Metrics That Matter
By Regina Brady
Analyze performance to optimize future e-mail campaign results.
One of the wonderful things about e-mail marketing is that we have many different ways to measure its performance, and we can use this information to better understand our customers, to learn from each campaign effort, and to improve future mailings.
Any good e-mail reporting system should provide a wealth of detailed data on campaign performance. It's important to take both a "macro" and "micro" approach to e-mail reports. Look at overall results, and then dig into the details to optimize campaign results in the future.
It all starts with the standard metrics that are collected. Here's what to look for:
1. Delivery statistics. Has your e-mail been delivered? Most programs break out invalid addresses and bounces. Invalid addresses have a problem with their syntax, such as a space in the e-mail address, a misspelling of the domain or a mal-formed extension. There are hygiene programs that can correct invalid addresses, although you should apply them cautiously so that any correction does not result in your sending e-mail to someone who has not given you permission to contact them via this channel.
Bounces are e-mails that did not reach their destination and were returned to the sender. Technically, information is included with the returned message that allows the sender to know why it was undeliverable. These generally are categorized as hard bounces and soft bounces.
Hard bounces indicate the domain or the individual at the e-mail address no longer exists. These addresses should be retired immediately, or you can drill down further to examine reason codes and selectively retire these names.
Soft bounces indicate a temporary condition that makes the e-mail undeliverable. The mailbox could be full. A server could be down. Resend soft bounces for up to four campaigns, then retire any names that could not be delivered successfully.
Be careful: ISPs look at bounces by sender, and a high percentage of bounces may result in your e-mail being blocked.
2. Open rates. This data allows senders to know how many recipients opened the message, offering marketers insight into how well their brand and subject line resonate. Generally, open rates only are available for HTML e-mail. They can be misleading (recipients may have their preview pane set to display the first few lines of a message, and this would be counted as an open) and should be used as a relative measurement to compare and contrast campaigns over time. Look at total opens and unique opens. Unique opens measure the number of individuals who open a message, not the total number of times a message was opened.
3. Click-throughs. This is the number of recipients who clicked on a link or image in an e-mail and followed a hyperlink to a Web site for additional content. Each link in an e-mail should be tracked separately. You'll want to examine the total number of click-throughs, unique activity and specific link results. Total click-throughs measure the number of times any link was accessed. Unique click-throughs measure the total individuals who clicked on any link. Specific link results allow you to determine what content or promotions in the e-mail captured the target audience's attention.
4. Unsubscribes (opt-outs). The number of recipients who ask to have their e-mail addresses removed from the list distribution. Most direct marketers will see unsubscribe rates of 0.2 percent.
5. Referrals or pass-alongs. This metric measures the number of recipients who forward the message along to a friend or colleague.
6. Sales. Some reporting systems also can tie in buying behavior generated by an e-mail. For e-commerce sites, this information is critical to determining success.
As you examine overall results, you'll also want to compare and contrast your overall metrics with industry averages.
With basic reporting metrics in place, you're armed with the tools necessary to both analyze performance and enhance subsequent campaigns. Here are some areas to examine further:
7. Length of the relationship. New subscribers to your e-mail programs often are enthusiastic, and your open rates and click-through rates may be higher for this group. If you've recently had a large influx of new customers, this group's performance may mask a problem with older customers. Try to understand whether there are certain timeframes in which your recipients begin to show vulnerable behavior. If click-throughs decline, for example, after individuals have received your e-mail program for four to six months, you may want to plan a stimulation effort to reactivate these customers.
8. Results by domain. It is helpful to analyze behavior for the top domains that represent the largest portion of your recipient base. You may be able to spot a delivery problem with a specific domain and then work to correct the issue. Or, you may see that performance of free e-mail accounts, such as Gmail, Hotmail and Yahoo!, is significantly lower than that of ISP accounts. In this instance, you might develop a contact strategy to obtain a "better" address from these customers.
9. Results by offer or content. Understand which offers or content appeal to your audience. Look for general patterns of receptivity so you can fine-tune your program. Also examine this behavior with an eye toward developing meaningful subsets of your file. The time you spend developing segments based on activity can help you develop future offer, creative and contact strategies.
For example, you may identify a group that responds to sales and consists of low margin customers. You might choose to either give the group more discount offers or find a way to stimulate buying behavior to improve margins. Or customers who respond to particular content might dictate the inclusion of additional content in this area of interest. If you can tailor your communication to segments, you'll be more relevant and results will improve.
10. Results by activity. Identify those customers who are actively engaged with your program as evidenced by frequent click-throughs, opens and resulting sales, as well as those who are inactive. Inactive customers are vulnerable, and once you pinpoint this group, you can develop strategies to re-engage these potential defectors.
11. Results over time. Examine the consistency of your campaign results over time. How do your October results compare with the previous month? Are patterns stable or did certain campaigns do significantly better? Delve into the details to understand what's working and what's not.
12. Results by data attribute or combinations of data attributes. You likely have multiple data fields that define your customers. Look at behavior by specific data attributes to determine whether there are differences in behavior. For example, do males respond differently from females, or do recent offline buyers perform differently from online buyers? There are many ways to slice and dice data. Direct marketers often employ recency, frequency and monetary value (RFM) analysis. This is another way to develop subsegments in your customer base.
Once you analyze e-mail-specific behavior, link pertinent information on customers to your master file or offline database. There may be additional data fields on your master file that would be helpful to add to your e-mail database and vice versa. This also will help you build an integrated communication plan for customer touchpoints.
It takes time and effort to conduct the appropriate analysis of your e-mail programs, but the approaches outlined here are similar to the rigor applied to any direct marketing campaign. Careful investigation of campaign results will allow you to develop hypotheses, test and ultimately improve results.
Regina Brady is president of Reggie Brady Marketing Solutions, a direct and e-mail marketing consultancy in Norwalk, Conn. She can be reached at (203) 838-8138 or firstname.lastname@example.org.