Marketing automation can be an extremely powerful tool, particularly for B2C ecommerce marketers that have to create a highly personalized experience for hundreds of thousands of customers a month. However, used improperly, marketing automation can have the opposite of the intended effect and create a user experience that's simply evil.
1) Treating B2C as faceless entities and B2B as real people
It's impractical for most ecommerce retailers to manually go through all of their customer records and engage personally with each of them like they do in the B2B world, and this is where marketing automation gets it's great value. In the B2B sales world, a real human being can manually review a given contact record and make human decisions about how to create a personalized and contextually relevant buying experience. In the B2C ecommerce world, that's impossible given the sheer volume of customers we deal with.
Because of this, many ecommerce marketers have decided that if they can't treat manually each contact as an individual person that they're going to treat all contacts the exact same. They blast their entire email list three times a week with a coupon and call that email marketing. They show everyone the exact same content on their website. They sometimes even ignore messages and comments coming in through social media because of the sheer volume of them instead of using marketing automation to solve these problems by collecting contacts into similar buyer personas that can receive a more targeted engagement.
2) Branched Logic vs. Lifecycle Nurturing
Probably one of the biggest mistakes in marketing automation is the idea that the customer should agree with you about what their next step should be. Traditional, "branched logic" marketing automation goes something like this:
- Contact does X.
- Send them an email asking them to do Y.
- Time passes and contact doesn't do Y.
- Send them an email asking them to do Y.
- Time passes and contact doesn't do Y.
- Send them an email asking them to do Y.
- Time passes and contact doesn't do Y.
- SEND THEM A REALLY AGGRESSIVE EMAIL ASKING THEM TO DO Y.
- Time passes and contact doesn't do Y.
- Stop sending the contact any more emails - they're obviously a bad contact that's never going to buy. Add them to the main list and blast them three times a week with a coupon along with the rest of the schmucks.
The sin here is assuming that the "buying cycle" (which we call a cycle yet somehow always illustrate as a linear process) involves a set journey that customers go through, and if they've done action X that action Y is the only plausible next step for them to do. It doesn't, for example, account for the fact that if I'm in the comparison phase and I discover a new feature I didn't know about that I might move back into the research phase of the buying cycle and that the types of content you should send me should reflect that change.
Instead, flag specific lifecycles that contacts can be in and nurture those with relevant content. The buying cycle is more complex than a linear, branched process - and people can go backwards. Respect that.
3) Ignoring the customer after the sale
We've talked before about the limitations of transaction-oriented unit economics and how that can ruin both the user experience and the success of your marketing campaigns. Ecommerce marketers may know intuitively that it's easier to sell to people who have already bought from you, but many of them are not doing anything more sophisticated than sending coupons three times a week to everyone who's bought before.
People are very excited in the intermediary time between completing the order and when it's delivered. There's a massive anticipation people experience. Use that time after the transaction to inform them when it will be delivered, send them relevant blog articles about how to use your product (see section: product adoption, the forgotten war), and introduce them to your social community. Also give them a chance to share their excitement with their social community - capitalize on that excitement to make acquiring your next customer cheaper! You can also follow up after they've received the product with a survey, such as a Net Promoter Score survey, that gives you valuable data about how your products are received and can also nip potential customer service issues in the bud.
After they've had a chance to use the product to address to reasons they bought from you, don't ignore customized remarketing. Up-sell, cross-sell, and re-sell marketing activities are the most powerful levers in increasing ecommerce customer LTV - yet again many marketers simply send them "specials" and coupons of what they want to promote instead of sending the customer products that are relevant to them. If you've gathered pre-transactional intelligence about the contact, such as why they want to buy a specific product and what they're using it for, you can create a highly relevant experience that respects that fact that a new purchase starts at the beginning of the buying cycle again. Just because I bought an HDTV doesn't mean that you should just send me a coupon for 3D glasses. I didn't even know those had advanced beyond the red and blue plastic lenses - much less that movies didn't look awful through them anymore. Once you introduce me to that, I start at the beginning of the cycle again.
4) Limiting yourself to just email
Customers don't think of their relationship with a brand within the confines of a specific "channel" such as email. They have a holistic relationship and concept of the value of a brand that extends across all digital and offline channels. Telephone interactions, live chat on the website, social media interactions, and visits to physical locations are all relevant to them and whether or not they buy from you (or buy from you again). Don't restrict your marketing automation to just email.
For example, you may want to have a special list of contacts that are flagged as a promoter through your post-transactional NPS survey (see above) and use marketing automation to detect if that customer has a Twitter account and then trigger an email to your social media team to follow and engage with them on Twitter. You may want to have dynamic content on your website that reacts to behaviors customers have taken, such as specific product detail pages they've viewed or other information you have about them. Since the customer views relationship between them and your company from a holistic perspective, your marketing automation should work in tandem with the various customer touch points.
5) Focusing on products over problems
Amazon was at the bleeding edge of technology a decade ago when they decided that someone who bought the same product as someone else might be interested in other products that someone else has purchased. It was a huge win! However, ecommerce marketing has grown beyond that.
The psychographic dimensions of buyer persona cohorts are strongly tied to the "why" of a customer's engagement with you. What are they using your product for? If you're selling camping gear, is this person solo backpacking across the Appalachians or are they camping with their family? The reason they're buying from you - what modern marketers would call the "job-to-be-done" - is equally as important as what they're buying from you. Instead of making the entire context of your engagement with them around what they're buying, identify the why (either through behavioral analysis or - and I'm being really radical here - asking them) and use that in your marketing automation to frame the conversation of your persuasive messaging.
6) Relying on Big Data alone
As referenced above, your marketing automation should rely on more than just the Big Data behavioral analytics. You'd be surprised how much information people are willing to give you if they understand why you're asking and what you're going to do with that information.
For example, I've rated over 700 movies for Netflix - a significant investment of time - because I know that they're going to use that information to improve their product recommendations for me. I know that they're focusing on improving my experience with them, so I'm willing to answer almost any questions they ask about my entertainment preferences.
Make the case to your prospects and customers - if you really are using their information for noble and not nefarious ends - as to why they should choose to share information with you and you can get very specific information that can help your marketing automation become laser targeted. Selling camping gear? Ask them where they're thinking about camping and when they're planning to go and tell them that you'll send them a guide customized to that region. Based on that information, you can not only send them highly targeted product recommendations (camping in the Northeast can be very different than camping where I'm from in Florida) but also make your emails come at the right time to help know when you need to shift from education to urgency building.
Which brings us to our final sin:
7) Ignoring the 4th dimension
In physics, depending on your theories around space-time, there's an all-important 4th dimension that plays an incredibly important role in the functioning of the universe: Time. Physicists treat time with equal weight and care as they would depth or length.
In ecommerce marketing, time is an equally important variable as what you're sending the customer. The most highly relevant message at the wrong time can still fall flat and fail. In fact, DiningIn.com - who will send awesome food to your home instead of you having to go to restaurants - saw a 4,200% increase in sales from their email marketing simply by starting to send people emails about food at the time of day when it will be most influential.
If I'm buying breakfast food, why would you send that to me at 2PM? If I frequently order lunch for my office, is sending that at 7PM a good idea?
Respect that 4th dimension of time and make sure you're getting the right message to the right people at the right time.
What about you? Have other brands sinned against you with their marketing automation? Let us know in the comments!
Originally published Dec 20, 2013 10:00:00 AM, updated January 18 2023