CHALLENGE
When feature film editor Andy Jenkins developed a training DVD five years ago, he decided to try selling it directly to video professionals via an online store.

Things went very, very well. The folks behind 'The Blair Witch Project' loved the DVDs so much they provided testimonials for the site. Soon Jenkins had customers at most major film and TV studios.

Inspired, Jenkins left his film career behind to continue launching more ecommerce sites, one for each passion he and his staff felt they had a critical mass of knowledge in. For example one store offers hundreds of tapestries to consumers who love decorating. Another helps wanna-be entrepreneurs understand how ecommerce works.

Among other tactics, the team drove new traffic to the various sites by "spending copious amounts in PPCs" and encouraging affiliates. But the only way this business model would be sustainable for the long run was if Jenkins could turn traffic into opt-ins and then make those email lists sing.

Soon he was sending dozens of completely different campaigns each month.

How do you manage an email program gone wild without spending night and day ceaselessly composing new campaigns? And which tactics can you use in all marketplaces -- and which won't fly?

CAMPAIGN
First Jenkins set seven specific email Best Practices rules for all site marketing teams to follow for campaign quality control and maximum results.

Rule #1. Clear overt opt-in

The majority of the sites require double opt-in, so visitors must type in their email address and then go to their email boxes and click another link to be formally added to the list. None of the sites offer pre-checked boxes or ever send promotional email to names that have not raised their hand requesting it. Even if you're a customer, you won't get promotions unless you asked for them.

Why so tough? Most of Jenkins' markets are extremely niche. You don't want to annoy folks in a tight vertical because you hope to sell them for years to come.

His only true mass marketplace -- the entrepreneur wanna-be crowd -- is deluged with messages from less scrupulous mailers. Which means in turn those mailers are filtered more heavily than average by ISPs and end-users. Sticking with double opt-in helps keep Jenkins' IP address off the blacklists that are watching him.

Rule #2. Clear overt opt-out

"If they're not going to be a customer, then I don't want to bother them," says Jenkins. "I want big lists with a lot of people ... but if they want off, I let them go -- goodbye!"

"I have a gut feeling it works. I've spoken to customers who opted out and told me, 'You didn't flood my inbox so I came back and bought something from you.'"

This means going far beyond the usual unsub link at the bottom of campaigns -- in fact Jenkins routinely puts a big fat unsub link at the very top of messages. He's perfectly happy if as many as 50% of new names use the link to leave the list because he knows his campaigns are not causing seething resentment (or spam complaints that could land him on blacklists) in the marketplace.

Rule #3. Fast thorough response to spam complaints

Jenkins' store databases track all important metrics for each opt-in, including the time and date they opted in, the site page they opted in at, the IP address they came from and precisely how many emails have been sent to the address.

"I hate to position it this way, but I have ammunition against spam complaints. You can tell them, 'You double opt-in subscribed on this date, this is your IP address, we mailed you X times.' That eliminates any kind of harassment."

Rule #4. Unusually useful metrics reports

Jenkins has customized the sites various email reporting systems in three critical ways:

-> Easy to group and track names by past behaviors, so no report is for "the entire list" but rather for segments proven over time to react uniquely -- such as newbies versus multi-buyers

-> Metrics include ROI based on sales, not just clicks.

-> Reports present critical data in bar charts and graphs rather than just spreadsheet cells to aid in visual impact and quick decision-making.

Rule #5. Segmentation by customer behavior

Based on initial reports, Jenkins decided to divide up his campaigns and efforts to target and track efforts by four different types of email-driven shoppers:

a. Opt-ins who've never bought ("just suck up server space"); b. "Immediate customers" who bought within 24 hours of visiting the site for the first time; c. Converted prospects who took longer to make an initial buying decision; d. Repeat customers.

Rule #6. Reiterate conversion messaging in Welcomes

Instead of using a generic "welcome new subscriber" message, Jenkins features each site's key benefit and offer statements in the Welcome. So a Welcome might include reminders about free shipping, easy returns and famous-name testimonials.

This way the Welcome serves as an extension of the store's home page, continuing the conversion process, instead of just being a bland transactional message from the email server.

Rule #7. Unremittingly test, test, test

Jenkins has tested, and continues to routinely test, all the email basics for each marketplace, including - Text vs HTML - Day of week and time of day - Full-article in email versus summary plus hotlink - Ratio of "editorial" emails to sales pitches to a list - Varying copy and offer.

With these best practices in mind, the team next set out to garner as many good opt-in names as possible. Most ecommerce sites rely mainly on an opt-in offer in their check-out process. But, Jenkins points out this could be a huge mistake; you're not capturing emails from shoppers who need multiple trips to convert.

"Getting people onto your list is an important conversion component that you have to work into your ROI. Let's say 45% of people who come to a site leave in the first 12 seconds. It's incumbent on me to keep those dollars on the site. If they sign up for a list we can remind them about our 110% guarantee, our low prices...."

