9 minute read
There’s so much personalized content on the internet that most of us probably engage it without thinking twice about it. Personalization is in action when we cross our Facebook feed, or pick a movie to watch on Netflix. It chooses our music when we use Pandora, and gives us ideas on what to buy next on Amazon.
Personalization is what drives some of the heaviest consumer trends in ecommerce.
Nonetheless, it’s taken a while for personalization to become an accepted marketing tool. 2014 was supposed to be the year for personalization, then 2015, and again this year.
So what makes this year any different?
Better data collection/testing methods are making personalization a science. It has become increasingly efficient, effective, with higher ROI YoY.
Numerous studies indicate that a majority of marketers and executives now believe that personalization is critical to reaching their customers.
And as time goes by, consumers becoming more aware of personalization. A VentureBeat survey found that “77 % of digital natives expect a personalized web experience.”
But just because personalization is becoming more commonplace, it doesn’t mean that all companies can deliver it properly, “Less than 10% of tier 1 retailers believe they are highly effective at personalization, and nearly one-third report having limited or no capability to support personalization efforts” according to a Gartner Survey.
What is personalization?
Personalization is a data driven process that tailors web content for a specific consumer or demographic. Personalized content can be anything from emails, news feeds, advertisements, landing pages, site design, or recommendations.
It works by aggregating data from basic tools such as analytics, cookies, and heat maps. Websites also utilize a vast array of software to track users. In many cases, individual data is compared to similar users in order to personalize content.
Information that’s used to create personalized content usually falls into two categories: Implicit and Explicit data. Implicit data comes from user browsing patterns, and explicit data is gathered from deliberate user action. Implicit data includes metrics like:
- Geographical location
- Page views/button clicks/visitor frequency
- Purchase history
- Time of day/date
- Referring URL
- Type of device
- Keywords used in search
Examples of explicit data include:
- Surveys or website forms
- Signing up with an email address
- Customer Relationship Management (CRM)- communications between company and consumer
- Native login
These data sets are then run through algorithms to create increasingly individualized customer experiences.
The name of the game is targeted convenience. Personalization makes it easier for a consumer to find what they’re looking for, and for site owners to optimize conversion rates.
Types of personalization:
There are multiple mediums of content personalization, but I’ll mention a few notables:
- Email: Personalizing email headers with customer names has been used for a while. But for the most part, just blurting out someone’s name isn’t really a selling point. Better tactics include A/B testing emails using metrics such as time, sending personalized emails “post task completion,” or matching personalized emails with landing pages.
- Advertisements or Personalized Retargeting: This method of personalization depends on user’s browser history to determine what ads they see. Main issues include privacy, and has even been called creepy.
- Chat Function: While this is usually left for companies with higher revenue streams, nothing beats personalization like a real human. Having a chat screen pop up during a site visit may be a good option depending on the level of support that customers need.
- Predictive: This method can be boiled down to simple straight line analytic logic using consumer data. If customer completes action x, they are likely to complete action y. Examples of companies that use this include Netflix and Amazon.
- Segmentation: This method is a great marketing tool in its own right, but can also be helpful for personalization. Segmentation compartmentalizes visitors by metrics such as age, gender, location, or anything else that defines a consumer. Segmentation hopefully achieves a goal of sending the right person the right information at the right time.
- Customization: This is user driven personalization in which website elements like filters can be used to narrow a search, or change preferences on a social media site. These functions help consumers find what they need.
- Weatherization: This method uses the weather in any given user’s geographical location to change featured items. Retailers might simultaneously showcase rain boots to someone in the Pacific Northwest, and t-shirts to someone in the Southwest on the same home page.
Prominent examples of personalization:
Most famously, Amazon utilizes purchase histories and information from similar users to create a landing page that has recommended items of interest. A study shows that consumers found Amazon to be the best at personalized ecommerce, with grocery stores at a distant second. In July 2015, Amazon “eclipsed Walmart as the most valuable retailer in the country,” according to the New York Times.
44 % of US consumers now head directly to Amazon for product research, instead of other methods like Google, according to this 2015 Bloomreach report. It’s a significant leap from 30 % in 2012. The same report found that Amazon posed the biggest “threat” to digital retailers, and that personalization was the top most marketing priority.
