Thirty-five percent (35%) of what consumers purchase on Amazon comes from product recommendations according to McKinsey.
The same study reveals that 75 percent of what we watch on Netflix comes from product recommendations using sophisticated algorithms and predictive models – artificial intelligence – to analyze transaction data and what’s trending on social media.
Product recommendations are convenient for your customers, and instant upsells for you.
As for email, product recommendation emails have always generated sales. But historically have been created and sent as a single-item promo in an email blast…
- Artificial intelligence and email
Amazon also uses artificial intelligence to make product recommendations in their order confirmation emails as well as on their website. The same “frequently bought with” algorithm you see on their website delivers product recommendations at the right time in an order confirmation email.
The “artificial intelligence” already in use by Amazon is gaining more traction in the retail industry.
Sephora, the beauty retailer, makes product recommendations at the bottom of the confirmation email with a rotating banner of products they are promoting. It’s just a small leap of data integration to make those recommendations more personalized to the buyer’s actual purchase.
- Automated marketing campaigns are just the start
Segmented email campaigns get more opens, click-throughs, and sales conversions. More importantly, in today’s crowded inboxes, your consumers expect relevancy and personalization from you. Fail to deliver, and you get booted from the inbox. Let’s say you were an iGaming brand looking to incentivize Canadian players to come and play their hand at some online poker. You can mine the data you already have about all people who have expressed interest in things like “slots online at www.mansioncasino.com/ca” and use this data create a list of players to email. This makes the offer highly relevant to those players in Canada who have expressed interest but haven’t come play at your site before.
Personalization through segmentation has gotten so much easier with marketing automation tools. Email automation platforms empower retailers to customize email communications with rules and conditions in addition to demographic information.
As Chad White, from Litmus explains, “data science is making email marketing smarter by powering better automation and personalization. However, data science isn’t a “set it and forget it” solution. They’ll zero in on the data points and behaviors that lead to truly valuable insights.
And, you can go as granular as you want with “if/and” rules to customize your communications even further based on reader behavior.
An abandoned cart email is the best example. “If” a consumer visits the check-out page, but doesn’t complete the purchase, “then” they will get a reminder of items in their cart and possibly an incentive to complete the purchase.
The next step in artificial intelligence is to create algorithms that will automatically identify:
- Products associated with the products just purchase for upsells and cross-sells
- Customer segments that give you the most conversions for specialized promotions and campaigns
- Recency and frequency to help you identify your most loyal brand ambassadors for specialized promotions and campaigns the minute they cross the threshold number of interactions
- Email timing
One of the reasons confirmation emails get the highest open rates is because it’s immediate and relevant to your buyer at that moment.
For all other promotional emails, most us of have had to test the “best time to drop an email” to our full list to determine the aggregate best time to send an email.
It’s worth testing because we get higher campaign performance when more people open the emails.
Artificial intelligence is already allowing retailers to send email messages customized to when each reader is most likely to open the email based on their past “open” behavior.
For example, if a customer in New York almost always opens their email first thing in the morning, that consumer would be tagged for an early morning delivery of future promotions.
Likewise, if a consumer in Chicago doesn’t read personal email until after work or evening, then that’s the optimal time they will receive the same promotional email.
Just like email list segmenting has gained wider acceptance and application in recent years, most serious email marketers will embrace artificial intelligence in the next few years due to consumer demand for relevant, customized communications.
With the promise of increased sales based on product recommendations, upsells, and cross-sells, it’s a worthwhile embrace.