We've all been in a situation where we set up a search, only to find on the summary page of the setup process, that it will return a gargantuan amount of mentions 😅.
This article will help you to better understand the different tips and tricks that you can employ in order to better refine your search setup and ensure that the number of results estimated will fit neatly into your data allowance.
More on available operators here.
Lets go through creating a new search together! First, create a new topic search around the topic of your choice (for this example, lets look at Coronavirus). Select your data sources and either the Wizard or Boolean setup.
Click through the next steps until you get to the Summary stage, where it show you a preview of the results from the past month:
Now, this search is showing a sample of 9 millions mentions from X from the last month! This is a huge amount, but if you want to reduce the scope, you can implement the following operators to the search setup:
Sample operator
Location operator
Language operator
Exclusion operator
AND NOT RT operator
AND NOT REPLY operator
Sample operator
The Sample operator allows you to sample your real-time tweets down to the specified percentage. In the given example, this will collect 10% of the tweets that are available on X that contain either the keyword coronavirus or the keyword covid. The Sample operator must always be capitalised and must be used with the AND operator.
Now that we have added the Sample operator into our, lets go to the Summary stage, and we can see that the volume has significantly decreased from 9m to 900k mentions:
🚨 Important: The SAMPLE operator is only available to use on real-time Twitter data. You can find out more information about it here.
Location operator
Another way we filter down the results is by using the location operator. Go back to the Topics stage of the setup, and type in AND LOCATION and then any countries you want the content to come from. In this example, I want to collect mentions from the UK only, but you can add as many countries as you like.
If we then click through to the Summary stage, we can see that 274k mentions have been collected from the past month, with the location operator added:
In conjunction to the location operator, you can also add in the sample operator to the boolean and this will further reduce the volume of data.
🚨 Important: We are not able to get location data for the following data sources: Instagram, Reddit and YouTube.
Language operator
Similar to the location operator, we also have the language operator available which you can add to your boolean to see results from a particular language. Simply type in AND LANG and then an opening bracket, then a list of all the different languages will appear for you to choose from:
Exclusion operator
Use the AND NOT operator to exclude posts that contain certain terms from being collected into your results, effectively filtering out any noise.
AND NOT RT
Use the AND NOT RT operator if you want to restrict your data collection to just original tweets and exclude all retweets. This operator must always be capitalised and can be added at the end of the boolean expression.
For example: (apple OR juice OR fruit) AND NOT RT is a valid expression.
However writing this: (apple OR Juice OR fruit AND NOT RT) is not a valid expression.
AND NOT REPLY
Using this operator will avoid collecting tweets that are replies to other tweets. Like all operators, the AND NOT REPLY must always be capitalised.
For example: (apple OR juice OR fruit) AND NOT REPLY is a valid expression.
However writing this: (apple OR Juice OR fruit AND NOT REPLY) is not a valid expression.
💡Tip: One final step, always make sure you add a second set of parentheses around the entire keyword expression before adding AND operators.
We hope you enjoyed reading this article! 📚
If you have any questions or would like to learn more, please don't hesitate to reach out to our support team via live chat. 🚀