Another successful company that utilizes personalization algorithms is Facebook. Their news feed continues to evolve in order to maintain social pertinence. A recent change included what post is at the top when you log in. The social media site hopes to keep people engaged by keeping more relevant content visible to the user, and decreasing less interesting posts.
But personalization campaigns aren’t just necessarily meant for the web.
Remember Coca Cola’s “Share a Coke” campaign?
As it turns out, printing popular millennial names on bottles and cans increased sales by two percent over the summer of 2014. This marketing ploy reversed a decade old slump in soda sales. Although America’s love affair with soda may soon be coming to an end, we still have to give props to Coca Cola’s marketing team. They are known for their historically high success rate.
We’ll see if that pedigree comes in handy for the new “personalized” Diet Coke labeling campaign. 36 core designs will be run through a design algorithm that will alter them slightly, so that no two containers are the same.
Problems with personalization:
The elephant in the room is the combined issue of security and privacy. In the same VentureBeat report mentioned earlier, 96 % of American consumers are concerned with data privacy, and how companies are using it.
It’s clear that data privacy is a huge concern for consumers. Whistleblowers likes Edward Snowden (who exposed a world-wide NSA surveillance program) have left an indelible mark on the American psyche.
This SAS study found, “More than 69 % of our survey respondents agree that recent events in the news have increased their concerns about their data in the hands of businesses.”
When it comes to privacy and personalization, people might opt into it if they had control over what information they share. This study found that “two-thirds [of participants] would like the option for privacy controls. [And] 58% want personalization based only on user information they proactively provide.”
If you’re concerned about privacy, I suggest installing a decent anti-tracking/anti-cookie extension (like Disconnect or Ghostery) to your browser so you can see who is viewing your data.
Another problem with excessive personalization is known as the filter bubble, a term coined by Eli Pariser. Pariser noticed this trend when his Facebook feed showed less posts from his politically conservative friends. In a New York Times editorial, he claimed, “Democracy depends on the citizen’s ability to engage with multiple viewpoint[s]; the internet limits such engagement when it offers up only information that reflects your already established point of view. While it’s sometimes convenient to see only what you want to see, it’s critical at other times that you see things that you don’t.”
Over-personalization may lead us to be closeminded to new opportunities, and makes it harder to engage with ideology that isn’t necessarily our own.
The Future of Personalization:
We are currently at a major junction in personalization.
At the same time that personalization is starting to become more economically viable for startups, large companies are taking the practice into whole new dimensions.
The biggest players in the tech world, cheerfully deemed the “Frightful 5,” have all become vested in personal, virtual, assistants:
These “helpful”, digital, entities can help us find what need with voice commands, and they are poised to start handling the intricacies of our daily lives.
The monolithic companies that created these assistants have a large collection of personal data to their advantage, bringing predictive AI a little closer to home.
But there’s a fundamental flaw to having personalized content from one source alone. As critic Jarno M. Koponen argues, “The almost infinite sea we call the Internet has become a collection of confined ponds with their own walls and rules. Platforms build their own understanding of you, and usually they don’t let you control how your data could be used for your own benefit in other places.”
Koponen fundamentally agrees with Pariser in that too much predictive personalization is stifling. And he also believes that a universal discovery algorithm could blend “goal-driven and casual experiences” by offering more than just an answer to a question.
He describes the process as “ambient” and only running when it makes sense. The system would track user’s cultural preferences and activity, maybe even sense emotions. It would make suggestions that you might have over-looked to achieve a “micro-eureka” moment.
This idea starts to make sense when we look at niche markets or communities that are not always so easy to find. We are often saturated with what we’re already comfortable with, so having user controlled personalized discovery algorithms could not only help people find new and interesting things, but help blend digital and physical personas.
The emergence of more intuitive predictive personalization, and smarter AI, leads us to beg the question:
How much of our personal lives should be determined by data algorithms for the sake of convenience or discovery? and will it ever really feel organic?
But that’s probably best left for another post.
- Personalization is a trending, yet under-utilized service.
- Personalization is allowing companies to achieve a closer relationship with their consumer base.
- Proven methods have only increased in efficacy. Personalized content is reaching wider audiences on every scale.
- Personalization is continuing to evolve, and will do so for the foreseeable future.
Thanks for reading!