As a marketer, you know how hard it is to keep up with customers today. They do in-depth research for most purchases and expect brands to assist them in a personally relevant way at every stage of their purchase journey.
It seems like a tall order, but some of the same technological advances that empower these impatient and demanding consumers can help marketers too.
Machine learning is one automation tool that can boost your efforts to meet people’s high expectations. Machine learning can help create automated marketing campaigns, at scale, that place the right message in front of the right customer at the right time at the right price, delighting people and delivering the business results marketers want.
Having worked with hundreds of brands who have adopted machine learning to improve their marketing, we’ve found that the top-performing brands are applying these five rules — and asking themselves these questions — to drive success in the age of automated marketing:
Machine learning is only as good as what you ask it to optimize. Top performers optimize for profitable growth and take a holistic view of their marketing, while others are obsessed with efficiency or measure too granularly, missing the forest for the trees.
For example, by focusing on long-term profitability instead of short-term ROI, HomeAway was able to dramatically turn around its business, increasing 2017 revenue by 115% year over year.
As another example, Google has worked with some top-performing financial services companies that optimize for product purchases made online, by phone, or in person while others optimize only to online information requests. A machine learning algorithm optimizing for actual products purchased in every sales channel will drive a lot more sales, much more effectively than one optimizing only for an information request made online.
Or consider online financial services company Betterment. Instead of focusing only on search or only on video, it made them work together. Betterment used custom intent audiences to engage YouTube viewers who recently searched for financial keywords on Google. The brand significantly improved its YouTube campaigns, and also saw a 245% increase in brand searches on Google.
You’ve probably heard the general rule that 20% of customers drive 80% of profits. Yet many marketers acquire new customers as if they are all equal. Top marketers use machine learning to put more of their money toward marketing to more valuable customers and less of it toward people not likely to drive business results over the long term. That means they’re automatically reaching people with a higher customer lifetime value (CLV).
For example, Google worked closely with a travel company that learned its best customers were not always the ones buying the most expensive trips, but were customers booking multiple trips over time. The company increased its focus on these high-value customers instead of caring mainly about short-term ROI and acquired more of the people most likely to help grow its business over time, automatically, using machine learning.
Top marketers also focus on increasing the CLV of their existing customers. By earning more over time from each customer they acquire, they can afford to invest more to acquire other new customers. Better yet, they can acquire more customers than their competitors.
Top marketers increase CLV by improving cross-selling and lowering churn using machine learning. For cross-selling, they forecast which product each customer should buy next and proactively market it to them. For churn reduction, they identify high-value customers at risk of churn and retain them with unique offers.
After insurance giant Allstate launched its first cross-selling and churn-reduction campaigns, it increased retention by 2.4X and found that cross-selling was 4X more effective than reacquiring customers.
In a world where online marketing will be automated, the power of your brand, the personalization of your ads, and the emotional connection you create with consumers will matter even more.
For search ads, machine learning can create hundreds of tailored ads for a single keyword by using a new tool called responsive search ads. It creates unique ads from a few headlines and descriptions, and automatically serves the right ad to the right customer.
On YouTube, advertisers can use machine learning to personalize content at scale. Frito-Lay identified the most popular YouTube content categories — everything from gaming to 90s fashion — among its target audience. It then used YouTube’s Director Mix tool to quickly create different creative variations for each of the top categories. Finally, it set up the campaign so that the relevant creative was served to the right person at the right time. If someone was about to watch a music video, for example, they might have seen music-related creative.
It doesn’t matter how beautiful or effective your ad creative is. If you have a poor mobile site experience, users won’t convert. Top marketers understand the value in having fast, frictionless mobile experiences. With automated marketing, machine learning bidding algorithms automatically drive more customers for better-converting sites. Underperforming mobile sites are at a disadvantage.
We’re seeing brands adopt new technologies such as Progressive Web Apps or Accelerated Mobile Pages to improve the speed and experience of their mobile sites. For example, Alibaba, which already had a great track record of improving conversion rates, launched a Progressive Web App and saw over a 76% lift in conversions.
Top performing advertisers tend to outperform other advertisers on all five of these rules. Imagine competing against an advertiser who is focusing on profit maximization not efficiency, who is focusing on acquiring the best customers, who makes more money for every customer it acquires, whose site converts better and who has better, more-engaging creative.
By: Nicolas Darveau-Garneau