Personalized recommendations are designed to deliver motivation tailored to each customer. Achieving that advantage at scale calls for adaptive automation.
“Show me the offers I want to see.” That’s what customers who have been given recommendations by Amazon and Netflix for over a decade have learned to expect of eCommerce today.
There’s no one-size-fits-all solution offering that will get everyone’s attention. Targeted marketing only works if the right message gets to the right person at the right time and through the right channel.
While that is beyond human capability, it is within reach of those who tap into the power of adaptive automation. As a result of collecting millions of data points over several years, Remarkety can deliver big data insight to derive insight into what appeals most to each customer segment in any type of eCommerce vertical.
Remarkety automates the process of customer analytics through its machine learning models that enables it to identify what appeals to that particular customer. That means that Remarkety clients can deliver personalization for both product recommendations and for the types of promotional offers that get customers to finalize their purchases.
The ROI of relevant recommendations
Customers prefer shopping where they get personalized recommendations. “Shoppers that clicked a recommendation were nearly twice as likely to come back to the site,” reports Salesforce.
That jibes with Merkle’s 2021 Consumer Experience Sentiment Report. About half of the 1300 customers surveyed said that “personalization makes it easier to find products of interest.”
Beyond appreciating the items, customers come to appreciate the brand that delivers on what they need: “88% of consumers view a brand’s products as having higher quality if they feel like the brand is listening to their needs.”
This positive association nearly always translates into repeat business, which raises the LTV of the customer. Both according to Merkle and Accenture 91% of customers say they’d make repeat purchases from brands whose recommendations reflect that they understand them.
The conversion rate for the online shopper who responds to a recommendation is 70% higher per session than for those shopping without the recommendation, according to eMarketer. The benefit holds even on return sessions to the extent of a 55% greater conversion.
The average order value (AOV) is also boosted by recommendation. Shoppers who clicked recommendations had an average spend of 10% higher, according to Salesforce. Even more significantly, “the per-visit spend of a shopper who clicks a recommendation is five times higher.”
It all adds up to a substantial boost for eCommerce. Salesforce found though only 7% of site visits came through recommendations, those visits accounted for 24% of orders and 26% of revenue. According to Gartner personalized messaging delivers a 16% lift overall.
Why machine learning is key to effective product recommendations
What would take a person days to figure out, machine learning can do in an instant. ML has the ability to consider a lot more factors than a human being can and can do it many times over in rapid succession.
It always refers to the customer’s purchasing profile to identify the segment that sheds light on what would be of interest in future purchases. In addition to that, machine learning constantly adapts its recommendations in response to any signal, like what people are currently buying, what holiday or season is coming up, what’s currently in stock, etc.
Remarkety’s product recommendation feature applies predictive analytics to generate personalized product recommendations. It leverages data both from the individual customer and the customer’s segment, drawing on browsing and purchase history to deliver different options for personalizing your recommendations to your customer:
- Complementary product recommendations triggered by a purchase; like the eye liner that matches the eye shadow, the bowl in the same pattern as the platter purchased, etc.. This is key to cross-selling and upselling.
- Best sellers from customer’s favorite category: for the customer who has demonstrated an interest in pet items, for example, this can show what other pet owners are buying now. This comes together through machine learning aggregation that segments the customer correctly to offer relevant recommendations.
- What’s new or trending now. This has to take into account near real time data to show what is popular at present. Even if such recommendations are not tailored to individual taste, customers do like to be informed about what’s new.
Less is more is the guiding principle behind automated recommendations via email or SMS. Customers appreciate these messages because they save them time in searching. However, sending them too many options removes the feeling of personal curation and comes across as spam.
The Coupon Dynamic
Every marketer knows that one sure-fire way to grab customers’ attention and motivate them to complete a purchase by offering a coupon. Online coupon usage has been growing steadily in the United States over the past several years, rising from 126.8 million in 2015 to 145.3 million in 2021, according to Statista estimates.
But not all coupons function the same way. Some are set up like traditional paper coupons that give the same offer to everyone. Those are called static coupons.
Dynamic coupons, in contrast, can be personalized with different “unique” coupon codes that correspond to specific customer segments. That makes them not just more effective at motivating the customers they target but also valuable in terms of revealing which forms of discounts prove most effective.
Because they are unique to individuals and campaigns, dynamic coupons help identify what pushed a particular customer to convert. Then businesses can apply the promotions that get results for the targeted segment.
Types of promotions
1. Free Shipping
This kind of promotion is pretty much expected for online orders today thanks to the practices set by Amazon, Target, Walmart, and other big names in retail. Often it will require some minimum order, though every once in a while an eCommerce will push a free shipping promotion or even free expedited shipping — like Amazon Prime — on any amount to incentivize a purchase immediately.
2. Dollar amount off
These promotions frequently range from $5 to $25, though they can go even higher with sliding scale offers to promote making a larger order like $30 off $150 and $50 off a $200 purchase. Given that $15 off is common for women’s clothing retailers, it’s likely there’s been some research that shows that figure motivates those customers.
3. Percentage off
The most common percentage off coupon is 20%, but discounts can be lower, too. Incentivized branded credit card purchases from Target, Walmart, and Lowe’s build in a 5% discount. Banana Republic frequently adds another 10% off for its branded card users. That would be on top of its typical 40% off offers.
4. Free gift with purchase
These are used to incentivize buying more than one thing to get up to the qualifying amount. It’s often something like a bag, like a shopping tote or a cosmetic bag , depending on the brand offering the promotion.
5. VIP offer, referral reward, order preview, or subscription incentives
These promotions can be anything of the above extended to customers to incentivize an action like subscribing to email or SMS or for sharing a link with a friend. It can also be used as an exclusive promotion to make customers feel special Remarkety insight into dynamic coupons can let you know which offer is most likely to convert for these customers.
The role of machine learning in dynamic coupons
Marketers don’t have to just select among the five categories above to serve promotions to their customers; they have to figure out how much each type of promotion should be.
For example, if you want to go with percentage, is 10% off the way to go, or is too low to motivate this segment of customers? You don’t know for yourself until you try, and then you might have lost that sale for the customers.
What Remarkety does is draw on its vast data to analyze what kind of offers the segment in question has responded to. It may find that 15% off is the sweet spot for one targeted segment. Another may ignore percentage discounts altogether but alway start shopping when given a $10 credit. Others may place an order every time they get an offer for a free gift with purchase.
Remarkety has already collected millions of data points and run the big data analysis to allow you to reap the benefits of machine learning for automated offers that are guaranteed to get the best results. Dynamic coupons reveal which promotions convert your customers, and adaptive marketing can apply the best offer.
Coupled with your automated email marketing program, the savings incentive grabs your customer’s attention and motivates conversion. Remarkety’s stats show that they garner a 76% improvement over the average open rate and an 82% improvement over the average click rate for email marketing.
Adaptive automation will become essential for businesses to compete with the Remarkety advantage of responsiveness, agility, and effectiveness.