Whether you’re building a new site or optimizing an existing one, you should be aware of what your competitors are doing in terms of their websites’ content and optimization.

Why?

Because you are competing with them for the same visitors. In the new era of OC/DC, it’s all about content, and how we optimize its distribution and conversion. Winning at OC/DC is extremely important as it results in both capturing more web traffic and converting it at a better rate. It can also be measured in very tangible and direct ways (revenue from new sales, new users, advertising revenue, etc.).

Understanding your competitor’s content can help you win at OC/DC by creating (or refreshing) your content so that it is stronger than theirs. Here’s a few tactics to get there, in no particular order.

  • Eclipse: the idea here is to build superior content that “eclipses” each of your competitor’s most important content pieces. There are multiple ways to measure what’s important, but that’s for another post. Focus on making your content demonstrably better than theirs (tighter headlines, smoother grammar, easier to read, more frequently updated, better designed, you get the idea). Copyblogger is rich with ideas on how to do this.
  • Differentiate: create content that covers topics they don’t. Build your long-tail superset. In other words, zag.
  • Never Sleep: keep adding and refreshing your content so you’re a moving target and not easy prey. This will make both of the above tactics hard for them.

Headlines

Since most content is usually summarized through headlines, there is no doubt of their importance. They contribute significantly to engagement and conversion. Researching your competitor’s headlines can be of huge help to aid in understanding their overall brand positioning and content strategy.

Page Titles, Meta Descriptions, and URLs

It’s also useful to research how your competitors are using page titles, meta descriptions, and URLs to position themselves within search engine results. Not only is your search engine result position important, what they look like side-by-side is also meaningful. In other words, optimizing how your search results are displayed may give you an edge and win over visitors, even if even if yours is ranked slightly lower than them. Ever clicked on that third or fourth search result over the first or second because it felt more relevant due to a superior title? It’s also no secret that these tags are also used by search engines to calculate relevancy between search queries and page content.

The case study & framework

Here’s a case study in how to analyze your competitors’ content. You can follow the same methodology used here as a framework to study other markets or categories. When doing so, keep in mind that you may have to adjust things for your particular case.

In our case we’ll explore three Austin car dealerships that sell cars from the same manufacturer. They are competing for the same traffic and there is significant money on the line.

To study the competitive landscape at a content-level, you may also want to develop a simple taxonomy that enables you to categorize content for easier analysis. For local car dealerships, this could be something like: brand, car models, inventory, corporate, local, and other. While we won’t go into that level of detail here, keep this in mind as you analyze your own case. The idea is to look at the quantity of content that each of the competitors has for each taxonomy. This should reveal their weak and strong points.

The research

To get the raw and aggregated Headlines, URLs, Titles, and Meta Descriptions for the three car dealership sites, I used SiteCondor‘s* API and put together a few scripts that perform further content analysis. The scripts are open source and included in the github repository for our SiteCondor Ruby client.

* Full-disclosure: I’m a co-founder of SiteCondor.

You can also use SiteCondor to download CSVs, so if you’re not technically inclined this study is still feasible without having to connect with the API and do any programming. The API just gives you more flexibility and enables you to both A) automate the process and framework for any sites you’d like to analyze and B) integrate it with your own tools or products.

SiteCondor not only extracts these elements, but it also aggregates them. For a given unique title for example “Honda Civic 2014,” it will give you all of the URLs for pages that contain that title. SiteCondor also sorts the results so that unique Titles, Meta Descriptions, and Headings used the most are shown first. URL results are not aggregated as they are intrinsically unique.

Below is a screenshot for Titles used by one of the dealerships:

SiteCondor-Titles-Delearship

This quickly reveals this particular dealership is primarily targeting specific car models for sale. They are also appending the location, followed by their brand.

You can easily repeat this analysis for each element and dealership to draw some high-level conclusions. You should look at the volume each dealership has (number of pages, unique title count, etc).

While analyzing things at this level is useful, we wanted to go deeper and analyze things at the keyword level as well. This is necessary for large sites, as the long-tail of different titles could be of significant size.

For example, a given site’s most used unique title may be used 50 times and contain no brand-related keyword, but the same site may have 5,000 other pages with a unique titles that contain the brand-related keyword. Performing a keyword-level analysis across entire sites can reveal these strategies. To do so we programmed the scripts so that for each of the elements (Headings, URLs, Titles, and Meta Descriptions) they would:

1) tokenize the values into keywords, remove stop words and unnecessary symbols
2) analyze the frequency of those keywords across the entire site, taking into account how many times each unique value was repeated and weighting them accordingly (e.g.: if “Honda Civic” was used in 20 pages, that title would increment both the “Honda” and “Civic” keyword frequency count by 20. If another title used was “Honda 2014″ and found on 5 different pages, it would make “Honda” frequency count 25 and add a new keyword of “2014” with a frequency count of 5.

[Note: We coded the scripts so they can stem the keywords but decided not to do that for this post to keep things simple. Additionally, while the scripts gave us full results for the whole sites, we limited the results displayed below to "top 20" keywords in this article, also for simplicity's sake.]

