On Monitoring an Online Story from Inception

I was interested in seeing how a particular story was being distributed online.

What sparked my interest was a PR firm report on the topic. I was mostly curious to see if I could replicate in a short period of time at no cost, what the firm was probably earning a fair dollar for. To clarify, I only saw the report, not the assignment, fee, or anything else in that regard.

Tangential person I am, I wound up testing Google search across a few different terms on specific date ranges. I then recorded the number of results per day and graphed them to try to gain a sense of how consistent the different searches where in their growth curve.

The numbers along Y axis are search results, and X axis are days from article release with 1 being release day.

I first searched the exact title of the article in “” and used date range function beginning on day one, ending day one, two, three, etc.






What I noticed in the early results was that the headline was included on a lot of pages to encourage people to click on it, but the article itself wasn’t included, which resulted in a lot of pages rendered.

I then searched  company name and a keyword from title of the article.






I got about a two day delay, but roughly the same results. So it seems that  Google simply needed some time to index the world and the results were reasonably close.

My next search was the company name along with a variation on the keyword. This particular keyword had two variations, so I conducted a search using the variation that was not in the title.






In this instance the results dropped dramatically, roughly flattening out at 800.

I got even more specific and went with company, keyword variation, and country where the article was originated from.






It dropped the results by about 100, but I was feeling pretty comfortable that these were more of the articles rather than headlines or links to the articles.

So I next tried the country and keyword without the company in order to get a sense of how much of a needle in a haystack this article might be without already having all the specifics and got this response:






In this instance the range jumps to between roughly 20,000 and 60,000 results.

Finally, I went to a very unique keyword in the story, a person’s name, and searched with quotes. It should be noted the name was not in the title.






I scanned through several pages of results and most started with either the first sentence of the first paragraph, or the first sentence of the second paragraph. It really seems that this is the true image of extent to which the article spread.

Finally, using Google Trends I compared the two versions of the keyword and the variation was the common use of the word and the keyword used didn’t have enough instances to even rank. This further indicates to me confidence in the results. If conducting this type of search, be careful and watch your results because Google will change your search for ‘miss-spellings’ be sure to switch results back.

I look forward to replicating this a few more times to see how the patterns compare to one another.

If you have any experience/confidence in this type of tracking would appreciate any tips on types of keywords and searches are most valuable.


Comments are closed.