Make Sense of the Metrics Morass
Clear objectives, effective measurement, and appropriate metrics are critically important for all marketing efforts. Search engine marketers have additional campaign metrics challenges because of the nature of our industry. Typical search engine marketing (SEM) campaigns involve tracking hundreds or thousands of keywords simultaneously, post-click behavior changes based on position, and multiple engines with different syndication networks. Marketers have learned that not all search-result click-throughs are created equal. Some visitors find what they are looking for while others just leave.
The challenge is to translate the site’s business objectives into campaign objectives based on the right metrics. Metrics help you get the most out of your campaign, so make sure you understand how to pick those that fit your needs. The most common SEM metrics are: cost per order (CPO), cost per action (CPA), return on advertising spending (ROAS), and return on investment (ROI). In the interests of sharing best practices, let’s review the common metrics, how they interrelate, and where they come from.
Offline media, such as direct response television (DRTV), direct mail, and print advertising, use the CPO metric regularly, and online banner campaigns are often judged based on CPO. If you sell something online, then you might want to measure CPO and use the data to manage your campaign. You can measure CPO for a whole campaign or for individual keywords and engines. CPO targets can be set on a campaignwide basis or even on a much more granular level, including by product or keyword.
Some marketers have a hard time setting a CPO target. It’s not always easy. Since an order has an immediate profit value and an impact on the lifetime estimated value for a customer, you can use both the immediate profit and long-term value in setting your CPOs.
If you need a justification to set a CPO target higher, consider adding intangible values to your formula, such as brand building (through user experience of the brand/site) or gain of market share. In addition, if you know you often get phone or mail orders generated by your online efforts but not captured, adjust the measured CPO targets accordingly. By adding an adjustment factor to the measured CPO, the campaign can be managed closest to the actual CPO.
Orders are great as a metric, but for some businesses tracking orders online is difficult or even impossible. This may be because the business is a service, the item is not appropriate for online purchase, it’s a high-involvement purchase, or the sales process for that product or service is complex. In that case, to differentiate between good and poor traffic, the campaign metric must be an action that correlates to a future sale.
Actions that marketers may measure to determine traffic quality typically include registering for a newsletter, requesting information, downloading a white paper, reaching a particular important page within the site, or visiting the “contact us” page. To select actions to measure, understand that those actions are proxies for what you’re really after—revenue. Think about how your sales and revenue generation process works. If you have more than one action that contributes to the likelihood of a sale, assign a weight to each action relative to the others.
Measuring the ROAS is an appealing concept. It’s the amount of revenue generated per dollar spent on advertising. For example, a ROAS of $1 means you’re generating $1 in sales for every $1 in advertising spend, and a ROAS of $10 means you generate $10 in sales for every $1 in spending. ROAS is a useful metric, but only for some marketers. If your profit margin is the same or very similar across products, the ROAS can be quite appropriate as a campaign metric.
With a predictable profit margin, it’s easy to set a target ROAS. For example, if your profit margin is 25 percent and your ROAS is $4, then you are breaking even (spending all your profit on marketing spend). If your ROAS is $8, then half of your profit goes toward customer acquisition.
Of course, ROAS does not take into account lifetime value; it simply looks at the return for the initial order. Also, when using ROAS and all these metrics, keep in mind that gross profit doesn’t take into account your overhead, allowances for returns, bad debt, and so on.
ROAS starts to break down when there is a large variation in profit margin on products or if some products tend to predict a higher lifetime value. For example, an online office supply store may have a lower margin on office phones than on binders, but customers who buy binders may be more likely to order again. Neither factor is taken into account with ROAS.
It would be hard to find a more overused word in marketing than ROI. True, ROI is difficult to accurately measure and calculate. Most marketers use gross profit when calculating ROI, and that is a great start. If you really want to know the long-term profit generated by an investment in advertising, marketing, or any other activity, you need to capture all sorts of information that is more difficult to find.
The slipperiest of variables is lifetime value, which is how much profit a particular customer or type of customer contributes to the bottom line over time. Then there are factors that could influence profit negatively, which sometimes get left out of a lifetime value formula that looks at revenue, such as how often that customer calls the tech support line or how many returns are made.
If you don’t have access to more sophisticated data, leave out complications such as lifetime value, and use the immediate ROI based on the purchase transaction that occurs online (or offline). For example, if you spend $100 advertising an electric stapler and sell $800 worth of staplers, on which you make $300, your ROI is 300 percent. ROI is often the metric used in planning when setting CPO and/or CPA targets for a campaign. After all, you probably want to set your CPO or CPA at a level where the ROI is acceptable.
Putting It All Together
These metrics and formulas are tools that allow marketers to make better decisions about marketing, shedding light on the good and bad portions of a campaign. Marketers who understand and use these tools will outperform those operating in the dark. A campaign can have objectives that are measured using more than one of the above metrics. If I sell products but also have a monthly newsletter, I can set both CPO and CPA objectives and even meld the two by assigning weights. If 1 in 20 people who sign up for a newsletter eventually orders, then a conversion to a newsletter subscription is worth 5 percent of a sale.