By One of the challenges we face at SparkToro is deciding between comprehensiveness vs. relevance in the audience data we deliver. In a nutshell, it’s the difference between knowing all the behaviors an audience might have (including the weird, random affinities, strange correlations, and surprising overlaps) vs. the ones that are most useful and compelling to a marketing campaign. Historically, we’ve erred on the side of comprehensive, but the more we talk to customers and get feedback, the more we realize relevance *really* matters.
Today, we’re taking a big step in that relevance direction with new, upgraded keyword data in all of our audience research reports. Here’s an example:
I’ve analyzed an audience of high income homeowners seeking modern lighting design solutions, and SparkToro’s new keyword data is shining through 💡💡💡😉. Whereas our previous system returned a more comprehensive list of terms and phrases people search for in Google (but not necessarily as on-topic and relevant), this new system is BOTH surfacing more relevant keywords AND filtering out the off-topic ones.
Here are a few examples that might help you grasp what we’re changing:
- If you researched an audience of consumers seeking modern lighting design, we’d show fewer keywords related to news items about energy prices (something this audience does search for, but isn’t particularly useful/relevant for marketing lighting design to them) and more keywords
- If you researched travelers seeking vacations in Costa Rica, we’d ignore the off-topic keywords around flight numbers, Disneyworld prices, or fitness trackers (we’re not sure why these audiences search for those last two with higher-than-normal-affinity, but they do), and show you the Costa-Rica-vacation planning search terms that really matter to reaching this audience.
- If you researched Infosec executives at Canadian companies, our keyword lists will no longer include oddly correlated, but hard-to-see-relevance terms around the Olympic Games, the best carry-on suitcases, or high thread count bedding (maybe Infosec execs aren’t sleeping so good?). Instead, you’ll see searches related to cybersecurity, hacking, phishing scams, password hygiene, and all that useful infosec stuff.
We’re also removing some historic pieces of the app that were almost never used by anyone (despite Rand’s hope that they might): SERP features, trending keywords, PAAs, and Related Questions. In fairness, we provided these more because they came with our keyword datasets than because anyone asked for them… So it should be no surprise that <1 in 1,000 SparkToro report users ever clicked on them 😅
These new, more relevant keywords join the AI Prompt data we added last month to help make SparkToro a one-stop-shop for knowing what your target audience searches for in traditional engines, prompts AI tools for, and how much they use those platforms. Not too shabby, eh?
Stay tuned for more upgrades coming soon to an audience research tool near you (OK, it’s us, we’re the tool). 😎