As I said yesterday, if you are searching for an ambiguous brand using Brandwatch, you’ll need to use more complex syntax in your Query. Here are some more top tips for getting it right:
Use NEAR/n instead of AND
Using NEAR/n to link brand and context terms closer together will often help to improve the overall relevancy of mentions returned.
Play around with the proximity used with NEAR/n to see if you can improve capture volume whilst still maintaining good quality results.
Use several search groupings
Using parentheses in your Query string, you can create several different search groupings if you want to.
Parentheses can be used to group together terms that require the same operator being applied to them, to nest together terms that need to be found together on a page, or to close off separate parts of the search string.
You can also use line breaks in your Query to separate out parts of your string, so it is easier to see what’s going on.
Brandwatch Query strings can be up to 4096 characters long, giving you more than enough space to construct highly specific Queries.
Keep it simple if you can
If your brand is not ambiguous (like in the case of IKEA) you don’t need to link it to additional context terms, unless you specifically want to search for mentions about a particular sector of the brand.
With these types of brand searches, just remember to also think about sub brands or unique product or campaign names that might be referred to on their own without a mention of the main brand.
Most Queries will require some sort of exclusions, whether this be irrelevant sites, authors or mentions, or if there are certain types of mentions about the brand you are just not interested in.
When you’re adding exclusions to a Query string, make sure that they are specific to the type of mentions you want the system to ignore.
You can test any exclusion terms that you are not sure about by changing the NOT in your string to AND then check that the mentions returned are the ones you want Brandwatch to ignore.
Remember to change the operator back to NOT once you have tested that the exclusion terms are capturing mentions you’re not interested in.
Test, Refine, Test, Refine, Test…
Always test and check the returned results continuously in the Query creation window as you build your Query and add new brand/inclusion or context terms.
Use the test results to highlight additional terms or phrases that could be added to your string, or terms or phrases from irrelevant mentions you want to exclude.
Before you save your Query, make sure that you scan through a good proportion of the results, checking throughout to see how accurate the mentions returned are.
For those on mention based accounts, it is important to note that from the moment you save a Query the mentions count towards your allowance, so make sure the you aren’t paying for irrelevant data you don’t want!
Once you’ve saved your Query, always review the mentions in a dashboard before you start using the data to gain insights about the brand.
A good technique is to click on spikes in the history graph on the summary component – to check that it is being caused by relevant mentions about the brand – and also view the Topics component to see if there are any obscure terms that are coming from irrelevant mentions.
Pronouns technique for big brands with lots of noise
When writing Queries for really big global brands, you will often see lots of advertising, sales listings and other types of noise returning that can cloud up results and make it difficult to get to the juicy insights.
If you are mainly interested in finding out what consumers are saying about a brand, there is a technique that we use to help pull out such conversations.
Someone writing from a personal perspective will often use personal pronouns to express their or others opinion or experiences, so by adding them to your search string and instructing the system to find mentions that include them as well, it can help you get to the mentions that are likely to be most important to you.
((I OR Im OR Ive OR me OR my OR mine OR myself OR he OR she OR his OR her)
NEAR/25 (“Mc Donalds” OR “Mc Donald’s” OR Mcdonalds OR “Maccy Ds” OR “Maccy D’s” OR “Mc Ds” OR “Mc D’s”))
Knowing when to stop
This is one of the most difficult elements to writing Queries, as we don’t know exactly what’s out there so it’s impossible to know how many good results your Query should be returning.
Asking yourself these questions will help:
- Is the volume you are seeing on testing what you would expect for the type of brand? i.e. huge global brands should return mentions in their 100,000s (in certain languages)
- Have you thoroughly tested the query and checked a good proportion of the mentions returned for relevancy?
- Are the results you checked >90% accurate to your search objective?
Work in progress…
In order for your Queries to continue to track as much activity as possible, it is important to periodically determine whether any new inclusion or context terms need adding.
If any new campaigns are launched that you want to keep track of, make sure you check that your Query will pick up the mentions with the current terms, and add new ones if necessary prior to the campaign launching.
In summary, follow these steps:
- ….and spend time working on your Queries and optimising searches as much as possible (particularly when dealing with ambiguous brands) so you are not working from and making decisions based on bad data.
If you don’t have time, let us do the hard work for you – contact us about our Query-writing services.
If you’re already using Brandwatch, hopefully the above tips will help you with your Queries. If you ever get stuck, please don’t suffer in silence.
There are lots of things Brandwatch can do to help: User Guides, extra Training or you can even get us to write them for you. Speak to your account manager or the support team to find out more.
Similarly, if you’d like more information on how Brandwatch can help give your business the edge in social media, just get in touch to find out more.
Keep calm and Query on!