out as soon as possible.
Finally, these data are a good reminder that you never know when something like this will happen. You need to be prepared. If you work at a news outlet (or any company or business, really), build your following now. Start measuring your impact now. You can’t suddenly have the reputation of the @statesman; that has taken them more than three years of dedicated effort to achieve. And you can’t go back and measure certain kinds of Twitter activity after it’s happened. Once it’s passed, it’s gone. The real-time speed of social media means that events can flare up and then flame out in just a few hours.
Fortunately, no one besides the shooter was killed at UT on September 28. But on that day, when no one knew what was going on and many people were assuming the worst, Twitter and the @statesman helped communicate vital information to those who needed it.
*If you want to learn more about how we calculate and define our metrics, including reach and exposure, read this.
Here at TweetReach, we’ve never been ones to turn down a good drink – and based on results from two TweetReach Trackers we’ve been running, neither have Twitter users. Over the last few weeks, we’ve been tracking keywords, hashtags and @replies related to both Jose Cuervo tequila and Captain Morgan rum – and an interesting battle of the booze battle has ensued. Cuervo is drinking El Capitan under the table by a measure of 7:1 based on impressions across Twitter, or 4:1 when you look at unique Twitter users. In September, tweets related to Jose Cuervo generated 7.8 million Twitter impressions, reaching a possible 2.5 million unique users, while about 1 million impressions (reaching 650K unique users) have had a bit of The Captain in them. See the graphic below for exact figures.
Some additional bottom-of-the-glass analysis also shows that engagement among those inspired by Cuervo terms on Twitter is much deeper, and in fact, may be bringing more exposure to the brand. For example, about 45% saw a Jose Cuervo-related tweet between 2-7 times versus about half of that amount (22%) for Captain Morgan.
Interestingly, neither brand’s website links to Twitter in any way, but both do have a Facebook presence. Neither have much in the way of official, active Twitter accounts either. So, even if you don’t officially participate in a particular social media channel, this shows that your brand can still be talked about and discussed there. And when celebrities and other accounts with thousands or millions of followers tweet about you or your product, like fake Gary Busey did about Jose Cuervo (which was retweeted 500 times generating 240K impressions), it becomes pretty clear why brands should consider participating in that activity.
Great news - Twitter has upgraded its search architecture!
What this means for you is that the search index is much bigger than it used to be, so we can go back further in the archive to retrieve tweets for reports. TweetReach reports will now include tweets from the past 7-8 days, and this number could grow. Before yesterday, we were limited to tweets from just the past 5 days – that’s a 50% increase! And don’t worry, full reports are still only $20.
Get a report using the new, extended index at TweetReach.com.
We’ve all been there. Your company or client has just completed a product launch. There seemed to be a lot of people talking about it on Twitter, so back slaps and fist bumps abound. And then the inevitable happens: your boss wants to see Twitter metrics and results – with charts – on the success of the program by the end of the day. Gulp.
It’s typically been a challenge to say exactly what kind of influence or reach your campaign generated on Twitter. So what do you do? At some places, it sets off a three-alarm fire drill as the team scrambles to gather results from a variety of sources that are tossed into a weak-looking PowerPoint slide or two.
Sound familiar? It happens at agencies and companies big and small, all across the country. But it doesn’t have to be this way.
A number of agencies find the TweetReach Tracker to be a helpful resource. This subscription service gives anyone the ability to track keywords, phrases, hashtags and URLs across Twitter, in real-time, during a campaign, product launch or designated period of time.
And the best part might be the sweet-looking charts and graphs that show reach, engagement, contributors and activity, along with key influencers who have used your keywords and phrases. Some of those charts are pictured here.
It’s easy to track pre-, during- and post-campaign Twitter volume so you can provide context and rationale behind key social media activities. Now your boss or client can see real results. With no last-minute scrambling.
We came across this piece over the weekend from the BBC about monitoring social media, and how it’s no longer something companies can ignore. In the past, an unhappy customer would complain to the company and then tell a few friends in the neighborhood. Today, the “neighborhood” is global and an individual can have quite an impact with a well-placed tweet, video or status update that goes viral — and not the good kind of viral most companies are trying for.
Some companies find not monitoring Twitter or Facebook unthinkable, while others haven’t even thought about it. The article mentions a number of tools and services out there that can help not only pull in what people are saying, but sentiment as well. But this is a tough technical challenge – terms and phrases change so quickly. Even Urban Dictionary has a hard time keeping up.
If you’re not quite sure how to get started monitoring your social media temperature out there, TweetReach has a free service to look at the last 50 tweets, but you can also get a much broader report for just $20. You can quickly learn who’s saying what about you or your company, your brand’s influencers, and all the people seeing tweets about your brand on Twitter.