This is the second part of our on-page word count test, where we look how many words should a page have. In our first test, 1000 words won. For this test, we increase the number of words tested. The content was identical lorem ipsum ranging from 1,100 words to 4,000 words, in 100 word increments. The concept is that, we will find where we will get diminishing returns. Clearly, an infinite number of words is not going to make us rank so there is going to become a point when the results become muddled, or maybe they inverse? This will show us where the cut-off is or the sweet spot is in terms of how many words you should actually have on the page.
Each page maintained a keyword density of 2% and had the same Meta Title. If you recall in the first test, we put 100-1000 words on the page and the pages ranked nearly perfect, from 1000 down to 100.
The results on this test is also just as amazing as the first one. Our hypothesis was that the pages with more words will continue to outrank the pages with fewer words. However, at some point, we would see diminishing returns in the form of muddled or inverse results, perhaps in the mid-3000s. We were correct in thinking that we would start getting inverse results. We were just wrong about how soon it was!
1300 words looks like the sweet spot. In both these charts you see that 1300 takes the lead and then rankings decline with the additional words (except for 1100 which tanked for some reason).
Also, you’ll notice that some pages in the test are ‘missing.’ Initially, every page was indexed and displayed. There was then a shift pushing the pages around a little, while some pages dropped from the index. In testing, losing pages like that happens from time to time, especially once the rankings have locked in.
It’s amazing in that, they pretty much rank in inverse order. From 2100-4000 words (perhaps even after 1500), there isn’t an advantage
Did you think we were closing this test series at that? Now way. Move on to the third round of this test by going to our test articles lists.