Oct 24, 2012

Tips for E-commerce Email Marketers

I got some nice feedback when I posted about our new revenue charts. So I thought I’d post some more tips for e-commerce marketers. First, go install our eCommerce360 feature. It connects MailChimp to e-commerce carts like Shopify, Magento, MivaMerchant, osCommerce, and more. Life won’t be the same after you install that. At the very least, connect your Google Analytics account to MailChimp. Then, you can go beyond simple open and click tracking, and get stats like this:

Done? Cool, now we can actually connect user behavior (like purchase activity) with your email marketing.

The Chivery is a website that sells funny t-shirts, and they were kind enough to let me use their account as an example here. These are just some quick ideas I had while skimming through their MailChimp account. Hopefully you’ll find this useful too.

First, learn more about your customers

You can build segments of your list to learn more about people. You don’t have to build a campaign to do this. A segmentation tool is available in the Lists area.

Go to Lists –> View –> Subscribers

Then choose "segment" to slice and dice your list:

Let’s see who our top customers are (the ones who’ve spent the most money) to learn more about what makes them tick:

As you can see, our eCommerce360 plugin pulls data from your cart right into your list, so you can build segments of your subscriber data based on purchase history (with some integrations, you can actually import historical purchase data into MailChimp). In the example above, MailChimp found only 85 of these people "who spent more than $1,000" in the subscriber list. I hate the number 85 (I’m superstitious), so let’s change that.

Let’s scale it back to "spent more than $250" but are ALSO very engaged with the list (they open and click a lot):

There. I’ve got 1,696 of these clicky-spendy customers.

Now, click "View Segment" then go to each member profile (CTRL+Click and open as many as you can in new browser tabs) and learn more about these uber-customers. What’re they buying? Within every member profile, we include their purchase history, including total revenue:

It’s a great way to see the lifetime value of each subscriber on your list.

So what do the top spenders have in common? I clicked six random members from this segment of the list, glanced at their member profiles, and noticed the following:

  • 5 out of six signed up in March of 2011. I wonder if there was a marketing campaign or promotion that got a bunch of people signed up that month (according to their Facebook page, there was a big t-shirt giveaway).
  • 4 out of 6 of these people purchased shirts that were 2XL in size, and they bought multiple shirts each time. Two of them also purchased 3XL shirts. That’s interesting. I opened up three more customers in the top 10, and sure enough, they all bought XL or higher. One of them bought seven3XL sized shirts. Fascinating. This leads to more questions:
    • Should their website be modified to allow for "filter shirts by size"?
    • Should they take the time to create an entire "channel" on their site for these customers ("Big and Tall")?
  • All of them bought the same particular item: a shirt that’s branded with the Chive’s logo (they’re loyal to the brand)
  • They also bought small shirts and women’s clothing. Gifts, perhaps?
  • They’re all @gmail or @hotmail addresses. In fact, of the top 20, the majority are @gmail or @hotmail, with @yahoo coming in 3rd place.

Top Five Customers

Those were simply random people I clicked from a segment of big spenders. On the Dashboard, MailChimp includes a handy little stats block that tells you exactly who your top 5 customers are:

I looked up each one, and noticed that three of their top five (by revenue) customers are from the southeast US (see: MailChimp geodata). The #2 top customer is in the United Arab Emirates. Wow. That’s different. He (or she?) is buying LOTS of stuff for women, and has been buying items every month from May through August, but stopped buying in September. Wonder why he stopped after so much activity. We could conduct a survey on all users who fit this profile, or–you know–just email the guy and ask him what happened.

Granted, that’s just a super fast skim of only six customers. You’ll obviously want to study your customers closer than I did. But just going by that little amount of research, let’s see where that takes us.

Looking at their revenue chart under /Reports/ I noticed that they had some spikes here and there, where revenue was up to 10 times the average. I hovered over the spikes to see if those were holiday sales. Nope. In fact, almost every major holiday campaign they sent was on the low end of the revenue chart. These customers really tune out on their vacations, or something. The only exception was Mothers Day:

Otherwise, those three big spikes to the right were simply emails introducing new products to their customers. Innovation really does open the door for marketing.

Social Data

If you want to learn about your customers, there’s no better way than by analyzing social networks. Well, besides actually talking to your customers face-to-face. But we’re talking about a list in the hundreds of thousands here, so that’d be a loooooooot of chit-chat. So let’s use MailChimp’s social listening tools:

Tweet Trends is a totally free feature we introduced a while back that’ll analyze your list’s Twitter activity. While it doesn’t give you user-specific Twitter data, it gives you a nice, aggregate, birds-eye-view of your audience.

To get to Tweet Trends:

When you first activate, it may take a few hours to populate the data. Patience, Daniel-san. Because once the data’s loaded, it can be really cool. Here, we can see the people that our customers follow on Twitter:

As you can see above, The Chive’s audience is into sci-fi, comedy and TV stars. And Men’s humor. Tweet Trends will also tell you common hashtags used by your customers.  I noticed that #USMC and #GoNavy are common hashtags used by their customers.

