If you are new to creating Searches using Boolean Syntax, please review Introduction to Simple Boolean Operators to gain a general understanding of how these searches are structured.

Common Keywords vs. Unique Keywords

One of the first things you want to consider when crafting your search is how unique your terms and keywords are.

A common keyword is one that doubles as a brand or product, as well as a word or proper noun, such as “Apple” or “Target.” Some ways in which they might be used outside of a brand mention are:

• A word in any language
• The name of a place or thing
• A person's name

A unique keyword is one that is distinct, in that whenever it is mentioned, it is referencing the brand. Examples of stand-alone unique keywords or unique combinations of keywords include:

• “Starbucks”
• "Nike"
• “IndyCar”

When approaching a search with a common keyword, the optimal outcome would be for that keyword to become unique (i.e. isolated to be only referencing the keyword you want). To do this, it is useful to identify unique characteristics of the keyword that disassociate it from the common keywords it would be confused with. Some ways to do this are through:

• Case sensitivity
• Term inclusion
• Term exclusion

Search Strategy for "Unique" Keywords:

1. Unique keywords are the simplest type of search to create, as you often do not need any additional key terms or requirements to make sure you are only getting relevant mentions of the particular brands.

2. Use the various ways your brand is mentioned (e.g. “BP” or “British Petroleum”). Also, don’t forget about possessive forms (e.g. “BP’s” or “British Petroleum’s”).

3. Often, adding the AND operator (inclusive) will actually ELIMINATE valuable results. So in this case, it may be best to keep the search quite basic.

Search Strategy for "Common" keywords:

1. Use the text: and headline: operators to lock in the upper-case of common words to help eliminate common uses of the word. For example: when locking in the case sensitivity of the word "Apple" -- you eliminate most uses of the noun "apple."
2. Use the various ways your brand is mentioned (e.g. “BP” or “British Petroleum”). Also, don’t forget about possessive forms (e.g. “BP’s” or “British Petroleum’s”).
3. Unlike with unique keywords, you will need to add in inclusive keywords using the AND operator to include specific brand or industry jargon. This could help narrow your mentions to only those that are relevant to you.
   • For example, if your company is Apple, some types of keywords you might want to use are:
     • Unique Competitor Brands: "Microsoft", "Samsung", "Google"
       
     • Unique Products/Services: "PC", "Android", "Google Home", "iPhone",
       
     • Unique Keywords/Jargon: "Artificial Intelligence", "Machine Learning", "Diversity & Inclusion", "Data Privacy", etc.
       
     • Campaigns/Events/Initiatives: "Think Different", "Super Bowl", "Best Places to Work", "F/8", "CES", "Consumer Electronics Show", etc.
       
     • Executives Names: “Tim Cook”, "Mark Zuckerberg", "Jeff Bezos", etc.
       
     • Partnerships: "Cisco", "Accenture", etc.
   • NOTE: You will be aware of most of this information for your own company/brand, and most of this information can be found with a quick Google search or by browsing your competitors websites.
4. If you are still finding irrelevancies, then you may want to consider using the NOT operator. Look through the data you are pulling to identify irrelevancies and attempt to find a theme amongst them.For example, if your company is Avant and you find that there is an NFL football player named “Jason Avant”, you would likely add NOT “Jason Avant” to your search.

Boolean Tips & Tricks

Building a Targeted Search for a Brand

To search for mentions surrounding only a segment of your business, you can build a targeted Boolean string.

For example, if you want to build a dashboard for Amazon Home Decor (not Amazon as an entire enterprise), you may build out something like this:

"Amazon" AND ("home decor" OR "lighting" OR "furniture" OR "home goods")

Creating Direct Competitive Comparisons

Competitive comparisons can be tricky if the companies do not overlap in all aspects of their business. In these circumstances, we recommend targeting only the overlapping part(s) of the business in your competitor’s search (see the previous section).

For example, imagine you are building a dashboard for Netflix. One of Netflix’s competitors is Amazon, but you would not want to search for everything “Amazon” since they have several other business segments that do not compete directly with Netflix.

By narrowing the Share of Voice search to (“Amazon” AND (“Instant Video” OR “video streaming” or “instant videos”)) you are able to see a truly apples-to-apples comparison because you are comparing Netflix (a video streaming company) to Amazon’s video streaming business segment.

IMPORTANT: If you are using an operator that searches a specific type of coverage (i.e. language, country, or mediatype) remember to apply those same operators to your Share of Voice searches so you are seeing an apples-to-apples comparison.

So, if our search for Netflix looks something like this:  

("Netflix" OR "Netflix's") AND (tk_location:"United States")

We may want our Share of Voice search for Amazon to look something like this:

("Amazon" OR "Amazon's") AND ("instant video" OR "instant videos" OR "Amazon video streaming"~10 
"Amazon’s video streaming"~10 "Amazon videos stream"~10 "Amazon’s videos stream"~10 OR 
"Prime Video") AND (tk_location:"United States")

The search above will yield all United States coverage that mentions Amazon (singular or possessive) and any of the following phrases: “instant video”, “instant videos”, “prime video”. It will also yield all United States coverage that mentions Amazon (singular or possessive) when it is mentioned within 10 words of the words (“video” and “streaming”), or (“videos” and “stream”). This is to account for mentions of Amazon Prime Video when the product is not directly mentioned. For example, it will capture mentions of “Amazon’s new video streaming platform” and “you can stream these videos through Amazon”.

Setting up a Target Publication Dashboard

Target publication dashboards are useful for narrowing results of a brand, industry, campaign, etc. to a specific list of media outlets.

Here is how you would organize your search if you were only interested in mentions of the “Summer Olympics” from the New York Times, Wall Street Journal, and USA Today:

"Summer Olympics" AND site_urls_ll:("nytimes.com" OR "wsj.com" OR "usatoday.com")

NOTE: You can also use a similar structure to exclude specified publications. For example, if you wanted to see all of Amazon Prime’s coverage in September 2015, except mentions from Consumer Reports, CNET, and any of the three major personal blog sites, you could search:

("Amazon Prime" OR "Amazon Prime's") NOT site_urls_ll:("consumerreports.org" OR "cnet.com" OR "blogspot.com" OR "blogger.com")

Creating an Industry Dashboard

Many of our customers use industry dashboards to gauge trending topics, gain competitive intelligence, find impactful articles to share and discover influencers. Unlike brand searches, where you often want to look at every mention, the key here is to find coverage where the industry is the core focus. Often, especially if your industry keywords are common, you may want to require that the terms appear in the headline or multiple times in the body of the article. However, if the jargon is unique to your industry, it may suffice to simply require the keywords to show up once, anywhere in the article.

Here is an example of an industry dashboard for the frozen meals industry:

title:("frozen dinner" OR "frozen dinners" OR "frozen meal" OR "frozen meals" OR "tv dinner" OR "tv dinners") 
OR (("frozen frozen frozen"~99 OR "tv dinner"~99 ) 
AND ("frozen dinner" OR "frozen dinners" OR "frozen meal" OR "frozen meals" OR "tv dinner" OR "tv dinners"))

The search above specifies that “frozen dinner,” “frozen meal,” or “tv dinner” (singular or plural) must be in the title of an article for it to return. Or, an article will return if the word “frozen” or both of the words “tv” and “dinner” are mentioned at least thrice, along with at least one mention of “frozen dinner,” “frozen meal,” or “tv dinner” (singular or plural).

NOTE: The "near" operator treats the same term written x times as x mentions of the term (e.g. “festival festival festival”~100 means the word “festival” must show 3 times within 100 words of itself).