Archive for the ‘search’ tag
You already know TweetReach reports are great for measuring the reach of hashtags and Twitter accounts, but how about an individual tweet? What if you want to analyze the reach of a tweet (and any retweets of or replies to that tweet)? Our reports can do that, too! In fact, that’s where our name comes from and one of the original problems we set out to solve more than three years ago. Tweet. Reach. TweetReach.
There are a few options for measuring the reach of a tweet. You can paste the entire text of the tweet into the search box. Since Twitter works best with shorter search queries that’s ideal if you have a shorter tweet. And if you have a longer tweet, you can select a few unique words or a phrase from the tweet to search for.
Let’s try it with this tweet from @Disney.
Since this is a pretty short tweet, we can search for the full tweet text (minus the URL to keep it simple): disney #DisneyFact: An estimated one million bubbles were drawn in the making of The Little Mermaid. We also included the original Twitter handle, minus the @ sign, to be sure we’re catching all attributed retweets of the original tweet. Here’s the TweetReach report for this query:
This report includes 108 total tweets, which includes the original tweet. So that’s 107 retweets. However, you can see that the original tweet only has 73 actual retweets, according to the Most Retweeted Tweets section. What’s going on?
This is where it gets a little messy. Some people will retweet a tweet with Twitter’s official RT button (we call this a new-style or automated retweet). Some will copy and paste the tweet and add “RT @username” to retweet (old-style or manual retweet). Some will modify the original tweet by adding their own commentary or abbreviating the text (modified retweet or MT). Some will simply quote the tweet without adding any RT language (quoted RT). Twitter typically only associates that first type (new-style RTs) with an original tweet to count them as retweets.
But in a TweetReach report, if a tweet starts with “RT @username”, regardless of how that retweet was generated (new-style or old-style), it will count as an official retweet. But if there’s anything in front of that retweet, such as commentary or other characters, then it will not count as a retweet, but it will show up in a report for that tweet. So that’s why the above report only shows 73 actual retweets of the original tweet, but there are 108 total tweets in the report. One of those tweets is the original tweet, 73 are official retweets, and the 34 remaining tweets are modified or quoted retweets. So the full reach of this @Disney Little Mermaid tweet and all its various retweets is 1,322,791.
A few more examples:
Search for: SFGiants amazing pic bradmangin melky cabrera 7th inning
Search for: wired “Hot New Characters Will Invigorate Game of Thrones”
Search for: tweetreachapp measure share of voice on twitter four steps
Tips for measuring the reach of a tweet:
- Keep search queries short
- Include handles without the @ sign
- Put exact phrases in quotation marks
- Select unique words for your query
- Leave out URLs to keep it simple
PS – Have you ever tried our TweetReach Labs Retweet Rings tool? It’s a fun, animated visualizer to see how retweets spread.
There’s one question our support team gets asked more often than anything else – how far back can TweetReach reports go? And it’s no wonder we get this question all the time; it can be pretty damn confusing. How long are tweets available? Why aren’t they available for a week or more? Why does this seem to change from one day to the next?
First, a bit about how TweetReach reports work. Our snapshot reports – both the 50-tweet free report and the full $20 report – are generated from Twitter’s Search API. You type in a search query, which can consist of one or more hashtags, keywords, usernames, URLs, and so on, and then we run that search through Twitter’s Search API to find all matching tweets. So our snapshot reports are dependent upon the tweets accessible through Twitter’s Search API.
It probably goes without saying that Twitter handles a lot of data. A lot. Twitter currently processes around 200 million new tweets a day, resulting in more than 350 billion tweet deliveries every single day. By our (very rough) estimation, there have been something like 1.75 trillion unique tweets posted in the past 2.5 years. Without getting too technical, let’s just say that it’s pretty hard to keep a service of this magnitude running. Because of this scale, Twitter can’t possibly keep trillions of historical tweets accessible to anyone at any time. Which is why when you go to Twitter Search or run a TweetReach report, you’re probably only going to find a few days worth of tweets. It’s just too hard to keep any more reliably and consistently available.
One of the things we love about Twitter – or at least that we have long since learned to live with – is that it can be a bit unpredictable. It’s a huge application with hundreds of millions of accounts; there will be occasional fail whales and things are probably going to change from one day to the next. One thing we know for sure is that it will continue to get harder and harder for Twitter to make older tweets available through search. The good news is that we’ve been doing this for a long time and have a number of ways to deal with these inevitable changes.
- Every day at TweetReach, we look at hundreds of reports to understand how far back search is going on that day, and we post current search conditions on our helpdesk so you can always be up-to-date.
- We are experts at constructing search queries so we can get the most and best possible data from Twitter.
