Machine Learning: What Will it Mean for Marketers?


The next wave of computer technology is upon us – and it’s Machine Learning. And if businesses want to succeed in the marketplace of tomorrow, they had better understand what machine learning is and how it can be applied to their business.

The danger of waiting is that your competitors, or some future company that does not even exist yet, will use machine learning to disrupt your industry and drive you out of business.

What is Machine Learning

Machine learning is a field of computer science that gives computer systems the ability to “learn” with data, without being explicitly programmed. It is an extension of artificial intelligence, which uses pattern recognition and computational learning to help solve problems better than humans ever could.

To most of us who are not actively studying or working in this field today, it might sound a little like science fiction. But computer technology is racing forward, with very real applications of machine learning already in place today.

Why Machine Learning

The reason that machine learning is so popular right now is a combination of the technological progress made in recent years and the potential it holds for a large swath of fields. For companies who are going to lead the way on innovation, the opportunities to deploy machine learning across their organizations is enough to excite everyone from board members, to executives, on down the line.

Machine learning will make decisions – both day to day and long-term strategic decisions – faster and better than humans. It does this by analyzing all of the data available to it using algorithms that get smarter over time.

What Does This Mean for Marketers

For marketers, machine learning will mean that some of the most complex parts of our job will get much easier. Machine learning might be used to:

  • Optimize who we target with our advertising, by analyzing who is most likely to become a paying customer on an individual basis
  • Write promotional copy that speaks directly to each person in our marketing funnel, by understanding the right combination of persuasion and emotion to use to convert new customers
  • Identify customers who are most likely to become repeat buyers and intervene with special offers at the right time to advance their loyalty
  • Intervene with users who we are at greatest risk of losing to competitors with a promotional offer that keeps them coming back
  • Recommend new products and services that our customers are most likely to purchase from us based on customer behavior and competitive offerings
  • Determine the appropriate price of everything we offer to maximize sales or profitability, including when and where to use promotions to spur greater levels of activity and address local or seasonality slowness

The above list just scrapes the surface of what might soon be possible with machine learning. To take full advantage, companies need to invest in research, data collection, training for current staff, and hiring experts. The time to start is now, because the sooner you can make strides in this field, the larger your advantage over the competition will be.

What Do You Know About Your Customers

Here is the simple truth – the more you know about your customers, the more successful you will be as a marketer or small business owner.

  • Knowing your existing customers will help you find more people like them
  • Knowing your existing customers will help you create products they want
  • Knowing your existing customers will help you provide better service

What should you know about your customers? In short, as much as you can.

You should know the things that relate to your business – how they found out about you, why they chose you over your competitors, how they use your product or service and whether it solves their problems, how they purchased, when and how much.

You should also know things that don’t relate directly to their interaction with your business. You should know things like where they live, how old are they, what do they do for a living, what are their hobbies and interests, where do they get their news, where do they shop, what else do they spend money on.

The world of big data, machine learning, and artificial intelligence will allow companies of all sizes to take advantage of information to improve their business. Marketers will have new ways of reaching out to potential customers with proven ROI. And with more data, those efforts will be more fruitful.

If you don’t know enough about your customers now, today is a great time to start learning. Collect data through user behavior tracking, focus groups and studies, surveys and questionnaires, interviews and direct customer outreach. Marketing departments should put someone in charge of customer knowledge and data analysis.

You may be surprised by how much information you already have on your customers. Start asking questions internally, streamline and protect that information for later use, find the holes and begin to fill them.

Why Marketers Should Care About Privacy

People the world over are up in arms over the lack of privacy individuals have in the digital age. And while some of what they are saying gets lost in translation, it ultimately boils down to the fact that it seems that no matter what we do, someone is watching us – whether that be the government, the banks, or advertisers.

What ever happened to privacy, anyway?

As marketers, we want data. To get that data, we take advantage of new technologies that collect that data from consumers, sometimes safely and other times a little less so.

We might say we care about privacy, but we’ll do almost anything to get data if we think it will help improve our numbers. And so we easily cross that moral line in the name of personal success.

Should we care more than we do?

I think the answer is yes. We should care because we are in the business of trust. If consumers can’t trust us, can’t trust what we’re telling and selling them, then we have no chance of success. And collecting data we have no right to is a surefire way to lose their trust.

How do we keep the trust?

I don’t mean to say we should stop collecting and using the data available to us. That would be a far greater sacrifice than is necessary. But we should be clear with people what data we collect and why. There are benefits for consumers when we collect their data. And they should know what they’re getting out of the deal as well, so that they can make the right decision about whether to share that information or not.

The Problems with Analytics

Data is a good thing. The more we know about our customers, our website visitors, our business processes, the better off we should be.

But there are problems that come with analytics. If you are working in a data-driven company, or are in the process of changing over to a data-driven culture, be sure to avoid these common missteps.

1. The data is wrong.

The first problem that many companies encounter is that the data they are collecting and analyzing is incorrect or incomplete. Often you’ll find that there are gaps in your data, or things don’t match from one system to the next. Then you have to go back and try to piece your data together manually to get a more accurate look at your business.

Sometimes, it is not immediately obvious that the data is wrong. So you start using it to make important business decisions. And you don’t find out until it’s too late that those decisions were made looking at incorrect information.

You need to be very confident that your data is correct and complete before relying on it to make important decisions that will impact the future of your business.

2. You’re using the wrong metrics.

Know what problem you want to solve or what question you want to answer before you start getting too involved with data analysis. It is far too easy to spend a lot of time and energy analyzing one metric or set of metrics, when it’s something else entirely that should be commanding your attention.

Just because you are able to see something, doesn’t mean it’s important. Be sure to prioritize your data analysis based on the impact it can have on your business.

3. You’re missing the big picture.

Analysis paralysis is real, people. It’s when you get too bogged down in vast world of analytics and are unable to pull yourself out and look at your business as a whole.

Sure, it’s easy to think that data will solve all your problems. If we improve this metric and that metric, the business will naturally improve with it. But if you start thinking on a smaller scale, focusing too intently on the numbers, you may find yourself unable to see the forest for the trees.

How to Know More About Your Customers

It is my hope that yesterday’s post on data convinced you of the importance of collecting and mining customer data, and encouraged you to think in new ways about the data that you already have.

But what if you don’t have the data you want, or need, to really make a difference?

You need to figure out how to get it. And there are a number of things you can do to change the way that you collect data and get more of what you need.

  1. Ask your customers – through a combination of surveys, sign up forms, and checkout processes, you can collect a lot of information directly from your customers. They are your best source of data, especially when they willingly provide it to you. Make sure that you have the systems in place to collect and store all this data, trackable back to each individual customer.
  2. Add third party data – there are a number of large companies whose primary business is collecting data on consumers at the individual and household level. If you can afford it, you can marry their data to your customer list, adding new pieces of information that you might not be willing or able to collect directly.
  3. Track web activity – even the most basic website analytics platforms give you thousands of different pieces of information about the behavior of people on your website. And you should be able to connect your customer database to your analytics platform to tie web activity directly to individual customers based on an IP address or login.
  4. Track purchases – some companies struggle to get the data they need because they’re not tracking the things they should be. All purchase behavior should be collected and stored with the customer data, including but not limited to the date, products purchased, amount spent, checkout method (in store, online, phone, etc.).

The key to knowing more about your customers is to identify what you know already, figure out what you want to know that you don’t already have, and then find the easiest way to get it. Often times, when you get to that point, you’ll find that the information you’re after is easier to get than you expected.

So stop waiting and start collecting.