To start this week's post, I'd like to offer a quick Amazon.com hack: if you place an item in your shopping cart but don't complete the purchase, you'll likely receive an email a couple days later with a discount code. #Savings
Amazon's triggered messaging program was designed to frequently communicate with customers through automated rule-based messages in an effort to increase conversions. Not all triggered messages are discount focused - Amazon's roster of triggered messages include shopping cart abandonment emails, welcome series, inactive subscriber, purchase confirmations, promotional emails powered by collaborative filtering (discussed recently here), and many more. Amazon is a retail giant for a number of reasons, but is their rule-based messaging a reason why?
There are definitive benefits to triggered messaging - messages are sent immediately in response to recent behavior, which is a strong signal of interest and create a way to automate engagement with customers. But there are several drawbacks to rule-based messaging; before your organization invests in this type of system, here are several important considerations to keep in mind.
Rule-based messaging actually adds work for your team
The goal of marketing automation is to reduce the amount of work for a marketing team and increase the marketing team's effectiveness - on paper, triggers do that. However, building and scaling a rule-based messaging program actually adds a great deal of work for marketers. Systems based on business rules require marketers to create and maintain detailed customer personas. Marketers then have to generate business rules for each segment, defining the customer actions that would trigger a response. At that point, marketers still have to write the actual message and make sure the correct item, promotion or discount will populate.
While this doesn't sound like an inordinate amount of work so far, adding more customers requires additional segmentation efforts, a larger bank of prepared emails, and more business rules that are difficult to maintain effectively.
Because the more business rules you add, the more they conflict
Triggered messages don't communicate with each other. A customer could fulfill several business rules in a single browsing session and, in return, could receive either the incorrect triggered message or several (or all of them) at once. For instance, imagine you have a customer that hasn't been to your site in 6 months that receives a re-activation email offering 15% off site-wide. When they show up, they browse Men's shirts, fulfilling an action that places them in the "Men's Shirts" segment and triggers a separate email for a 20% discount. Then, they add a shirt to their shopping cart but don't purchase, triggering a cart abandonment email. In that short period of time, this single customer fulfilled a series of actions that each trigger responses - but which one is actually going to create a conversion?
Customers don't want 3-4 messages a day
According to Brightwave Marketing in a November 2016 webinar titled "Millennials + Email: How to Engage Email Natives," 47.1% of millennials unsubscribe due to high frequency of emails. The webinar also describes millennials as the group that is most accepting of email as a communication channel, so high frequency with other groups is just as problematic, if not more-so. So, to keep customers engaged without over-saturating their inbox, sending with the right frequency is imperative.
Even if your message arrives at the right time, the content might be incorrect
Even if marketers designate enough business rules to send only one message for their customers, that message will be in response to each customer's singular, most recent action. While recent behavior does indicate some level of interest, the last action may not directly relate to customer preference moving forward. For example, I'm still receiving triggered re-engagement emails for pairs of women's slippers (I'm still a man). Until the customer acts again, systems are limited to the small amount of information provided. While timely, triggers can only surface pre-prepared content and merchandise - which can't be managed effectively with millions of items, customers, and business rules to maintain.
The goal - smarter content within fewer messages
Rule-based messaging systems are ineffective because they take the customer's most recent action and send a pre-programmed response without considering the full user context. The goal of email should not be to respond to a single action, but instead present the best content in a single message based on the full context of actions taken by the user - and not necessarily at the exact moment of interaction.
To create smarter content within fewer messages, you don't need pre-programmed rules. You need AI that can learn from the full breadth of user actions, and then dynamically adapt based on the user's response. It makes the consumer feel as though they're receiving a real curated experience from a personal shopping assistant who understands their needs, rather than a robotic response that they can use to receive discounts (i.e. the Amazon shopping cart Hack).
We as consumers don't want more messages - we want better messages.
To learn more about Jetlore, the AI-powered Prediction Platform proven to surface the best performing content for each customer in email, click here.
Join the conversation on social media! As always, join the conversation on social media - follow us @Jetlore #GoBeyondPersonalization.