|
When preparing to begin a pay per click marketing campaign, it really helps to go into the situation with an idea of what the numbers look like. By "the numbers", I mean what kind of return on your investment you can expect assuming different bid levels and different conversion rates. It's really all just conjecture because you don't know what your click-through rate will be, what your bids will have to be to generate adequate traffic, or what your conversion rate will be. But you don't want to go into the situation blind. So the following technique is a way for you to at least apply some logic to the situation, rather than just guessing.
ROAS ("Return on Ad Spend") is defined as the total dollars generated by the advertising divided by the cost of the advertising. The following discussion and data illustrate an ROAS-based approach to setting a baseline bid level for your campaign. Every ad group and possibly every keyword will have its own bid once you actually set the ad groups up. But this type of analysis will give you an idea of what your average click price needs to be in order to generate a positive return on investment, assuming various conversion rates.
Below are two examples of the kind of analysis we often do when beginning a new campaign to give the client and ourselves an idea of what he might expect his return on ad spend to be. It also helps us establish a benchmark average bid.
The first thing you're going to need is some data to plug into your spreadsheet. So log into your Google, Yahoo!, or MSN account and create a new campaign or ad group. When you get to the keywords section, add a large group of keywords related to your industry. You are not going to actually use this campaign; you just want to get an idea of how much traffic there is for the keywords and how much you might have to pay for clicks. Going through the process of setting up a campaign lets you generate some data to work with.
The data for the examples below was generated in the Yahoo! Search Marketing interface. Yahoo! is convenient for this kind of analysis because of its sliding scale bid tool that gives immediate feedback about expected traffic and costs at different bid levels by just moving the slider around. We used a single broad set of related keywords to generate the necessary data. In reality, your keywords will be divided up into logical groupings, or ad groups, but for analytical purposes a single ad group was sufficient.
For our first analysis, we are going to assume a bid level of $.75, which Yahoo! estimated would result in 21,993 monthly clicks at a cost per click of $.50. For illustration purposes (and to make the math easier), we assume average revenue per sale of $100.
We know that our total cost is going to be $10,997 (21,993 clicks X $.50), so we can use this information to estimate what our total sales, and thus return on ad spend, will be at different conversion rates:
Bid: $.75
Est. Monthly Clicks: 21,993
Avg. Cost per Click: $.50
Total Cost: $10,997
Avg. Revenue per Sale: $100.00
Performance per Conversion Rate
Conversion Rate: 0.25%
Estimated Sales: 55
Revenue: $5,498.25
Return on Ad Spend: 50%
Conversion Rate: 0.50%
Estimated Sales: 110
Revenue: $10,996.50
Return on Ad Spend: 100%
Conversion Rate: 1.0%
Estimated Sales: 219.9
Revenue: $21,993.00
Return on Ad Spend: 200%
Conversion Rate: 2.0%
Estimated Sales: 439.9
Revenue: $43,986.00
Return on Ad Spend: 400%
Conversion Rate: 3.0%
Estimated Sales: 659.8
Revenue: $65,979.00
Return on Ad Spend: 600%
As you can see, if we convert at 0.25% (1 in every 400), we are going to generate a return on ad spend of 50%, meaning we are bringing in fifty cents for every dollar we spend. That is not very good. We probably need to convert at a rate between 1% and
2% to show a reasonable return on our investment.
Now we might want to see what the numbers look like if we lower our bid. If we bring our bid down to $.50, Yahoo! estimates our estimated monthly clicks to be 14,948 at a cost per click of $.38. So now the return on ad spend analysis looks like this:
Bid: $.50
Est. Monthly Clicks: 14,948
Avg. Cost per Click: $.38
Total Cost: $5,680
Avg. Revenue per Sale: $100.00
Performance per Conversion Rate
Conversion Rate: 0.25%
Estimated Sales: 37.4
Revenue: $3,737.00
Return on Ad Spend: 66%
Conversion Rate: 0.50%
Estimated Sales: 74.7
Revenue: $7,474.00
Return on Ad Spend: 132%
Conversion Rate: 1.0%
Estimated Sales: 149.5
Revenue: $14,948.00
Return on Ad Spend: 263%
Conversion Rate: 2.0%
Estimated Sales: 299.0
Revenue: $29,896.00
Return on Ad Spend: 526%
Conversion Rate: 3.0%
Estimated Sales: 448.4
Revenue: $44,844.00
Return on Ad Spend: 789%
We can generate a higher return on ad spend at the lower bid. However, we will generate less total revenue, and may convert at a lower rate. In addition, we have found that the accuracy of Yahoo!'s estimated traffic lessens as bids decline. But if you are on a very limited budget in an industry with plenty of keyword inventory, as in our example, then you can probably get away with bidding low and still generating a lot of clicks.
You could take the analysis a step further and look at it in terms of profit, rather than revenue. For instance, if the above estimated revenue per sale of $100 translated into $50 profit per sale, you could estimate your return on investment at different conversion rates by dividing the return on ad spend in half. If you substitute profit per sale for revenue per sale in the above analyses, then you will generate your ROI rather than ROAS.
One caveat about setting your initial bid this way: due to the quality scores that both Yahoo! and Google apply to keywords and ads, it may be necessary to begin the campaign at a considerably higher bid than your research indicates in order to generate a higher click-through ratio, which plays a part in determining the quality score. Once the ads are generating clicks, you can back off the bids more in-line with your desired ROI.
Another caveat is that oftentimes a search engine's estimated traffic at different bid levels is not very accurate. It's hard to fault the search engines for this, when there are so many variables that can affect how much traffic a campaign will generate. So your actual performance and required bids will probably vary quite a bit from what the chart shows. But the above described method for setting an initial bid at least gives you a logical starting point. Once the campaign is running, you can make adjustments as necessary.
The question of how much to bid may also be determined by your budget for the campaign. At higher bids, you are going to burn through your budget quicker. If there is so much keyword inventory related to your business that you are able to use up your entire budget almost regardless of what you bid, then it might make sense to bid lower...if ROAS is your main consideration.
If there is a branding component to your online marketing, then you may want to bid higher for higher positioning on the page. Another consideration is that not all sales are made immediately. If you position yourself as the leader in your category (high on the page), you generate more immediate traffic and more potential future business from prospects who visit your page but don't immediately do business with you. This branding component is not reflected in the kind of analysis we displayed above.
An analysis like this can easily be prepared using a spreadsheet, and we highly advise you to do something similar to get a feel for your own numbers.
|
Rate This Article
Add Comments
Send To Friends
Print 
Download as PDF