We live in an age of instant gratification where our customers want anything, anytime, anywhere.
42% of millennials say they will spend more time reading content if it is tailored to their interest. 50% of consumers shop on multiple devices and channels — email, SMS, web, push, ads.
The modern marketers answer to this is being able to send the right content, at the right time, to the right channel.
In this article, we share 10 secrets that successful marketing teams use to get into the head of their customers and drive sales by showing them what they want.
These insights are based on the work we’ve done building recommendation engines and personalized experiences for online businesses impacting over 100 million users.
1. Do less. Focus on high impact areas.
The most important principle when undertaking big data and personalization is that you should do less.
Senior management or executives often want to implement the whole shebang at once. This approach often leads to disappointment and many years of marketing, engineering, and data science work with non-consequential business impact.
Marketers have too much on their plate and too little time. Furthermore, personalization involves developing algorithms and software and making them actionable. As with building all forms of software, it is better to focus on high impact areas first.
We use the ICE framework to prioritize tasks based on a combined score of impact, effort, and confidence.
2. Find ways to get unknown customers into your funnel
For most consumer businesses, the unknown customer accounts for something like 80–90% of traffic. This is the largest opportunity area in the funnel.
One of the strategies around increasing sales conversions lies with getting to know your unknown customers. Below is a great example from Zalora, whereas when users are about to bounce, they give you $15 off your next order in return for your email (and your gender!).
This email makes you a known customer that they can send nurture with emails and drip marketing.
3. Real-time recommendations
We’ve all heard about recommendations that have been popularized by Amazon and Netflix.
A lesser known principle is real-time recommendations where recommendations are modified on the fly based on the actions a user has taken on the website.
Personalized recommendations can increase sales conversions by over 100%. However, this uplift only applies to 5–10% of users who have bought from your business before. 90–95% of web sessions come from users who are unknown or who have not bought before.
That’s where real-time recommendations shine. By adapting recommended products based on a user's behavior during his/her session, a business is able to provide an intimate and personal service much like a savvy salesperson tastefully suggesting products to a potential customer in a store.
A small, statistically significant uplift in sales conversion on unknown users is enough to impact sales in a meaningful way.
4. Personalized emails
Real-time recommendations can be used not only on websites but also in the emails and other messages that are sent to customers.
In the email from Zalora Taiwan below, personalized recommendations account for 80% of campaign sales, despite being buried at the bottom a lengthy email campaign.
5. Fear of missing out (FOMO)
This has proven to be true in so many business and social situations that we have an acronym for it — FOMO.
Here is an example from Booking.com (that I personally have fallen prey to) where they remind us about why we would want to be making a booking right away to avoid disappointment later.
Another example that retailers use to simulate a busy store space in the online context.
6. Predictive receipts
Receipts are that email that customers really want from you. Make it count.
Furthermore, receipts need to be sent out each time a purchase is made, which makes them a perfect opportunity for marketing automation.
You could personalize receipts to maximize the goal that you are looking to achieve. For instance, you might include personalized recommendations to encourage the next sale like Huckberry:
Or you might give the user the chance to write a product review to increase the amount of social proof you have about a product.
7. Predict churn to improving timing of win-back campaigns
Most brands use a rule like months from last purchase to identify at-risk customers. For instance, many brands consider a customer to be at-risk if he/she has not bought for 12 months.
The problem with a rule like that is it only accounts for the recency of purchases. It does not take into consideration other important factors like the time between purchases, the frequency of purchases and recency of engagement with the brand.
A rule-based model is about 25% less accurate than a predictive churn model.
By contrast, a predictive churn model takes into consideration factors including recency, the time between purchases, the frequency of purchases and the recency of engagement with the brand. As such, it is more accurate at identifying customers who are truly disengaged and at-risk.
Read more about this in our blog post.
8. Get customers to tell you what they like in a fun way
This is one of Amazon’s newer features where they let you discover products you like. What is happening behind the veneer of the playful user experience is Amazon is learning about you and using that data to improve the accuracy of their recommendation engine.
The act of clicking ❤️ on a product shows more purchase intent than viewing or clicking on a product (but less intent than an add-to-cart or purchase). This data is used to populate a matrix of relationships between users and products and in turn allow Amazon to more accurately determine what you and other users might want to see based on limited interactions with the site.
A similar experience on mobile is the Tinder-styled swipe left and right interaction on products.
9. Identify valuable customers with predictive customer lifetime value
Being able to predict the future sales from each customer is key to figuring out which customers to focus on. Getting customer lifetime value right is critical for most business.
In fact, the higher the dollar value of the goods you sell, the more important getting this right is. For instance, knowing the value of each corporate or high net worth client might mean multi-millions of dollars of incremental business for a bank.
Adding up historical sales is not a good way to predict future ones. Predictive customer lifetime value tells you with greater accuracy which customers to focus on.
Read more about this in our blog post.
10. Bottom-up customer personas
Coming up with customer personas is often a mixture of gut feelings from touch points with customers, research reports, and creative agencies. Should big brands be comfortable making multi-million dollar marketing decisions based on an ideal customer that is based on gut feel?
Probably not. An alternate way is to group existing customers with similar buying patterns and interests and understand customer personas based on what data tells you.
Combining this with our predicted customer lifetime value analysis, we know how valuable each of the segments is as well as how valuable a customer in each segment is. Read more about this concept in our blog post.
To wrap things up, here are the 10 things we covered in this article:
- Do less. Focus on high impact areas.
- Find ways to get unknown customers into your funnel
- Real-time recommendations
- Personalized emails
- Fear of missing out (FOMO)
- Predictive receipts
- Predict churn to improving timing of win-back campaigns
- Get customers to tell you what they like in a fun way
- Identify valuable customers with predictive customer lifetime value
- Bottom-up customer personas
If you enjoyed reading this, please feel free to share so others can find it too!
If you are keen to implement personalization for your business, feel free to reach out to us at email@example.com. You can find more writing on predictive marketing and retail digital transformation like this on the Metisa blog.