Some folks collect stats, but our team collects GIGS of stats to better analyze what’s going through the pipeline.
Lately, our team has done some tactical drilling into our article databases to show me the most commonly mentioned keywords sorted by category. I suppose you’d call it a keyword density report and as the number of articles climbs towards 80k, we’ve got a statistically quantity to review the data with some level of confidence.
So, what do we do with this data now that we have it?
I’d agree the data would be valuable if it helped an author to figure out what to write about based on what folks are surfing for, but in this case, we have data that is AFTER THE FACT and not before the fact.
…meaning, this keyword density data shows us what our authors are writing about, but it doesn’t really tell us what the marketplace is searching for…just what the marketplace is producing in terms of commonly repeated keywords and keyphrases sorted by top level category.
One of my fears with this data should we release it (currently the decision is to not release it) is that we would edge closer to the keyword whores of the world that write crappy articles based on keyword density metrics…something I have no respect for. I think it’s ok to know what your keyword density is, but it’s better to care about departing valuable information, secrets, tips and expert advice than it is to hit some keyword density value.
One of the ways we’re considering using this data: To help identify authors who produce articles that are too mechanical, too perfect, and too contrived. It’s important that we separate ourselves from the article authors that are only here to game the system.
The difficult part is to figure out how to get our servers to help us tell whether an author is genuine in their article content vs. when they are clearly writing to nail some keyword density metric given to them by their SEO/SEM boss.
Conversely, we could also use this keyword density stats to tell an author when their article repeated a keyword or keyphrase too many times so that they can make their articles in a better natural language that doesn’t look hype-y.