There are so many social media monitoring tools out there, how can we tell the good ones from the bad ones? In all of the reviews I’ve read about social monitoring tools, they often seem to gloss over the one feature that I think is the most important, the query. In these tools, the query is the search string you specify to tell the tool what you want to have monitored. Have you ever heard of the computer cliché, “garbage in, garbage out”? Well, that term applies to monitoring social media as well. The social mentions that are found by your tool need to be exactly as you intended, otherwise any fancy analytics and reporting features will be skewed from irrelevant results being pulled in. The two biggest reasons why the results may be full of garbage are likely because the terms you are monitoring aren’t specific enough, or the tool itself has limitations on how specific your search can be.
We’ve explained the situation in depth in our white paper, “The Importance of the Boolean Search Query in Social Media Monitoring Tools ”. The white paper includes more details and examples of reasons to use social media monitoring, what to monitor, ways to express variations in keywords, Boolean search operators, post-search filters, the types of tool user interfaces for queries, and other things to consider when evaluating a social media monitoring tool to ensure it provides the best results for what you need.
Here’s a quick look at the problem.
Let’s say you have a company, brand or product name. Of course you know how to spell it and what the official way to represent it is. But what about everyone else? It’s important to do some brainstorming to make sure you’ve included every likely variation you can think of in your search for mentions. Other terms can also be monitored, like the names of people who are influencers, competitors and their products or industry terms to supplement your market research and look for opportunities.
There may be common misspellings or it may be often abbreviated or represented differently in a Twitter handle. It may be represented by a stock symbol or a nickname. Where does the apostrophe go in a company name such as “Actor’s/Actors’/Actors Studio”? And if you’re monitoring for additional keywords or industry terms, there might be variations for each of those as well! Just think of how many names you can come up with for beer (brew, draft, ale… ) Is there the likelihood that your keyword will be abbreviated, like svcs for services or doctor represented by Dr. or doc? Anything that includes a location name opens a can of worms for variations – like New York City, NYC, Manhattan, The Big Apple, and so on.
So let’s say you’ve done some research and have a wonderful and thorough list of keywords and variations of them to monitor. Can your social media monitoring tool support searching for all of these variations without having to create a separate query and alert for each variation?
Each tool is likely to have a different type of user interface where you can specify the query. In our sampling of social media monitoring tools, we found many to be inadequate when we wanted to construct a complicated Boolean search string. All tools seem to allow the basic Boolean search operators, AND, OR and NOT. You choose which words must be in the results (AND), which are optional (OR), and which results you want to exclude if they contain certain words (NOT). Here’s a rudimentary example that looks for mentions of two types of pie but not interested in finding recipes:
apple OR peach AND pie NOT recipe
Some tools let you type in the Boolean search string just as above. Other tools go about this by giving you fields for each operator, so you might have:
Looking at the above screenshot, how would you specify wanting any of five types of fruits (apple, peach, rhubarb, cherry, blueberry) that must appear with three possible types of dessert (pie, cobbler, tart)? That user interface is too restrictive to specify this and there are too many combinations to enter as separate keywords (this particular tool only allows five keywords and we’d have 15). If you could enter it as a Boolean search string, you could simply specify it as:
(apple OR peach OR rhubarb OR cherry OR blueberry) AND (pie OR cobbler OR tart) NOT recipe
One of the problems we found was that it’s impossible to be very specific without the use of these additional query operators such as parenthesis to specify the order or groupings of what you’re searching for, wildcards to allow for variations in the words without having to specify every possible tense or suffix, and proximity operators that let us specify that we want one word to follow the other within a certain number of words to add more context to the keywords we are searching for.
This is only the tip of the iceberg. Some tools have ways to indicate if special characters are included in the search, like if you were looking for M&M’s, while others may not support such characters at all! You may be able to specify whether upper or lower case matters, only include or exclude mentions on specific web site URLs, specific author names, web page titles or geographic locations. Also be aware that some tools have limitations on how long the query string can be, or how many keywords you can search on.
Want to see if your social media monitoring tool can handle a complex Boolean query? How about seeing if it can easily search on every variation of ways to express or misspell the store name P.C. Richard and Son. People may confuse whether the name has plurals, how to abbreviate the PC part, or whether or not to include spaces. When we looked at these variations, we came up with 40 permutations!
We were able to boil down all of these variations into a short Boolean search string. This brand name is the perfect example because it requires the use of Boolean operators along with parenthesis, double quotes, special characters and proximity operators. We took this example search query and tried it out on a sample of social media monitoring tools with widely varying results. It really highlighted each tool’s limitations. Some totally failed with no way to specify multiple variations of the brand name, some would require the query to be broken into multiple separate queries (which would be a lot of unnecessary work to manage all the social mentions as numerous separate queries, alerts, analytics and reports), and others required some tweaks but gave us successful results. See our white paper for a detailed step by step case study of how we created a complex Boolean search string to cover the 40 permutations of P.C. Richard & Son.
So, if you’re shopping around and doing a social media monitoring tools comparison, take the time to do some due diligence and create a list of which features are required, and which are nice to have. Most importantly, seek out the list of which Boolean search operators are supported and find out if the tool’s user interface is flexible enough for you to create complex queries. The query functionality should be the most important factor in deciding whether the tool will work for you, before you bother looking any further into the tool’s analysis or reporting capabilities. Being able to drill down to precisely what you want to monitor and not open the flood gates with irrelevant results saves time, effort and money, and makes the analysis of the results more accurate.