Why web-based search engines don’t cut it for market research

January 25, 2019
Brock Jones

Our recent blog post Why is it so hard to find your past research? discussed why you need a solution that can easily get you to the relevant data point in a document. So why not just load your studies into a document or content management system and slap on an internal site search? Then you’ll be able to retrieve every relevant document with every relevant data point right? Hold up…It’s not that easy. There are all sorts of problems relating to words, vocabulary, and language that must be overcome.

Let’s say you’re looking for data on what keeps people from purchasing jumbo-sized candy bars for your brand. It’s a pretty straightforward question, but it could be asked in several different ways:

Why don’t people buy jumbo-sized candy bars?

Reasons why not buy jumbo bars

Barriers to jumbo

How does a search engine ensure that these questions match up to the right results, however they may be worded? An internet search engine uses statistical techniques that take advantage of the fact that they are probing the entire internet. And the “corpus” comprising all of the language on the internet includes vocabulary from a myriad of domains, such as techie terminology, business jargon, sports idioms, you name it. In this way, they are drawing from a huge repository of words, sentences, and phrases that make it easy to detect patterns.

In your case, the corpus comprising your market research data isn’t nearly as robust. Add to that all of the jargon, acronyms, and other terminology that are specific to market research, your company, or even your brand, and you’ll find that a web-based search solution just won’t cut it.

That’s why your knowledge management solution needs a search engine that can understand the structure and meaning of the language it’s working with. For example, your search engine should account for the different ways market research study questions can be asked, along with the ways in which a search term can be phrased so that it knows that phrases like “Why don’t people buy…” and “Reasons why not buy…” are essentially the same question.

You should also make sure that market research terminology is programmed into your search engine. For example, you want it to know that “Barriers to…” refers to why something isn’t purchased. In addition, you want your search engine to be customizable to suit your specific vocabulary. Does your brand use the word “jumbo” to refer exclusively to “jumbo-sized candy bars”? Make sure that’s programmed into your search engine so that your users can search using the terminology that’s most natural to them.

Most people assume that search is a one-size-fits all solution — it’s certainly not. Make sure your knowledge management system has a search engine that’s designed around your needs. At KnowledgeHound we know that the data you’re often looking for is in the study questions, and that’s why our search engine is designed to match up your search terms with the studies that contain the questions you’re looking for.

Want to access the data behind this report for free?
Create a free KnowledgeHound account to explore and analyze the survey questions behind this study.