To that end Jenkins has "run the gamut" testing ways to get opt-ins that turn into repeat buyers.

Test a. Discount coupon offer on home page

Jenkins tested offering a 5% discount to new visitors to entice them to sign up for email as they shop (well before the check-out process). The offer appeared on site home pages. Everyone who signed up received an electronic coupon in their email they could apply right away to their first order.

Would the ultimate value of increased email opt-ins outweigh the immediate profit loss of giving out discounts?

Test b. Email offers on product pages

Jenkins tested adding email offers to landing pages for individual offers. If a visitor wasn't quite ready to convert immediately, they could sign up for email to learn more and hopefully convert later. Would this "maybe" option reduce the number of immediate paid conversions?

Test c. "Name squeeze pages"

The team also tried putting a page up that blocked access to the product page. To access product information, first you had to decide whether or not you'd hand over an email address. Key: there was always an incentive such as educational materials in PDF or in audiofile format that you'd receive by signing up.

Test d. Viral-style email offers

Hoping to benefit from pass-alongs, Jenkins created enticing educational PDFs and distributed them freely. Then to turn pass-alongs into email opt-ins, he added an offer for an additional PDF such as a tapestry design guide at the top of each page.

The mail-to click link started a message in the consumer's email system they could send to the site to get the extra guide and also be added to the list.

"You always have to warn them when they click on this link their mail program will pop up so they can see where they are sending the subscription request." In addition, in the Welcome message you must reiterate the fact that the name's been added to an ongoing list and offer a prominent opt-out link.

RESULTS
"My ROI is good. We're in the black," says Jenkins. The other metric he's proudest of is that after sending more than four million messages over five years, he's only received one serious spam complaint. And that was quickly resolved due to his careful record-keeping, "The host wrote me back and thanked me, 'You are on top of it.'"

His sites have continued to grow in popularity -- the tapestry site has been featured repeatedly on the cable TV decorating show 'While You Were Out.'

While many of the metrics he's tracked over the years have "turned out to be meaningless," segmentation by shopper type has proven very valuable. "Someone who buys right away has a 30% chance to become a repeat customer versus a prospect who buys after they've been on the list for a while."

Turns out immediate customers give the most response to new product announcements. Prospects who took longer to convert respond best to email campaigns featuring storewide sales.

Some tactics worked across all lists and marketplaces:

-> Best time: "When emailing for dollars, that is, when making an offer of a product, the very best time that I've emailed is 4p.m. ET on Thursday, no matter what the price point or offer.

"Maybe it has something to do with Friday being a pay day, or the euphoria that a person feels that tomorrow is the last day of the work week. Maybe it's all about being re-born on the weekends, and the offer presents an opportunity for them to do or get what they've always wanted."

MarketingSherpa note: These reasons may apply to Jenkins' uniquely enthusiast audiences, but not to other types of ecommerce sites. No one time of day or day of week is best for everyone. (But you already knew that.)

-> Text vs. HTML: "Text is not the clear winner, but has demonstrated a better clickthrough rate. I track individual links that are presented in an email, and by looking at what links are being clicked on, I can get a sense of how deeply into an email a reader gets. In this case, TEXT wins.

"Now, when it comes to physical products, HTML always trumps if, and only if, the relevant image that depicts the product has an offer (savings, free shipping, limited quantity) attached to it."

-> Full article vs. summary plus link: "I have published emails that simply state "Click this link to get the full story". 14% less response. However, when I send out both versions to the same list [staggered] separately, response goes up 6%."

Other tactics only worked for certain demographics and sites:

-> 5% discount opt-in offer: Didn't work in the B-to-B marketplace, but is now a staple on Jenkins' tapestry site (link to creative below). "Turns out only 70% of customers that purchase ever use it, so it doesn't lower our profits by 5%. It wasn't the enormous wonderful growth I thought it would be for overall sales though. It's right on the cusp."

-> Name squeeze pages and viral PDFs: Only worked in the wanna-be entrepreneur marketplace, mainly because "they are so thirsty for education." Up to 50% of the resulting new names opt-out within 72 hours, which Jenkins is happy about because the more non-buyers who opt-out, the lower the chance of spam complaints.

-> Name squeeze pages vs. opt-in offers on product landing pages: "Our opt-in rate went up with the name-squeeze, but of course close rates went right down. Overall net we increased sales 22%."

Useful links related to this article

Creative samples from various A Squared Artifact's email campaigns and opt-in campaigns: http://www.marketingsherpa.com/andyjenkins/study.html

PDF Creator: http://sourceforge.net/projects/pdfcreator/ Several of Jenkins' various sites:

Like a Pro Avid Training DVDs : http://www.likeapromedia.com

Yahoo Store Builder: http://www.yahoostorebuilder.com

A Squared Armory (replica medieval weapons and armor): http://www.a2armory.com

Stomping the Search Engines: http://www.instantseoexpert.com

Tapestry Standard: http://www.european-wall-tapestries.com