How To Measure Your Competitor's AdWords Budget And Estimate Transaction Costs

This method only works for competitors who have the same stock of goods and it only gives you a rough estimate. If that's so, you're probably thinking, what use does this method have in the first place? Well, I suppose you won't mind finding out some data about your competitors' income and spendings with no effort at all.

First thing you need to do, is to find out how many monthly visits your competitors get. I used a random website for this case, which happened to be sunglasses-shop.co.uk. I usually use Similar Web's Chrome Extension to count the traffic. To get more data about the website visit Similar Web:

Of course, talking about online store that sells mainly sunglasses, we should take seasonality into account. The number of visits varies from 521 000 in July to 156 000 in November. In December we see some rise because this store sells goggles as well. The avarage number of visits is nearly 339 000/month. I'll stick with it for the sake of the experiment.

Count the number of sales your competitor makes. Most e-commerce websites have sequential order numbers.

1. Place an order #1. If your order number is 003149, that would be 3149th sale on this website.
2. Place order #2 a week later. Say this time order number is (003281).
3. Count how many orders the website processes every month. Simple math is all you need: 3281-3149=132 so 132 orders per week, multiply by 4 to get number of monthly orders (132*4) ~528 orders per month.

Sometimes people place an order and don't proceed with payment or return the item. Number of such cases depends on a particular store, of course, but rarely goes over 10%. I'll take 12 orders per week (it's about 9%) to account for unpaid, refused and returned orders, and what's more important — to get another round number. Now that we know that website has 480 monthly orders and monthly traffic of 339 000 visits.

This is the part where your experience kicks in. If you're analyzing your direct competitor, your average sale should not be too different. I don't sell sunglasses, so I can only assume that average sale for sunglasses-shop.co.uk is £160 or ~$200. 480*200 =96 000, so our competitor sells $96 000 worth of goods every month.

Time to find out this website's AdWords monthly budget. It will take some time, but there's nothing to it.

• Enter the domain into Serpstat's search bar and go to "PPC Research" and then click on "Keywords." All keywords that the queried domain bids on in Google AdWords are displayed here. Now you need to export this list. Spreadsheets are best suited for calculations, so that choose between Excel and Google spreadsheets. I prefer Google Spreadsheets so that I'll use it for calculation.
  

• After exporting, you'll get the spreadsheet with lots of different columns, but we need just some of them:

1) Position (Pos in a spreadsheet)
2) The number of queries
3) CPC (cost per click)

If you want, you can also leave the columns with the domain, not to forget which competitor you're analyzing and the keywords to see the correlation between the keyword, and it's monthly budget. After deleting all irrelevant columns, you'll get something like this:

To estimate competitor's budget, we need one extra metric – CTR. As we all know the ad position effects on CTR: the higher the position is, the better its CTR is. Based on our experience and different researches, we created "CTR by position" table. Here it is:

Add these values to your spreadsheet somewhere. Now we have all data required for calculation. First, we need to count the budget for each keyword, here what we have to do:

We're almost there. To get the total budget of your competitor, just sum the budgets of all keywords. Here's what we finally get:

So that the total budget of this website is almost $13 000, which makes up about 13% of their sales number. Compare these numbers to your own data and see if your competitors are more efficient. If they are, I recommend running a domain analysis and looking up the keywords they use in PPC. We have a detailed manual on this, by the way.

And don't forget to share your experience in comments, if you give this method a try!