I have bad news for anyone trying to market to me.
There’s a high probability I don’t see your emails. You might have the perfect solution for a problem I didn’t even know I had. But you’re probably ending up in the Promotions tab in my inbox, which I never check.
All the money you spend on search ads for your small business accounting software? It’s targeted at keywords I don’t actually search on. Your organic content doesn’t show up, either.
When I ask Siri to recommend a great pizza place nearby, her database only has your address way out in Beaverton, not the one next to my office. So I will never taste your delicious Neapolitan slice.
And I didn’t install your email app because the App Store listed it below 10 other apps that do the same thing.
Digital marketing is difficult! That’s because robots — algorithms, mostly — are now the gatekeepers to your customers.
Is there any good news? Yes. Marketers who can work with our new robot overlords will be far more effective than they’ve ever been.
Robots are the gatekeepers to your customers
Depending on the estimate, organic search drives somewhere between 50 to 65 percent of all website traffic. When your customer has a problem, their first step in solving it, in many cases, is to Google it.
For example, if I’m interested in buying a paisley pocket square:
- I open google.com.
- I type in “paisley pocket square.”
- Google ranks the trillions of pages in its index according to relevance and displays the result.
As a purveyor of paisley pocket squares, I care the most about step 3. That’s where Google’s algorithms make a decision about whether anyone will see my site, which will have a direct effect on my revenue.
As a result, content marketers and SEO specialists spend a lot of time trying to understand Google’s algorithms. Hundreds of articles have been written about them for Search Engine Land and other publications, not to mention formal studies done by Moz and others. Moz will even show you a history of algorithm updates, together with its historical view of your search visibility.
Successful SEO practitioners spend a lot of time developing their understanding of Google’s software. And they also spend a lot of time adapting their strategies to suit it. For example, the rise of “content marketing” starts right about when Google changed their algorithms in 2011 to reward better content. Below is a graph from Google Trends of interest in the “content marketing” topic.
As another example, take social. In 2016, Twitter announced that it would start curating users’ timelines for them, provoking a medium-sized backlash. But this is common practice among social networks, as the The New York Times noted:
Twitter is not alone in its feed-fiddling. Earlier this month, Instagram began using algorithms to increase the visibility of popular posts, and Facebook has regularly altered its news feeds for years.
The use of algorithms to decide what posts to display is intended to make the networks more useful for their users. But of course, it also reduces advertisers’ reach — unless they pay. In 2016, organic reach on Facebook fell by half, with many publishers simply moving to other outlets or deciding to pay. Smart advertisers understood this and adjusted their mix to compensate.
Two more examples. Paid search, of course, uses algorithms to determine which ads to display, with the amount an advertiser is willing to pay being one input among many. And email marketers have to worry about the algorithms that send their messages to the “Promotions” tab in Gmail, or to Spam.
The problem is even more intense in the app world, because mobile devices’ presentation layer is tightly controlled by the operating system. When an app is installed, you have to ask for permission to send push and other notifications, and iOS or Android controls that experience. And even if you get the user to agree to receive notifications from you, it’s still iOS and Android that decide how they’ll be displayed.
Fighting software with software
So the modern marketer has to understand algorithms, and operating system design, and a bunch of other technical stuff.
Do we get anything in return?
Yes. Digital means software, but it also means numbers. Marketing is easier to target, easier to scale and easier to measure than it’s ever been before. And if we agree to accept the limitations of digital marketing, we can also take advantage of its capabilities.
Take email marketing as an example. I need to understand a lot about how email works in order to use it effectively as a marketing channel. How do I stay out of Spam? How do I make my emails look good in email clients? How do I measure open and click rates?
There is a lot to learn in order to run an effective email marketing campaign. But the upside is that even a very small marketing ops team can keep millions of leads in a database, automate communications to them, set up a huge number of individual segments and respond in real time. And we can test everything. We can understand objectively what performs better — and change our practices to match.
As another example, take mobile marketing automation. There are dozens of vendors out there, and they all solve a similar problem: delivering the best possible communication within the constraints of mobile operating systems. Very basic functionality like message previews is a great example of this; we can use third-party software to help us cope better with Apple’s and Google’s software.
So, think of it this way.
- We use email marketing automation software to help us deal with Google’s and Microsoft’s software (email clients).
- We use social media automation software to help us deal with Twitter’s and Facebook’s software (social media networks).
- We use mobile marketing automation software to help us deal with Apple’s and Google’s software (operating systems).
The algorithms that we need to understand are, however, getting increasingly complicated. In fact, they’re starting to write themselves.
What I wrote above about SEO (for example) actually understates the complexity of what digital marketers have to contend with. With the gradual rollout of Google’s deep-learning RankBrain engine in 2015, the search ranking algorithms aren’t really algorithms. We’re not even sure what they are. From Search Engine Land:
Even people at Google don’t quite understand how RankBrain does what it does, we’ve been told. Honest. But it’s ultimately designed to reward great content.
What does this mean for marketing technology?
Marketing technology vendors — who, of course, are marketers themselves — have rushed to add “deep learning” or “artificial intelligence” in any way they can. Some of these might end up being really useful applications: being able to automate simple interactions with customers through chatbots or predicting which of your customers will churn, for example.
But then there’s the second half of that quote. Ultimately, RankBrain is designed to reward great content. So, if I write something truly spectacular, does all that “playing nice” with Google go out the window?
Or if email filtering improves dramatically through AI, do I need to write great emails, or can I trust that my pretty-good emails will get through to the right people? Can I have a machine write my website content?
I don’t know the answer. I suspect that, at least for now, AI will only intensify our reliance on software, but mostly to help us create the right content, and choose the right customers, in the first place. Good marketers will look for more places to put software, not fewer, so they can focus on being creative and thinking deeply about their businesses.
Some opinions expressed in this article may be those of a guest author and not necessarily Marketing Land. Staff authors are listed here.
About The Author
Justin is a digital marketing nerd and marketing engineer. He specializes in using software and data to execute effective, focused marketing strategies that drive revenue. Justin is currently Director, Marketing Ops and Digital Acquisition at Urban Airship. He’s led digital marketing functions at other fast-growing organizations as well, including enterprise unicorn MongoDB, and Gates Foundation grantee One Acre Fund. Justin has an MBA in Marketing from The Wharton School, and a BA in Ancient Greek and Latin Classics from Columbia.