The second machine age is here. Algorithms are doing many of the jobs that used to be done by humans. In their book ‘The Second Machine Age’, Andrew McAfee and Erik Brynjolfsson say: “Computers are going to continue to improve and to do new and unprecedented things.
“The transformations brought about by digital technology will be profoundly beneficial ones. We’re heading into an era that won’t just be different; it will be better, because we’ll be able to increase both the variety and the volume of our consumption.”
The relationship between Big Data and automation
Data is the catalyst for automation.
Think about car insurance companies. How are they able to price things so quickly when there are so many parameters to consider? The reliability of a car, its age, the age of the driver, their style of driving, location, experience; it must all be taken into account.
Insurance is all about risk. The lower the risk, the less you pay. The high risk drivers pay for the lower risk drivers. There’s no way a team of human beings can calculate the right quotes for the right people. They need data.
When car insurance first existed, there was no data, only assumption. Maybe there was a single price for everyone. Now, there is a ton of big data insurance companies can dip into.
But they don’t do it one quote at a time – they use algorithms. A computer program can run thousands of queries on a database simultaneously, and an algorithm can carry out thousands of transactions before one human has finished making their morale-boosting cup of coffee.
It’s because of the mass of data we have been able to collect about shopping habits, earthquakes, the weather, or even the quality of wine, that we are able to create software to automate thinking.
Will marketing managers be replaced by automation?
A robot is not likely to take your job soon. People who work in the marketing industry are creative, generally, and they are not in the white collar catchment that is likely to be hit earliest by the second machine age.
The first in line for robot replacement are those who carry out repetitive mechanical tasks that could be handled by automation. I’m talking about secretaries, customer help desks, accountants, even financial advisors. These are all people who do a lot of the same things repetitively, following the same processes. Where the task is driven more by repeated process than by unique creativity, automation is a natural evolution.
That’s why we don’t see telephone exchanges these days being manned by teams of people pushing wires into holes to connect conversations. It’s all done by a computer program.
It is possible to automate marketing, if you really want to, but let’s not get confused. What we know as marketing automation is not true automation. That’s programmatic. When I talk about automation, I’m talking about algorithms doing the thinking, the planning and the execution.
Let’s look at some of the places where automation could play a role in marketing.
Subscriber list management
Remember that time before the existence of Mailchimp, Aweber and Campaign Monitor? Before there was an online email service provider to manage your mailing list? You would probably use Microsoft Access or a server-side CRM system to store your data.
Before we had computers and databases, mailing lists were probably kept on pieces of paper and direct mail would involve a typist who would create many labels or type addresses on envelopes.
Now, we can upload our mailing list to an ESP, and connect our website so that people can manage their own subscriptions. We can create email newsletters that send themselves at the right time.
What comes next? Algorithmic management of your lists. I bet you have a way of managing sales leads in your organisation. That requires some kind of manual management to keep changing tags or adding notes. What if all of that could be automated and a computer program could just tell you, “Call these people today and tell them this.”
Media buying requires two things. First, you need a good knowledge of the media landscape and its relative rate cards. Second, you need to have a good knowledge of the kinds of returns you might get from each media channel.
The best media planners are people who have done it for long enough to gain the integral knowledge needed to spend money in a way that maximises return. An algorithm could knock spots off even the best media planner.
If you had all the data about each media outlet, including rates, you would be able to spread your budget automatically across the marketplace using parameters. Imagine you are buying media for a TV advertisement. All you would have to do is say what kind of audience you want to reach and the algorithm could do the work for you. Just like the insurance company measuring risk based on all the data, you would be measuring opportunity based on all the data.
Over time, if you feed back response rates in order to measure ROI, you will be able to track which media channel, and which media outlet, brings in the best ROI. You can then tweak your algorithm to make smart decisions based on the likelihood of a good return.
The idea of automation in event management intrigues me. Every event requires a number of set parameters and behaviours that are repeated.
There’s the promotion – gathering a list of attendees and marketing to them, which you can automate.
There is also the event itself – imagine if you could track the movement of all attendees to see which stands they visit and for how long, and correlate this with who they speak to.
Measuring the value of exhibition stands by tracking footfall
Event organisers would be able to find out, based on data rather than assumption, the real ROI of the event, venue and specific stands.
Video creation and optimisation
Netflix already has the ability to create popular programmes based on the data it collects from viewers. We are probably only a few years away from automated editing becoming the norm.
It’s possible to run a whole broadcast channel using automation of playlists. What if we could upload a video and have an algorithm edit it, adding music to match the pace? Would there be a need for such a thing? In advertising, perhaps.
Consider Unruly, a company that was recently purchased by Rupert Murdoch’s empire for £114 million.
Here’s how the company describes itself: “Differentiated by a unique data set of 2 trillion video views and powered by a full tech stack, Unruly adds value by algorithmically evaluating content shareability and programmatically targeting custom audiences.”
Call centre management
Robotic call centres are not exactly loved by many people. What’s the point of telling a computer your password and date of birth only for a human to then ask for the same information?
Some call centre scripts are so badly designed that they create more bad feeling among people who are perhaps already unhappy. The failing here, though, is in the design, not in the idea of automation.
Call centre software is clever. Algorithms can filter emails based on language, making decisions on where to send messages. These systems are enhanced by humans, who can teach the algorithm when it makes mistakes.
If you are a call centre worker, expect the second machine age to change your job.
Running a survey
Remember focus groups and on-the-street vox pops? They still exist, but they used to be the only way to run an opinion poll. Now, you can do that online, and have the analysis delivered to your inbox without you having to think about it.
Who hasn’t heard of Survey Monkey? Google also offers a consumer survey service, where you can carry out research or a focus group for a relatively low fee with the click of a few buttons.
Reporting and analysis
This is the one area that will probably see the fastest change. How many hours do you spend poring over Google Analytics or sales figures, looking for patterns; highs, and lows?
Imagine how much creative stuff you could do if you weren’t chained to Excel trying to crunch some numbers in order to present progress to your board.
There are some great tools available such as Google Analytics, Spring Metrics and Kissmetrics to do the job for you
So, can a robot be a marketing manager?
Maybe one day, but there’s a lot of ingenuity and creativity involved in marketing. There’s also a lot of manual labour involved in the day-to-day marketing manager’s role that a robot could do, though. All you need is the data.
The starting point is to gather the data then turn that big data into action.
Automation is not about replacing specialists. Instead, it helps us achieve so much more.