What kinds of searches work well vs. poorly on SparkToro?

Our database is built on profiles’ self-descriptions (the bio/profile field in most social media websites and the about page on someone’s website) and the behavior of those profiles on the web (what sites and pages they link-to, share, talk about, etc. and the accounts they follow, amplify, reply-to, and engage-with). Thus, SparkToro tends to excel when an audience is describable based on their online behavior. And, conversely, it’s not great when an audience’s behavior isn’t visible or describable on the web.

For example, SparkToro’s great at finding data on audiences like:

  • Real estate agents (because they tend to use these words in their bio/profile text)
  • Snowboarders (because they tend to talk about snowboarding and winter sports online)
  • People who live in San Diego, CA (because they often put this information in their online profiles)
  • Folks passionate about crafting with polymer clay (because they often talk about polymer clay in the text of their public, web content)
  • Home cooks who share pasta recipes online (as they often use the hashtag #pastarecipe in their web content)
  • Fans of the website, narratively.com (or most any website with substantial traffic, as these audiences tend to share, link-to, talk-about, and follow these sites)

SparkToro isn’t so great at finding data on audiences like:

  • Men between the ages of 19-35 (because we don’t collect gender, age, or other demographic or personally identifiable information)
  • People who own homes (since homeowners rarely share their ownership status on their web profiles or in web content)
  • Individuals with a particular blood type (since this information is rarely shared online)
  • Consumers thinking about switching to a new car insurance provider (as this type of purchase consideration is rarely talked about in public social or web content)
  • Businesspeople who wear formal attire to their offices (as, again, this isn’t commonly shared/talked about online — though one might infer this attribute from various professions that are in online bios/profiles: lawyer, banker, etc)

When attempting to get data about an audience through SparkToro, it can be helpful to first consider attributes of that audience that are shared by that group in their online profiles and shares. If a word, phrase, behavior, or hashtag that people use online is indicative of the audience you want to reach, SparkToro can usually help you find it. And if they’re not, it can pay to instead target that audience through sources of influence they already follow (via “my audience follows the social account” or “my audience frequently visits the website” types of searches).