It's hard to get rid of the buzz of machine learning. In fact, every industry is talking about it.
So what is machine learning? according toHewlett PackardSay, "Machine learning refers to the process of computer development pattern recognition, or the process of continuously learning and making predictions based on data, and then making adjustments without special programming." In other words, it is machine analysis and processing a large number of The way information is, and continues to learn and improve over time.
Related:2019 year of study landscape
For an example of a machine learning algorithm in action, let us consider facial recognition-a field we see day in and day out. Today, iPhone users use their facesUnlock their phones.执法 部门Use facial recognition to discover fraudulent activity and arrest criminals. Google Photos allows users to sort photos by their people. These algorithms may not be very accurate in the past, butOver timeThey have been machine learningtraining.
This is not human intelligence, it is programmatic learning, and its scope of application goes beyond facial recognition and cross-industry. Take marketing as an example. Today's marketers are working hard to deliver relevant information to customers. Although humans cannot communicate with a large number of customers on a large scale, machines can. Not sure what it looks like in practice? In this article, I will explain five key uses of machine learning in marketing.
1. Recommend the most relevant products or content
For many years, digital marketers have used product and content suggestions. In the past-sometimes today-these recommendations were manually curated by humans. In the past 10 years, they have generally been driven by simple algorithms that display recommendations based on content viewed or purchased by other visitors.
Machine learning can achieve major improvements through these simple algorithms. Machine learning can synthesize all the information you provide about someone, such as his past purchases, current online behavior, email interactions, location, industry, demographics, etc. to determine his interests and choose the best product or the most relevant Content. The machine-learning-driven recommendations are based on his participation in the recommendations to understand which items or item attributes, styles, categories, price points, etc. are most relevant to each specific person-so the algorithm will continue to improve over time.
Machine learning-driven recommendations are not limited to products and content. You can recommend any content-category, brand, topic, author, reviews, technical specifications, etc. Using machine learning in this way, you can create relevant websites or email experiences that show visitors that you really know them and help them find things they like.
2. Automatically discovers important customer segments
Although machine learning allows you to provide more personalized customized experiences, segmentation is still a valuable tool for marketers. Through segmentation, you can create potential customers or customer groups based on meaningful differences to better understand these groups. Humans can discover obvious differences that they may already know-such as the difference between high lifetime value and low lifetime value customers or new customers and loyal returning customers However, due to the large amount of customer data to filter, there are many other patterns that are not obvious to humans.
The machine helps you identify segments that you are not aware of, and you can use this information to talk to these segments in a more meaningful way.
For example, machine learning algorithms may be able to identify millennials seeking to refinance their homes tend to exhibit certain types of behavior. With this knowledge, you can provide more targeted messaging for this market segment, have a different conversation with the market segment on your website, or talk to an agent over the phone and determine that it may be Other potential customers in the market segment. They show similar behavior
3. Identify potential problems and take action
Your marketing campaign generates a lot of data. Think of all the emails your company sends every day, or the number of people who visit your website, use mobile apps, or interact with your call center. All these interactions generate a lot of data-humans cannot view all the data in time. When something goes wrong-when the link is broken or the promotional code doesn't work, it may not always be obvious to you. Algorithm can filter all data and predictshouldWhat happened and notify you when things don't look right.
For example, suppose it is Black Friday, and one of your emails contains an incorrect link. Machine learning algorithms can predict the click-through rate and/or conversion rate that should be expected from the offer, and alert you immediately if the reality is much lower than it should be. With this knowledge, you can take corrective action before such an important day of the year causes too much damage.
4. From A/B testing to providing individual related experiences and offers
Testing is another area that can be improved through machine learning. Traditional A/B testing allows you to run tests between two or more digital experiences, find the option to produce the best results, and use that experience. This is valuable, but it is one size fits all and does not explain any differences in groups or individuals. Instead, it requires you to choose an experience to show to everyone, which means that many people won't be able to see the experience that works best for them. Machine learning has changed this game.
For example, instead of manually setting up tests between two home page experiences, you can provide the same experience to machine learning algorithms by waiting for the test to complete and selecting the winner. The algorithm will select the experience at the moment it believes it will provide the best results for everyone based on all available information. It will learn from these interactions to inform the next decision.
The same method can be used for promotions and offers. Machine learning not only offers the same 20% discount or static promotion to all customers, but it also allows you to show discounts only to those who need additional incentives. For those who don't need extra rewards, machine learning can choose other related experiences, such as promoting newcomers in the categories they like.
5. Decide how to communicate with everyone
How do you decide when and where to communicate with potential customers or customers? Does she like email? Push notifications? text? If so, how often should you contact her? These are all questions that machine learning algorithms can answer for you.
For example, you can use machine learning to generate a forecast score to determine whether sending this next email to this particular person will cause them to open instead of bulk and explode emails, you only need to send the same to everyone each day. Email, ignore, click or unsubscribe. If so, you don't send it. Instead, you can wait until you have something more relevant to him or her.
Machine learning offers marketers the potential to interpret and process large amounts of information in a scalable manner. In this world, we continue to accumulate more data than we know – and where we want to build personal relationships with customers on a large scale – this is an exciting development. Take a moment to understand how your organization can benefit from machine learning in the near future. Use one of these five areas to start dipping your toes in the water, and then start from there.
This article is reproduced from Entrepreneur, Original address