Ever wanted to compare how your website ranks in comparison to a competitor’s? There are lots of tools that show overall visibility (think SEMrush), but very few that show the more granular data at the level of keyword ranking positions. So where do you start?

This is what we aim to tackle today: how you can gather keyword data and use it to produce intuitive graphs which help you compare keyword ranking positions for two (or more) domains. There are four main graphs we will look at: Empirical Cumulative Distribution Function, Histogram, Box Plot and Stacked Bar Chart.

To highlight how you could be using data and graphs in this way, we’ve used the example of retailers selling trainers and analysed two very different domains ranking in this space – Sports Direct and Next.

Before diving into the graphs, it’s worth noting a couple of things on how we gathered the data for the example we have used:

  • We started with basic keyword research for ‘trainers’ in the UK market and took 1,000 of the highest volume search terms.
  • We then scraped Google’ search engine results pages to see who is ranking for these terms, up to position 20.
  • All analysis is based on our sample data, and not a total reflection of how Sports Direct and Next are ranking overall.

To help demonstrate how useful the below graphs can be, we have also made them dynamic, so you can input some of your own data and see how it changes the graphs. Click the button ‘Try with another data set’ to use your own data.


Graphs have been recreated – view below!



ECDF (Empirical Cumulative Distribution Function)

An ECDF graph allows you to plot an entire feature of your dataset on a single graph. In essence, it is a scatter plot (the dots are so tightly packed it creates a line) with values sorted from lowest to highest and plotted as a cumulative percentage on the Y axis. That sounds very complicated, but in reality it is very easy to interpret.

What can we take away from the below ECDF graph?

  • 60% of Sports Direct’s rankings for ‘trainers’ are in position 6 or lower. 60% of Next’s rankings are in position 9 or lower.
  • 80% of Sports Direct’s rankings for ‘trainers’ are in position 11 of lower. 60% of Next’s rankings are in position 13 or lower.
  • 100% of Sports Direct’s rankings for ‘trainers’ are in position 20 or lower. 100% of Next’s rankings are in position 20 or lower.

What does that mean?

Sports Direct are ranking better than Next for ‘trainers’ – with more of their pages showing up in lower positions in search engine results pages.

Histogram

A histogram buckets numbers together (often within a range) and then graphs as a bar chart showing the frequency of each bucket. In our case, we have bucketed keywords into groups between position 1 and 20, as we have collected rankings up to position 20..

What can we take away from the below histogram?

  • Sports Direct have 109 keywords ranking in position 1. Next have 71 keywords ranking in position 1.
  • Sports Direct have 54 keywords ranking in position 5. Next have 48 keywords ranking in position 5.

What does that mean?

Sports Direct have more keywords ranking in lower (better) positions than Next.

Box Plot

A box plot shows data distribution in a box shape. The entire line is the range, the middle line is the median, and the lines above and below the median line are one quarter and three quarters of the data – otherwise known as the interquartile range. The size of the box shows how close the data is to the median value. So, a small box would show rankings are all close to the median value, and a long elongated box would show that there a lots of rankings further away from the median value.

What can we take away from the below box plot?

  • Sports Direct ranking distribution range is between 0 – 20. 1/4 of the rankings are in position 2 or lower, the most frequent ranking position (median) is 4, and 3/4 of rankings are in position 10 or lower.
  • Next ranking distribution range is between 0 – 20. 1/4 of the rankings are in position 3 or lower, the most frequent ranking position (median) is 7, and 3/4 of rankings are in position 12 or lower.

What does that mean?

Sports Direct have more keywords rankings in lower (better) positions than Next.

Stacked Bar Chart

A stacked bar chart shows two datasets next to each other – the bar is broken into buckets based upon the count within the respective bucket. For example, we have created buckets based upon rankings positions between 0 to 3, 3 to 5, 5 to 10, 10 to 20, 20 to 50 and 50 +.

What can we take away from the below stacked bar chart?

  • Sports Direct have 284 keywords ranking in position 3 or lower. Next have 163 keywords ranking in position 3 or lower.
  • Sports Direct have a total of approximately 700 keywords rankings in our sample. Next have a total of approximately 630 ranking keywords.

What does this mean?

  • Sports Direct have more pages ranking in our sample than Next – Sports Direct have approximately 700 and Next have about approximately 630.
  • Sports Direct have more pages ranking in lower (better) positions than Next.

Ranking Distributions: A conclusion

Looking at keyword data distributions using graphs like these allows you to quickly see how domains compete in different spaces – i.e. ‘trainers’ or ‘children’s shoes’. However, there are a few caveats and additional things to consider:

  • These distributions do not take into account the search volume of keywords. For example, could Next actually get more traffic if they rank in position 1 for higher volume search terms?
  • It is a static view – there is no time dimension (although you could compare a single domain against itself to see how the distribution has changed over time).

When are ranking distributions most useful?

We find looking at ranking distributions most valuable when comparing two domains alongside different search themes. If we were to take our example further, we might look at how Next competes with Sports Direct for ‘children’s shoes’ or ‘boots’, rather than ‘trainers’. By looking at the graphs we would gather a very quick view of who is winning in each space. This understanding would then open the door for us to see if there is something one domain is doing to achieve a win (such as having more product listing pages) and what the other domain can do to better compete.

Can we help you?

Interested in knowing how well your site is ranking compared to your competitors? At Melt Digital we are specialists in SEO and can help you gather and analyse the data you need to find the answers. Get in touch today and see what we can do for you.
Get in touch by calling 0203 735 5070 or by emailing info@meltdigital.com.