- We built our Tracker to monitor and archive your tweets so that you don’t lose them after a few days.
- We’re here to help you figure out how to find the tweets you need. When in doubt about a search or a report, ask us!
This brings us back to the most frequently asked of our FAQs – how far back can a TweetReach report go? The simplest answer is that our one-time snapshot reports – both the free and the full $20 versions – go back as far as Twitter’s Search API does. And right now, the Twitter Search API goes back a few days (the exact number varies, so check here for current conditions). The more in-depth answer is that, if we know about your event, campaign, or promotion in advance, we can use our TweetReach Pro service to track and save your tweets for weeks, months or even years. TweetReach Pro comes with Trackers, which connect to Twitter’s real-time Streaming API instead of their historical Search API. This means we can actually save your tweets on our own servers the moment they’re posted to Twitter, and then you can access them later because we’re not dependent on Twitter keeping those tweets available.
So, if you’re confused about your search results or curious about what tweets you can retroactively access, let us help you. Seriously, we’re here if you have any questions – just ask!
Photo credit: Search. by Jeffrey Beall
Welcome to TakeFive with TweetReach, our ongoing series where we talk with notable members of the social media analytics and measurement community, pulling together insight, commentary and conversation around all things measurement. As always, please let us know what you think!
TweetReach: Welcome Jen! Let’s kick it off with a question about measurement. How important was measurement in your initial strategy for social media marketing and how has that evolved?
Jen Grant: The importance of measurement and proving ROI has become incredibly more important as our clients’ social media strategies have evolved. Primarily because more internal stakeholders are involved and excited to see results, but also because campaigns are maturing and we need to constantly adjust tactics for higher success rates.
TweetReach: What about about consistency in measurement? Agencies and marketers have had to use a variety of tools and metrics to analyze the performance of their social media efforts, resulting in inconsistent results. How important is the ability to measure and report on social media results in a consistent way to your agency and your clients?
Jen Grant: Consistency is extremely important! You’d be surprised at all the small details that make big differences when measuring social activities. Are you pulling numbers and running reports from Twitter on Mondays instead of Fridays? It makes a difference because Twitter’s API only holds data for 5 days and unless your brand is just as engaged over the weekends as it is on the weekdays, your numbers will be considerably lower.
Its also important to compare apples to apples. One simple way to do so is to compare the best piece of content from any given week.
TweetReach: For many, social media has enabled us to become more engaged with our communities. Most of us are in constant communication with our constituents, every day. How do you see integrating analytics and measurement into every-day social media activity. Is it important? How do you see this happening/evolving?
Jen Grant: One thing that social media analytics has helped me do is identify strategic partners within my social graph. I’m a firm believer in not making decisions solely on numbers alone, but I tend to get pretty strategic and scientific when I’m focused on a certain goal. I do extensive evaluations of people who I choose to engage with and consider “influencers”. Many of the considerations are subjective, but when I need to see reliable data and numbers, I rely on TweetReach.
TweetReach: Do you have any secret techniques, tools, or other Jedi strategies that you can share with our readers? Any best practices for getting greater reach for your content?
Jen Grant: Go by your gut. Do the legwork and research, but if your gut is telling you to go a certain direction, follow it — you’re almost always right.
TweetReach: Traditionally media success has been measured using reach, impressions, exposure. How important are these metrics when looking at social media campaigns? What else to you need to measure?
Jen Grant: The main reason I started presenting results in this format was because it was the only thing stakeholders were comfortable with and could relate to. After months of saying a brand’s “@mentions” had reached X, I had to throw in the towel, speak their language and make relative comparisons.
Other important measurements are engagement on blogs and other social networks, open rate of email subscribers, click through on campaign activities and overall engagement percentage.
TweetReach: Does size matter? David Armano has written about the importance of topical influence. What do you think? How important is the size of of someone’s social graph vs their influence in a particular topical area?
Jen Grant: I think size is relative. Someone’s reach is far more important. Also, since our agency is founded in search, the factors being considered by Google and Bing’s algorithms are even more important. We’ve done considerable research and confirmed the importance of many qualities held by influencers are rarely considered in traditional measurement.
TweetReach: Any examples of how analytics have helped you tweak a campaign or program for the better?
Jen Grant: Absolutely. We continuously compare reach and impressions for Blogger Outreach campaigns and tweak our selection process based on findings.
TweetReach: Any social media pet peeves? What practices irritate you the most when you look at the state of the industry?
