This post was originally published in California Mortgage Finance News, a California Mortgage Bankers Association publication.
“I believe that at the end of the century the use of words and general educated opinion will have altered so much that one will be able to speak of machines thinking without expecting to be contradicted.” These are the words of Alan Turing in a paper published back in 1950 entitled “Computing Machinery and Intelligence.”
In 2012 Target identified a young woman’s pregnancy before her father knew. They sent coupons for baby cribs, maternity clothing and baby clothes addressed to his daughter, who was still in high school. The confused and angry father went to speak to a manager at Target about the impropriety of marketing in this fashion to his young daughter. Days later, he found out that his daughter was, in fact, pregnant.
How did this happen? Target assigns a unique code to each one of their customers. This code is a “bucket” that carries data analyzed in real time and gets monetized through highly personalized promotions at the individual level. Every time you use your credit card at their store, call a customer service line, open an email or go onto their website…it all lands in your personal data bucket.
Combining this data with your entire purchase history and that of millions of customers, deep analysis uncovers patterns of behavior, trends and associations. This information helps artificial intelligence (AI) programs figure out what and how to market to you.
Another example…look at Amazon. Today, about one-third of Amazon’s business comes from “recommended purchases,” This feature is powered by a highly intelligent algorithm that makes suggestions to you based on how you’ve interacted with their site, the purchases you’ve made and other data associated with your browsing habits.
These are both examples of the application of artificial intelligence marketing (AIM). The more data you can feed these programs, the more they learn and the more accurate they become.
Artificial intelligence marketing defined…
The best way to describe it is the leveraging of big data that, when analyzed deeply, reveals patterns, trends and associations in consumer behaviors. Algorithms learn from the data and predict a consumers’ wants and needs and provides information or services accordingly. The more data it receives, the more it learns and the more accurate it becomes. This outcome mimics the intelligence of human beings, hence “artificial intelligence.”
Forward-thinking companies in the mortgage industry know they need to understand and embrace this technology to be competitive and sustainable in the years to come.
Artificial intelligence is no longer a buzzword, but is a hot boardroom topic. Forward-thinking companies in the mortgage industry know they need to understand and embrace this technology to be competitive and sustainable in the years to come.
Let’s look at two big areas that are being deeply affected by AI technology…
SEO (Search Engine Optimization)
Right now, as soon as you’re done reading this article, the power of AI is at your fingertips!
Google has added a strong component to their overall ranking algorithm called RankBrain. RankBrain is a machine learning AI system that has dramatically changed SEO. Machine learning is a component of AI which is simply the ability of a software program to learn from experience and produce smarter results over time.
In the past, one of the keys to marketers winning the SEO game was to identify keywords or keyword combinations searchers use when learning about a specific topic like home financing, mortgage CRM, etc. Then they would develop website content and blog posts, dropping those exact keywords and keyword combinations in the title and as many places as possible within the copy, sometimes to an annoying degree.
To stay ahead of their competition, marketers pushed the envelope and found creative ways to stuff keywords into websites and blogs along with low quality backlinks and more. This led to what became known as black hat SEO techniques designed to manipulate search engines to achieve ranking. These techniques landed up hurting legitimate sites and causing searchers to lose confidence in their results.
As a result, search engines like Google built countermeasures into their algorithms to punish these techniques and reward the good guys.
Now, with the proliferation of voice search, an AI technology (think Siri, Google, Cortana), SEO has changed forever. When you search by voice, you tend to use more colloquial phrases. People are much less patient now with trying different keyword combinations to get the results they’re looking for. Accordingly, search engines have placed more emphasis on natural language search optimization.
Enter Google’s RankBrain, a deep learning algorithm, that can take a natural language search query and break it down into its most important components to drive the best results. If it doesn’t recognize a word or word combination, it will look for words that have linguistic similarity before it returns the result. Based on how millions interact with the results, it will learn how to provide you with exactly what you’re looking for!
Try it now! Search for “how do you get money to buy a house” and see what happens. Here’s another one…” how do I make loan originating easier.” I purposely picked two phrases that are a bit more open to interpretation and left out words like “mortgage,” “down payment assistance” and “LOS.” In the results however, you’ll find the information you’re looking for and maybe some results you didn’t expect. When you click on the result you like best, the algorithm learns as it records input from you and millions of others.
What does this exercise teach us?
The key to doing well in search these days is simply writing great content. No need to keyword stuff or write unnaturally to gain ranking. If Google’s algorithm learns your writing is unnatural, it’s only a matter of time before you get hit.
Also, focus on strong incoming links and a high-quality website. Google also knows quality. It will compare your site to the best in your category and use that as an additional ranking factor.
Advertising Campaign Development & Measurement
Great strides have been made in how advertising campaigns are developed, executed and measured. As you probably figured out by now, developing general ad campaigns targeting broad buyer personas and using analytics that only consider clicks and impressions is antiquated.
With AI technology, you can get to the bottom of ”why” a user clicked on an ad and leverage that information for better targeting. This “why” is discovered through a deep analysis into that users search history, interactions with display ads, videos, previous contact with your website, social media interactions, historical downloads and even third-party data sold by companies like Acxiom, Bluekai, Datalogix and others.
Armed with this data, AI platforms can make instant decisions on the type of content (video, display ads, infographics, charts, etc.) to serve up along with creative and messaging. They can create propensity models to more effectively target the most relevant prospects and their propensity to click on any piece of content and convert.
AI platforms are learning to highlight creative and text elements that are working and not working within various content types to optimize your targeting strategy.
In the case of a mortgage lender, this saves money and a lot of human time manually figuring out how to drive a user to your virtual front door to start the mortgage application process.
AI is not just about processing data based on hard-coded rules, but learning from that data as well so your campaigns can be “smarter” tomorrow than they are today, providing an invaluable lift to your marketing efforts. This is why Forrester predicted a 300% increase in investment in AI in 2017 over 2016. That’s a big number!
How can you take advantage of AI in your marketing?
The good news is that you probably already are and may not even realize it! Here are some examples:
- Search: As a marketer, you may have already changed how you develop content that you publish simply by understanding that Google and other search engines now understand your intent. For example, in the old days if you’re a mortgage company in Los Angeles, you would feel compelled to use the keyword phrase “Los Angeles mortgage company” in your content. Now, that’s no longer necessary. If your prospective borrower lives in LA and is looking for a mortgage company, they simply ask Google “where can I get a mortgage” and up pops a list right in their area.
- Social media: Maybe you’re using Facebook’s ad platform. If you are, you’re leveraging the power of Facebook’s AI to make sure your ad or sponsored post is only seen by the most relevant audience so that you don’t waste money having it served to people who are not interested. Same thing goes for LinkedIn and other social media platforms.
- Advertising: Have you used Google AdWords lately? If you have, then you’ve already used Google’s AI in the form of its automated bidding system to capture results from searches.
You can also use it for curating individual newsletter content, marketing automation, chatbots and much more as well.
Now is the time to think more deeply about AI, investigate how you can do a better job of leveraging it and expand your knowledge base. This will be key in preparing yourself to compete effectively in the coming years!