By William R. Leake

So you’ve launched your Google AdWords campaign. Perhaps you’ve just taken it in-house, or are the new person at a marketing agency tasked with figuring all of these stuff out. You’ve got the ad copy written, the landing pages selected and built and the campaigns set up. Now, alas, you’re face to face with one of the most daunting questions a pay-per-click (PPC) search advertiser must answer, the dreaded “How much should I bid for that click?”

When dealing with hundreds, thousands, tens of thousands (or, in the case of some of our largest clients, hundreds of thousands!) of keywords, each one with potentially varying optimal bid levels, the question becomes even more complicated. But there are some straightforward concepts that can quickly help you make the right bid decisions for the keywords in your campaigns.
In a nutshell, while there are many complicated formulas in use by bid management tools or agencies, the basic formula we endorse is: the maximum bid price equals the value of an action (lead or sale) multiplied by the conversion rate of a click to the desired action.

ROI vs. Position Bidding
In the early days of pay per click, many advertisers focused on “position bidding,” which meant they were trying to control which rank they had in the paid search results. So you’d hear a lot of discussion around the topic of which position would get you more traffic or–less frequently–make you more money, and how much you’d need to bid to hold that position. Many of the automated bid management tools built their algorithms around this concept.

Google and Yahoo, however, have in many ways fundamentally altered the utility of this approach in recent years, and position bidding is no longer as desirable, easy, or–most important–profitable as it once was.

Most of the truly proficient search marketers are now using ROI bidding strategies rather than position bidding. The bid is based on a keyword’s financial return, not solely or primarily on its position. If being in a top position renders your campaigns less profitable, do you really want to be there? If you’re a large branding advertiser, focused more on reach and visibility rather than profit and revenue metrics, maybe you do. But for most of us, we are trying to make money. Therefore, ROI bidding strategies that can result in provable profits and sales will be the focus of discussion in this article.

There are two primary types of ROI bidding strategies: 1) targeting a cost per action (CPA) and 2) targeting a return on ad spend (ROAS). The following are the formulas that allow you to calculate each:
CPA is the cost divided by actions (typically used for lead gen campaigns); ROAS equals revenue divided by the ad spend or cost (typically used for e-commerce campaigns).

I’m going to assume that most marketers already know how to determine the CPA or ROAS that they want to target. But if you don’t, you can use the current average CPA or ROAS of your campaign and attempt to drive more actions or sales while maintaining your ROI ratios.

Foundational Formulas
Once you have determined your target ROI metric, you can begin to figure out how much to bid on a cost per click (CPC) basis for each of your keywords. For the rest of this article, I will use an example CPA campaign. But with some basic algebra, you should be able to apply the same principles to an ROAS campaign.

Let’s assume we are running a lead gen campaign and over the last 30 days we show an average $1 CPC and an average click-to-conversion rate (CVR) of 10 percent. Using these two variables, we can determine our CPA using the following formula:

CPA = CPC / CVR
So, in our case $1.00 CPC / 10 percent CVR = $10.00 CPA. Simple stuff, right? But when setting bids, we already know what CPA we want to hit, so this formula isn’t of much help. We need a formula that calculates a recommended CPC, not CPA. We can arrive at this by simply rearranging the same formula:

CPC = Target CPA x CVR
If we’re targeting a $10 CPA and a keyword tends to have a 10 percent conversion rate, this formula tells us that we want to pay $1 CPC to achieve our target CPA. If you can understand this basic formula, you are on your way to becoming an effective PPC search marketer. Unfortunately, this formula doesn’t answer all of our questions. There are still a few other variables that we will need to consider–such as tomorrow’s conversion rate.

Forecasting Conversion Rates
The harsh reality is that we don’t know what future conversion rates will be. But we can make a fairly accurate prediction by looking at historical conversion rate data. If conversion rates for one of your keywords have stayed between 8 and 12 percent every day for the past 60 days, you can be fairly certain that it will be within the same range tomorrow. So forecasting tomorrow’s conversion rate at 10 percent will likely get you very close to your target CPA. However, this kind of historical analysis is not going to work for all of your keywords, specifically your long-tail or low-volume keywords.

You likely have in your keyword groupings hundreds of low-volume keywords that might only get single digit click numbers each month. While they may drive very few transactions individually, they can have a very significant impact on your campaign’s performance in the aggregate. In other words, you certainly don’t want to turn them off. But that is exactly what will happen to many of them if you base your CPC bids on recent conversion rates. If a specific keyword has received two clicks and no conversions over the last 30 days, the keyword will have a 0 percent conversion rate. Using the formula discussed in the previous section, this would lead you to a $0.00 CPC bid.

To resolve this problem, you need to manage your long-tail keywords differently. A common solution is to use the average conversion rate for all keywords in a campaign or ad group to calculate the CPC bid for all keywords under a specified click threshold. So, if your entire campaign shows a 10 percent conversion rate, you might want to apply this average conversion rate to all keywords with less than 50 clicks in the last 30 days.

Other Variables
While the above strategies and formulas will give you a good start to making effective bidding decisions, there are still many other situations that will require you to depart from basic bidding strategies, such as the following wrinkles:

When forecasting future conversion rates, do you want to base estimates on yesterday’s conversion rates? Last month’s conversion rates? Last year’s? Each timeframe will likely give you a different conversion rate estimate, and not all timeframes will accurately predict future performance.

When placing a bid with a search engine, you define the maximum CPC you are willing to pay. This does not mean that you will actually pay that amount. The nature of bid gaps between ad positions may make it impossible to pay the exact CPC necessary to achieve your target CPA or ROAS.

Do your conversion rates tend to fluctuate predictably depending on the time of day, day of week, or season of year? While it’s not always an easy process to analyze such trends, they can help you achieve even greater bidding efficiencies in your campaigns.

Pay attention to both recent trends and past history. Looking at a keyword over long periods of time may give you more data and a more accurate estimate, but market changes and seasonality can confuse the issue. If the keyword or ad group is affected by market changes, focus on more recent history and trends. Use anticipatory bidding with upcoming sales, seasonality or product releases.

If your best traffic comes at a certain time of the day, use Google’s built-in tool to raise or lower bids during different times of the day or days of the week. For example, if people tend to research and buy your products during the work week, you might find that on Wednesdays you want your bids to be 20 percent higher and on the weekends, you want bids to be 50 percent lower. Be sure to continually test your assumptions. And remember, visitors may take hours or days between the first click and the lead or purchase, so be sure to tie your sales and leads back to the time of the click; don’t simply focus on the time and day of the sale or lead.

The good news is that–despite the mathematical formulas set forth above–PPC search marketing is not rocket science, so don’t be intimidated. Any diligent online marketer should be able to apply some logical bidding strategies and achieve success in their search marketing campaigns. But it will take time and diligence. Pay close attention to your conversion rates and don’t get emotional about the importance of low-performing keywords. And be sure to pay attention to more than just bids: If a keyword isn’t performing, you might need to adjust ad copy, the landing page, match type, or negative keywords, rather than simply pull the bid lever.

William R. Leake is the founder, president and CEO of Apogee Search, an Austin, Tx.-based search marketing solutions firm. He can be reached via e-mail at leake@apogee-search.com.