Hmmm, we may have a big military audience here. Taking a glance at this store’s Facebook page, I see comments from lots of soldiers, some stationed overseas:

The above comment was about a giveaway that The Chive ran, for a military-branded "Keep Calm and Chive On" t-shirt. Reading more comments on Facebook, it appears there were quite a few disappointed fans who were unable to snag a shirt. One thing I’d be tempted to do is look up those people in my list, then send them shirts if/when they’re available again. Just to surprise them.

I know that sounds crazy, but I think that’s the point of all this technology: to show your customers that you’re crazy about them.


Another place where you can learn more about your customers and get marketing ideas is Wavelength. That’s an app that we introduced early this year, and officially made public last month. Wavelength analyzes your newsletter list, then shows you other newsletters that your subscribers like. For The Chive, this is what Wavelength tells us about their audience:

Publishers on the same wavelength as The Chive include an edgy street art/street-wear site, sports apparel sites, and a jewelry daily deal site. I’m guessing The Chive’s Banksy tees is the tie-in to Obey. The sports sites make sense for their male audience, and are perhaps some places where The Chive can do some advertising. The jewelry site is intriguing. It sparks all kinds of ideas for them to perhaps expand into new audiences. Anyway, something to noodle on for later…

Now take what you learned, and go automate

Okay, so we know a little bit about our 1,696 most engaged, top-spending customers. Now what? Do you stop everything you’re doing and send them extremely targeted offers? I guess you could. But it’s a lot of work crafting a special little email campaign to just 1,000 members when you have a list of hundreds of thousands (or millions) of subscribers. That’d be great for sending them a one-time special offer. But it’s not scalable.

So you can automate the process. Set up some Autoresponders that automatically go out to those tiny slices of special customers.

Good email marketers focus on:

  1. Sending the right content
  2. To the right people
  3. At the right time
  4. Repeat 1-3.

#1, is the hardest part. #2 is all about segmentation and targeting, #3 can be linked to purchase behavior, and #4 can be accomplished with automation. When most people get into marketing automation, they start backwards, at #4: "Cool, I have the power to automate! Let’s set up rules! If a customer does this, send that. But if they also do this and that, send this!" After spending countless hours tinkering with fun automation rules, they realize they have no content. Content is really, really hard. Start there.

Fortunately, a lot of companies are on Facebook, which can be a treasure trove of useful, fun content for your customers.

Here’s a basic idea. Send a special "refer-a-friend" offer to troops overseas:

Or, instead of sending a coupon (to people who already love you and are willing to give you money), put together a very nice thank-you email, perhaps with a video just for your military customers. Make it funny, since that’s what they like. Or ask them to take a picture of themselves wearing the shirt while on duty out there, and post to your Facebook page. Nice way to get this crowd to engage on Facebook. BetaBrand does an amazing job of getting customers to post action shots on Facebook. Making them famous is way more fun than a coupon.

Or maybe send a nice thank you note about how you’re donating a portion of proceeds to the USO or something.

Some of their top customers bought 10 shirts at a time. This made me wonder if some of them were store owners. And remember the big spender from the UAE? Who are these extreme buyers? If they’re store owners, might there be a special bulk-buyer offering we can provide? What if we automatically send a quick survey to these people who spend so much per order?

You can send a link to a quick Wufoo or SurveyGizmo survey that just tries to find out who the heck they are, and why they buy so much. Make their responses go to your inbox. Over time, read through them all to see if you can discern any pattern or traits shared between them. Maybe it’ll give you some new product or marketing ideas.

Looking at The Chive’s Facebook page, I noticed they’ve got some Star Wars fans (who doesn’t?) So if/when a customer purchases their Darth Pinata shirt (lol):

They could send a personally written "Thank You" email with a link to their Facebook photo album: "Rare Unseen photos from Star Wars"

Now, wait just a minute here. It turns out The Chive also has a photo album all about Darth Vader helmets:

Jackpot. Now that would be a pretty fun thank you email.

Or, flip this concept around. Don’t just wait till someone buys something from you. You can use Star Wars fandom to generate awareness about your Star Wars apparel.

In the days leading up to Star Wars Day (May the Fourth be with you), send a campaign to customers with all the Star Wars content in The Chive’s Facebook albums (this album of photo-chops was pretty funny):

I’d include content from outside The Chive, too. Peter Cooper taught us how to make money by NOT talking about yourself. And since we know The Chive customers have some appreciation for art (see Wavelength results above: Banksy/Obey) perhaps a link to this archive of LucasFilms’ awesome holiday cards would be a nice fit:

In that email, you might also include a link to the Darth Pinata shirt. You might generate a list of people who’ve previously purchased the Vader shirt, and exclude them from this mailing (here’s how). I’d be curious to see if this is a holiday campaign that would actually spike on that revenue chart.

So there you go. My ideas are all probably awful, and need way more testing. But I’m just trying to demonstrate how you can gain a lot of insight using data sitting in your MailChimp account, and learn a lot about your customers. Then, you can use what you know to find the perfect content to send to your customers, write it in a way that relates to them, and maybe even use Autoresponders to repeat that connection over and over again.

Also See: