• by George Bilbrey , Op-Ed Contributor, April 8, 2015

Email marketing blogs and mailing lists are full of best practices born of many years of experience by many smart, hard-working email marketing practitioners.

While these best practices can be extremely useful, in some cases advice is given without a lot of serious analysis. By looking at actual data, we’ve found that one thing becomes clear: Best practices aren’t always right.  Let me give you two examples.

1. Village wisdom: Subscriber-level, engagement-based filtering at mailbox providers means that you should mail more.

At a recent Email Evolution Conference, a panel of mailbox providers indicated that they use data on the quality of engagement (messages read, deleted without opening, etc.,) in their determination of inbox placement at the subscriber level. This wasn’t news, but for several voices in the email ecosystem this became a “Eureka” moment. Some said that these new rules of deliverability meant that marketers should send more mail. The logic: If the only subscribers that don’t receive mail are the ones who don’t read your messages anyway, you should just send more. It’s not going to hurt the response from your campaign.

What the data says: There is an optimal point for frequency that varies by sender. In many cases, mailing more will destroy value.

For example, when we took a look at the data for apparel retailers, we found:

• Read rates fall pretty dramatically with each additional weekly message sent per subscriber. At some point the lower batting average (read rate) that comes with sending more mail overwhelms having more at bat. At that point, mailing more can mean that fewer people read your mail and fewer messages are read in aggregate.

• Complaints increase with each additional weekly message. As frequency approaches once per day, the effect is non-linear, with exponential growth in complaints. This has two effects: You lose the lifetime value of an active subscriber (only the active ones bother to complain) and you increase the probability of deliverability issues, especially as complaint rates near 1%.

• Inbox Placement Rates decrease with the frequency of mailing to inactive mailboxes. We found a definite correlation between sending to more unengaged subscribers and failing to reach engaged subscribers’ inboxes.

When modeling this out for clients, we found the optimal send frequency depends on many factors: how much you are currently mailing, conversion rates, customer “churn” rates, and average order values, to name a few. For many, following “village wisdom” results in over-mailing, which clearly hurts marketing performance.

2. Village wisdom: Offering incentives in welcome messages is a best practice that should work for everyone.

At several conferences, I’ve heard people say that by offering incentives during welcome campaigns you can drive new subscribers to interact with your messaging, building more engagement over their lifetime.

What the data says: Your mileage may vary.

Broadly, across the entire representative sample we studied, we found that offering incentives didn’t change welcome message read rates. Some brands offered incentives and saw high read rates, but for others -- most, in fact – this tactic very much didn’t work.

What does this mean for email marketers?  I would suggest three things:

• Challenge assumptions: Just because something is a best practice doesn’t make it true for you. In fact, a common-sense recommendation may be actively hurting your marketing performance.
• Use available data on what works for others:  Competitive intelligence products that show the performance of other marketers’ campaigns can give you a feel for what might work for you.
• Test with a control: Find the 1-2 levers (frequency, program mix, etc.) that can really make a difference, and try something new to see what works for you.