When beginning online marketing initiatives, sometimes one needs to better understand the competitive landscape one is was entering. It would be helpful to know who my competitors were, how they ranked in terms of online visibility and income, how big they were, and which marketing channels they utilized most. Competition being as it is however, this information isn’t always easy to home by. Fortunately, there is a lot of information to be found through various resources, if you know where to look. My goal with this post, is to share methods I’ve figured out along the way.
Imagine for a moment, that there is an unknown object inside of a box. You cannot open the box but there are various holes in the sides of the box that you can look through. If you look through just one hole, you may get some sense of what it is, or you might now. But as you start to systematically review all of the holes in the box, you’re able to interpolate a better approximation of what that mysterious object probably is. So to is my approach to figuring out these companies.
Exposing Key Terms
The first step ideally would be to use a tool such as Google Keyword suggest, WordTracker, or KeywordDiscovery to identify the primary “head terms” for your specific vertical. By head terms, I mean the 10% of key phrases that get 90% of traffic. Each of these tools also provides a competition metric that can be useful. I personally like WordTracker and the use a competition metric called KEI that compares total search volume compared to total number of competing documents for those terms. For this exercise, I’d start by filtering key phrases based on volume, and then further refine until I found the top 10 key phrases. Note these down, along with the KEI index.
Next, you’re going to go to Google and search these phrases that you looked up, noting all the competitors you can find. Depending how deep you want to go, consider how many you want to take note of. I personally noted around 50 competitors, with a goal of filtering them into four buckets ultimately – tier 1-5 competition. I use a spreadsheet to track all the various metrics I compiled for each company I identified.
Search Engine Results
One possible predictor of a popular or well-established company, is how well they rank in the search results. Google in particular has begun showing bias toward large established brands, as well their inbound link relevancy algorithm, naturally favors large brands. So take note of how someone ranks for this top terms and assign them an overall grade of A-F. Take particular note of how well they rank on the high KEI index (competitive) terms. You can also use sites such as OpenSiteExplorer.org to expose more quantified data such as MozRank and domain rank, which can be useful. I also find Google’s PageRank very useful and track that as a separate metric as well. Page Rank provides a 1-10 ranking of how substantial the inbound linking profile is for a give page and can be easily seen by installing the Quirk SearchStatus plugin for Firefox.
Competitive Ranking Tools
There is a set of tools that can be particularly useful for this exercise that attempt to track historic traffic for sites. Compete.com provides the last year of data for free (or 2 years for paid members) and allows you to compare up to 5 competitors on a multi-line graph, to see who’s truly getting traffic. Alexa also provides a similar service and historic graphing but goes an additional step of providing a ranking number for each site; #1 is the most popular site on the Interent whereas #100,000 considerably less popular. Both of these tools work in part by providing a tool bar users can install on their browsers to provide extended functionality and data while surfing. In exchange these companies get a sample set of data from which they can extrapolate general behaviors. It is worth noting however that due to the nature of data collection, these data sources (Alexa in particular) can tend to skew toward demographics that would be more male and early technology adopters. Assuming you are simply comparing businesses within a specific niche however, this shouldn’t be much of an issue. Quantcast is another tool of similar means (different data collection means) that can also be used. I would generally recommend getting data from all 3 of these as none of these is totally reliable. If you can triangulate a rough idea between the three however, you’ll be more assured of accurate data.
Alexa Traffic Approximation
Staying with Alexa for a moment, a few years ago, I found a wonderful graph that has been invaluable and proven to be generally accurate over the years. You can see from this scatter chart that numerous people have indicated their Alexa ranking and corresponding traffic. There are of course outliers but there is also a general trend line that begins to emerge. Thus you can take an Alexa ranking for a particular site, apply it to a site, and get a rough sense of how much traffic they might be *actually* getting! I would track this as a separate metric.
Are your competitors spending money on AdWords? Depending which vertical you are in, this could be more or less significant. If you are in B2C eRetail however, this could be substantial. There are amazing tools that have emerged in recent years that can infer and approximate how well someone is doing on AdWords and what their estimated daily spend is. So, if PPC is a viable marketing channel, this can be a very strong indicator, how well your competitors are doing. As for specific tools – I personally like KeywordSpy.com. KWS crawls Google SERPs and notes the PPC ads shown. The more frequently the ads appear as the bot crawls for that keyword, and the higher the ad position is each time, the higher the relative ranking that brand gets for that keyword. KWS then compares that to the keyword bid data to infer roughly how much money someone is paying per click, and collectively, their total daily spend. They also infer how well someone might be doing in terms of ROI, by looking at how consistently that brand has been present in their crawls. Someone who consistently ranks for a specific keyword must presumably be doing well. Pretty clever huh?
Another interesting metric can be found by looking at how well the company is represented socially. I would NOT simply look at Facebook followers as this is much more reflective of social activity than brand value. Also, it is very easy to manipulate this metric by simply ‘buying’ friends or advertising a Facebook fan page, for example. Instead, look at how many social mentions a brand is getting. For example, are people sharing stories, liking artless, or retweeting their message? OpenSiteExplorer.org has an excellent tool for assessing social mentions on the important channels, but requires a paid SEOMoz subscription to access that data. Klout.com also has a rather interesting and more in depth way of assessing a brands social value and each brand is thus assigned a 1-100 Klout score. At a high level, this alone might be enough to gauge social presence of a brand, though that OSE social mention data is ideal.
Many brands have an affiliate program, enabling independent agents or resellers to earn commission by directing sales to the brand. If a company has an affiliate program, it may be an indication of a more mature or established online brand. More importantly, if you sign up for the affiliate program and drive a little traffic to it, you can begin to see how well their products sell, what customer acquisition costs are, etc. If you are getting a 20% commission split as an affiliate, just multiple x5 to see exactly what their revenues are, and you can basically reverse engineer the entire cost and profit models of the business!
Finally, there are a couple of resources I’ve found that are particularly useful for fleshing out what the business must look like offline, in the real world. Jigsaw.com is a business intelligence resource owned by SalesForce.com that will tell you how many employees are known at the C** level, how many directors, sales, etc. I’m not sure the inner-working but this must be pulling from SalesForce.com as anonymous data shared by their customers. They use this data to infer the total employees of the company, and provide that and revenue as a possible range. Sometimes these ranges are a little too broad to be useful but its a good base to build on.
Next, I’d look for quarterly filings for any of these companies that happen to be public. You can get very precise data on head count, profits, losses, and margins if so. In my experience however, the companies I’m profiling rarely are already public. More likely, I’m profiling younger companies in up and coming sectors, in which case there’s likely to be a couple that have been recently profiled by Inc5000. If so, they’ll tell you exactly how much revenue, how many employees. CrunchBase is another useful tool for determining if any of these companies has raised funding along the way, and how much.
By compiling base ranges from Jigsaw and hopefully filling in exact numbers on 10-20% of your companies via other means, as well as having compiled all the above data, you should be able to extrapolate approximately what the revenues, costs, and size are for nearly every business on your spreadsheet. And now you know who’s who and can split them into tiers, and based upon your own resources and history, you know which tier you belong in, and are targeting. You now know precisely who your competitors are.