Jen Grant: Social media pet peeve? Here goes – I’m just gonna put it out there – Klout scores!! I welcome anyone from Klout to call and explain their stuff to me, but after hours of evaluation and numerous conversations with industry insiders, I still can’t find any accuracy between what my Klout score and analysis is vs. what is happening in real life. The concept is great, but the data is always wrong (for me and my clients).
TweetReach: Thanks for your insights, Jen!
Jen Grant is Director of Social Media for Intrapromote and has been immersed in social media for almost 10 years. Jen is a social media expert having positively demonstrated the business value of Twitter, Facebook and various social media tools and applications and excels in blog marketing techniques. Her experience encompasses business development, sales and sales management, marketing, operations, staff development, coaching and mentoring, merchandising, and setting a high standard for customer satisfaction.
Jen creates and implements proven social media strategies at corporate levels to connect businesses with consumers and expand brand awareness across multiple industries. Her experience in building marketing strategies that are scalable and can be executed for brands that have many subsidiaries or locations is an invaluable asset to Intrapromote’s customers.
Venturing into new territory can evoke fear in clients. By walking through both listening and engaging strategies and marking the progress with milestones and KPIs, Jen helps social media clients realize the far-reaching benefit of social media as a marketing tool.
Twitter supports a number of advanced search operators and filters that allow you to customize your search query and find exactly the tweets you’re looking for. Here are a few of our favorite Twitter search operators and how to use them (with tons of examples).
Find one keyword OR another
First, Twitter does not require an AND or + operator to search for multiple keywords. So don’t include them. Just type together multiple keywords into your query and Twitter will return tweets that include of those terms. For example:
However, sometimes you might want to find tweets that include one keyword or another keyword. Use the OR operator to separate those terms and your report will include tweets that mention one or the other.
You can also chain together multiple keywords to create a more complex query. The OR operator will attach to the word that immediately precedes it, very much like order of operations in algebra. For example, the following query will find tweets that mention social media metrics or social media analytics, because the OR links to the metrics and analytics terms.
There are several ways to learn more about the reach of tweets from a particular Twitter account, depending on the type of information you’re looking for.
- Tweets to, from and about an account - tweetreachapp
Run a report for a username but do not include the @symbol. This will return all mentions of that Twitter account (including retweets and replies), as well as all tweets from that Twitter account. This is the most comprehensive set of reach stats for a specific Twitter account.
- Tweets to and about an account – @tweetreachapp
Run a report for a username and include the @symbol. This will return all mentions of an account, but not any tweets from that account. This report will let you know how many people are talking about a certain Twitter account, and the ways they’re talking about it (including all retweets, replies, and mentions).
- Tweets to an account – to:tweetreachapp
Run a report using the to: operator and a username. Do not use the @ symbol. This report will return only direct replies to that account (where the username is the first word in the tweet). This reports is useful for learning more about how people talk to that account.
- Tweets from an account – from:tweetreachapp
Run a report using the from: operator and a username. Do not use the @ symbol. This report will return only tweets from that account. This reports is useful for measuring the reach of an individual Twitter account, and for learning more about the kinds of tweets that account is posting.
You can filter your search results to a particular time period by adding the since: and until: operators to your search query. Use these date filters to narrow down your results. And since you can access up to 1500 tweets per query, if you run a report for each day of a campaign using date filters, you can find more total tweets. For example:
social media since:2011-09-24
You can use one or both filters in a query. These dates correspond to around 12:00 a.m. UTC, so since filter dates will include tweets from that date, but until filter dates will include tweets up until that date. And no matter what, snapshot reports can only go back about a week, so you still can’t use these filters to access tweets older than a week.
You can exclude certain keywords from your search by adding a minus sign (-) before the keyword. This will filter out all tweets that include that keyword. This is particularly useful if your company/brand/client/product has has a common name and want to exclude mentions of others with that name.
These are some of our favorite filters and operators, but here’s the full list of advanced search operators if you’re interested in more. One word of advice – Twitter handles fairly simple queries really well, but tends to break with longer and more complex queries. We recommend that you only add in a few advanced operators per query and try to limit the total number of keywords and characters in a search query. Keep it under 5-8 words and 60 characters and you should be fine. And definitely run free TweetReach reports to test out your more complex queries and see what kinds of tweets they find.
If you ever have any questions about search queries and how to get exactly the data you need from Twitter, just ask us! We’re big Twitter search nerds and can help you figure out even the trickiest search queries.
Every day, thousands of people run a free snapshot report at tweetreach.com. These free snapshot reports analyze up to 50 tweets from the past 3-7 days about any topic. You can search for a keyword, hashtag, URL, username, brand or product name, or any combination of those. You can even filter your results to a specific date or use other advanced search operators. You’ll get a full analysis of those 50 tweets, including metrics about reach, exposure, and contributors and some pretty charts like this one here.