The script also analyzes the length distribution for each of these elements across each entire site , also taking into account how many times each of those unique elements was encountered.

Here are the results:

Headings

The table below shows the top 20 keywords used on H1 Headings by each of the dealerships. Weights are relative for each of the dealerships and calculated as the frequency of the keyword divided by the frequency of the most used keyword.

SiteCondor-Dealerships-H1-table

Here’s a few takeaways from the table above:

  • Round taxonomy priorities: Brand, Competitor Brand, Models, Location
  • First taxonomy priorities: Brand, Competitor Brand, Location, Models
  • Howdy taxonomy priorities: Brand, Year (Model), Location, Model, Competitor Brand
  • In terms of volume for most used keywords, Round was a clear leader, followed by First and then by Howdy

The screenshot below shows the top 20 keyword frequency distribution.

SiteCondor-Dealerships-H1-chart

All three dealerships have a similar distribution shape. However, the chart makes it very clear Round and Howdy are heavily leading in top keyword volume.

The chart below shows H1 headings length histogram for all headings.

SiteCondor-Dealerships-H1-length

Quick takeaways:

  • First tends to use longer headings, while Howdy has a slightly shorter length distribution, and Round uses shorter headings.
  • None of them had missing headings, indicating that these are not amateur sites.

URLs

Top 20 keywords used on URLs by each of the dealerships:

SiteCondor-Dealerships-URL-table

Takeaways:

  • Round taxonomy priorities: Brand, Inventory, Location, Other Brands
  • First taxonomy priorities: Inventory, Brand, Model, Other Brands
  • Howdy taxonomy priorities: Brand, Location, Model, Inventory
  • In terms of volume for most used keywords, the order is again the same: Round, Howdy, First.

URLs keyword frequency distribution (for top 20 keywords):

SiteCondor-Dealerships-URL-chart

Here again the chart shows the volume difference between each dealership site.

URL length histogram:

SiteCondor-Dealerships-URL-length

Round had the longest URLs (some probably too long).

Titles

Top 20 keywords used in Titles by each of the dealerships:

SiteCondor-Dealerships-Title-table

Takeaways:

  • Round taxonomy priorities: Brand, Location, Location, Other Brands
  • First taxonomy priorities: Location, Brand, Inventory, Model
  • Howdy taxonomy priorities: Brand, Location, Model
  • In terms of volume for most used keywords, the order is again the same: Round, Howdy, First.

Please keep in mind this is just quick taxonomy priority analysis for illustrative purposes. For a more detailed analysis, we suggest you add up the frequencies of each keyword in a taxonomy group. For example, looking at Round, adding up “austin,” “tx,” “round,” “rock,” “cedar,” “park,” “georgetown,” and “serving” keyword frequencies would show how they are actually prioritizing Location in terms of volume, whereas their top most used keyword is Brand related.

Title keyword frequency distribution (for top 20 keywords):

SiteCondor-Dealerships-Title-chart

Once again the chart shows differences in volume.

Title length histogram:

SiteCondor-Dealerships-Title-length

Round is using longer titles than First, whereas Howdy is using significantly shorter titles. None of the sites had missing titles or titles too short.

Meta Descriptions

As the patterns repeat, we will leave it as an exercise to the reader to read through the charts.

Top 20 keywords used on Meta Descritions by each of the dealerships:

SiteCondor-Dealerships-MetaDescription-table

Meta Description keyword frequency distribution (for top 20 keywords):

SiteCondor-Dealerships-MetaDescription-chart

Meta Description length distribution:

SiteCondor-Dealerships-MetaDescription-length

Conclusion

This article doesn’t attempt to deliver a comprehensive content competitive analysis for the three dealerships, so we won’t focus on drawing conclusions from the case itself. Instead, we would like to point out the following:

  • Performing competitive content analysis on Headings, URLs, Titles, and Meta Descriptions can help you devise a content strategy that wins at OC/DC
  • You can use SiteCondor to extract the raw and aggregated data needed for this type of analysis
  • You can use this framework as a jumping off point and tailor it for your own needs

We hope you’ve found this template useful. For completeness, we would like to leave you with a few more thoughts:

  • Remember to not over-optimize your content for search engines. Those days are long over and if you do that you will get penalized. Leave the black hat tricks for the circus.
  • Instead, optimize your content for discovery and conversion by human beings.
  • Use this technique to find your competitor’s weak areas, and go after those first (e.g.: if most competitors are weak on a given taxonomy that is interesting to your audience, target that one first).
  • Not comprehensive: while this kind of research can yield a lot of insight, don’t think about it as an all-encompassing competitive analysis. Mix and match it with other tools, data, and processes to complete the picture. SiteCondor job results are a great way to achieve this at the on-page analysis level, and there are many other tools out there that can help better understand your competitor’s distribution strategy with regards to social media, backlinks, rankings, and more.

Good luck!

About the Author

Sebastián Brocher is the co-founder of SiteCondor, a site analysis tool for digital marketing experts.

In addition to founding several startups, he holds a degree in Computer Engineering from the Buenos Aires Institute of Technology (ITBA), and is fluent in Spanish, French, and English.

 

Sebastian