Must Love Algorithms: Why Marketers Shouldn’t Fear the Machine
March 9, 2015
I’ve always loved creating new technology. There’s something deeply satisfying about building something new, standing it up, and watching it go into motion.
But even the most advanced technology is just a tool: something to enhance our capabilities — for people to use, for good purposes or ill, for inspired creativity or unimaginative triviality.
Yet new technologies always bring change, and change can bring uncertainty.
I recently spoke at NYC Media Lab’s Personalizationpalooza, alongside a number of peers using algorithms to personalize in different ways, and the question on many people’s minds got asked out loud: what about the human element? Are algorithms a replacement for people?
The answer is no.
It’s true that when it comes to digital distribution and how people experience content, personalization based on individual user interests will outperform human curation. I think the word I used was “crush”, and that’s about right — many of our clients have seen 50%+ pageview or revenue lift from turning on our algorithm.
Why does our algorithm work? Because different people are, well, different. People have distinct tastes and want different things — and no single individual, human or machine, can choose one combination that is best for everybody. Some people like Taylor Swift, some people like Hüsker Dü. (I like both, but not everyone would agree!)
Yet even the best algorithm can only optimize what’s already there. It cannot replace the essential human act of storytelling. Great content comes from great creators.
And that’s why neither marketers nor editorial teams need fear the recommendation engine, which is only a tool — a tool to free up more time for the fun stuff.
Remember all that time you used to spend combing through libraries of thousands — maybe hundreds of thousands — pieces of content, or working to merchandise a digital storefront?
What about all that time you spent creating dozens of segmented newsletters, manually slicing and dicing your audience in an effort to reach them with just the right message?
Now you can spend that time developing creative strategies, creating more great content, and focusing on new ways to build loyalty.
Of course curation, too, can be creative; and when it’s called for, we support that too. Some of our customers have built fantastic user experiences, with a combination of editorial curation in designated slots alongside algorithmic selections.
In the end, the right mix of content is just that, a mix. It can be providing what you know your audience loves based on data, and it can be breaking the top story of the day or launching your new product line.
Personalization technology is never black and white (or blue or gold). It represents an explosion of color, an expansion of possibilities — and we build for you so you can explore them all.
—Ian White, CTO & co-founder at Sailthru
Build an Email Newsletter Everybody Loves
3 Things We Learned from an Audience in Crisis
For years, we’ve collected data to help set benchmarks for publishers’ email programs. We know it’s helpful for publishers to see how their email...
Data & AI
Five Years of Personalization and Consumers Still Want More
This byline was originally posted on MarTechSeries.com When the first Retail Personalization Index was published in 2017, personalization was still a somewhat nascent strategy....