Sixty Nine Cents Analysis

So You Want to Actually Use the Sixty Nine Cents Analysis Method

I've been running this method on client briefs since around 2017, right after David Szabo published it on Search Engine Journal. The approach is straightforward in theory, but the execution has enough edge cases that most people waste a lot of time before they figure out how to use it properly. Sixty Nine Cents Analysis involves copying the top-ranking pages for your target keyword into a word frequency tool, finding which terms appear most often across those pages, and then building your content around those semantically related terms. Google's algorithms use natural language processing to understand topic coverage, so the more relevant terms you include in context, the better your chances of ranking. The whole thing is basically an exercise in matching the semantic profile of the existing top results rather than guessing what Google wants.

Sixty Nine Cents Analysis Step-by-Step

Here is how I actually run through it when I need to produce a content brief quickly. First, grab the top ten results for your target keyword. Not the top three. The top ten. You need a sample size big enough to smooth out anomalies. If someone with a massive domain authority ranking spikes for a single term on page one, you do not want that skewing your entire dataset. Pull the URLs, open them in a new tab, and copy the main body text from each page. Strip out navigation, footers, sidebars, and ad copy. You only care about the actual editorial content. I usually paste everything into a single document as a first pass so I can scan for anything that looks like boilerplate that leaked in. Next, run that combined text through a word frequency counter. I use tools like WordList or even just a basic web app where you paste text and get back a frequency report. The key thing here is to toggle off stop words. Words like "the," "and," "is," "of" — they are noise. You want the meaningful nouns, verbs, and adjectives. Most tools have a built-in stop word list. If yours does not, you can download one from the Stanford NLP group and paste it into the exclusion field.

Once you have your frequency list, look at the terms ranked by count. The ones appearing most frequently across all ten pages are your semantic signal. These are the terms Google considers core to the topic. I typically highlight the top fifty to one hundred terms. Not everything on the list matters. Terms that appear in only one or two pages are often noise from a poorly structured article, so I filter those out manually. From there, I build the content outline. I take the target keyword and place it prominently in the H1 and first paragraph. Then I structure the subheadings around clusters of the high-frequency terms. If "machine learning," "neural networks," and "deep learning" all show up frequently, those belong in the same section because they are semantically related. I do not just stuff the terms in randomly. The context matters. Google can tell if you are forcing a term into a sentence where it does not belong. For a recent project on topic modeling for publishers, I ran the analysis and found that the term "latent Dirichlet allocation" appeared with notable frequency across the top results, but the second-page results were dominated by pages that only mentioned it in passing. That told me the deeper technical pages had the semantic coverage that mattered. I structured the brief around LDA as the central topic rather than treating it as a throwaway mention, and the resulting article ranked within the top five for the primary keyword within six months.

One thing I learned the hard way: the method does not work well for highly transactional or navigational queries. If someone searches "buy running shoes," the top results are product pages, not informational articles. The word frequency will be dominated by price points, sizes, brand names, and other e-commerce signals that have nothing to do with creating useful content. In those cases, the Sixty Nine Cents Analysis gives you a distorted picture. I switch to competitor gap analysis instead. Look at what terms the top result covers that the rest do not. That is more actionable for commercial intent keywords. Another nuance that trips people up: the method assumes the top-ranking pages are semantically relevant to the keyword, which is usually true, but not always. I once ran this on a keyword where the number one result was a outdated blog post that had accumulated years of backlinks purely from a single viral share. The word frequency data from that page pulled in terms that were completely tangential to the actual topic. I caught it by spot-checking each source page before including it in the analysis. If a page ranks well but its content seems off-topic, remove it from the dataset. A single irrelevant page can drag your frequency count in the wrong direction. The tool selection matters less than people think. I have used WordStat, Keyword Insights, and even a simple Python script with NLTK tokenization. The results are roughly the same. What actually changes the outcome is how carefully you clean the input text and how thoughtfully you map the frequency data onto your outline. Rush either of those and you end up with a brief that looks data-driven but reads like aTerm stuffing exercise.

For the download link, I typically point people to the original free resources rather than gatekeeping anything. The core tooling is just a word frequency counter and a spreadsheet. There is no specialized software you need to buy. The process takes about twenty minutes for a standard ten-result analysis once you have your workflow set up. Before that, it takes longer because you are figuring out where each tool lives and how to export the data in a usable format. There is also a limit to what this method can tell you. It shows you what the current top pages have in common. It does not show you what Google will rank next year. Search algorithms change. New pages enter the results. Semantic associations shift. I run this analysis at the start of a content cycle, not as a one-time setup. Revisit it every few months if the keyword is competitive and the SERPs are volatile. If you are starting out, pick a low-competition keyword first. Something where the top results are mostly thin or poorly structured. You will see the frequency data map more cleanly onto a well-organized article. That gives you a sense of whether the method works for your niche before you apply it to a hyper-competitive term where the margin for error is much smaller.