Interpreting Your Results
When interpreting your results, it’s important to remember that the free report shows only the most recent 50 tweets for a search term. So the report you run right now could look very different than the report you ran yesterday, or even an hour ago. Even so, after time, you can start to get a pretty good sense of what kinds of numbers are appropriate for your particular search term. Here are some guidelines for how to interpret your 50-tweet report. We also have a detailed explanation of how to interpret a full 1,500-tweet report.
In a 50-tweet report, the overall exposure could be anywhere from a few thousand to a hundred thousand. On average, a free report will generate somewhere between 20,000 and 50,000 impressions. If your exposure is more than 50,000 in a 50-tweet report, that’s generally good – your message is spreading. Exposure is our count of total impressions generated by a search term.
For 50 tweets, reach will likely fall between 1,000 and 100,000. The reach number represents the total number of unique Twitter accounts that tweets about the search query were delivered to – it’s a measure of your potential audience. So, if in 50 tweets, your search term only reached a few thousand people, that’s pretty low. Are most of the tweets @ replies? Are many of the tweets posted by the same person or few people? On average, a 50-tweet report will reach somewhere between 10,000 and 30,000 people. If your reach is more than 30,000 in a 50-tweet report, that’s great. That likely means one or more well-connected people have tweeted about your topic and a wide variety of different people are tweeting.
If you divide your exposure number by your reach number, you’ll end up with your reach:exposure ratio, which will fall between 0 and 1. There’s a in-depth discussion of the reach:exposure ratio here, but basically, you want to aim for something 0.2 or higher, with most reports falling between 0.2 and 0.6. The closer this number is to 1.0, the more distinct and separate contributors are represented in this report. That means a variety of people from all over Twitter – each with their own unique set of followers – are tweeting about this topic.
Sometimes, a 50-tweet report will include tweets from 50 different people. That’s actually pretty rare; generally, most 50-tweet reports include tweets from 25-45 people. If the number of unique contributors is lower than 20, then one or more people are tweeting repeatedly about your term. Is this something you should be concerned about? It will depend on your situation, so look closely at the top contributors. Is someone spamming his/her followers about this topic?
If your search term hasn’t generated 50 tweets in the past few days (i.e., if the free report returns fewer than 50 tweets), why? If you ran a report that only looked for tweets from a specific Twitter account (from:username), then that probably won’t have 50 full tweets in it. But most other terms should get to 50 tweets in a week. What can you do to get more people talking about this topic? Start thinking about what you can do to increase the conversation around this topic.
You can actually learn a lot about a topic in just 50 tweets. Here are some ideas for how you can use the free report to help measure your impact and get more out of Twitter.
Track your numbers over time. Run a report for your brand or company every morning. Are your metrics growing? Who are your biggest advocates? What can you do to improve these numbers?
Monitor your competitors. Run a report every week for each of your competitors. Enter their stats in a spreadsheet. Use these baseline numbers as a guide to see how you stand up to the competition over time.
Count retweets. The query from:username OR “RT @username” will return tweets from and retweets of a particular Twitter account. It’s a great way to see the reach of your recent tweets and how many retweets you’re generating.
Find new blogs. Search for important keywords for your company or industry and add filter:links to your query. This will return only tweets with links in them, and could lead you to some new reading material.
Watch news spread. Enter a URL or short quote from a press release or blog post you just published. Run a report once an hour to see how the article is spreading around Twitter.
So, what are you waiting for? Give it a try and see how your numbers stack up! It’s totally free to run a 50-tweet report and you we won’t ask you to log in or give us your email address. And if you want more than 50 tweets, you can always buy the full report for your search query, which will include up to 1,500 tweets from the past few days .
Getting relevant, useful and actionable data from Twitter can be difficult. There are nearly 100 million tweets posted to Twitter every single day, and most of those are spam or simply irrelevant. Culling the mass of tweets down to the ones that matter to you – tweets about your brand, promotion, competitor or client – is something we work on every day at TweetReach. The first step in getting quality data is getting better at search. Crafting a good search query will save you a great deal of time and frustration later.
Join us for a free interactive webinar on November 11, at 11:00 a.m. PST, where our chief social scientist, Jenn Deering Davis, Ph.D., will share tips, tricks and best practices for creating effective Twitter search queries. We’ll discuss the following issues:
- Important Twitter search considerations, including key search operators and how they work
- Twitter search limitations and how to get around them
- Disambiguation and other data quality issues
“Search Twitter Like a Pro” Webinar
Thursday, November 11, 2010
11:00 AM – 12:00 PM PST
Sign up here
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.