Personalization is a concept that’s both old and new. In the pre-industrial era, everything was personalized: Clothes were made to measure and furniture was crafted by hand. Industrialization, with its push towards mass production, made personalization obsolete. The internet era revived personalization, as websites and apps can now use software to offer unique experiences to everyone.
What is personalization?
Personalization is the process of tailoring a product or experience to a specific individual. For example, a friend might personalize a book recommendation by using what they know about you to inform their suggestion.
Web personalization takes this idea and applies it to your online experience. Think: When music streaming services recommend new artists to you based on your listening history, or news sites send you curated email digests on topics you self-identify as interesting. The basic formula is that websites or apps collect information about you, and then use that information to tailor your experience.
In addition to web services, online ads can be personalized, too. The ads you see online can be personalized based on what a web service provider knows about you (based on your activity, which is tracked) or what they’ve inferred about you.
Personalization vs. customization
In the digital sphere, personalization is generally used to describe things that happen automatically, whereas customization refers to options that the user must take. For example, if an e-commerce site remembers your size preferences and automatically shows you items in your size, that's personalization. When that same site offers you options to filter by size, that’s customization.
How does online personalization work?
There are many different kinds of personalization experiences, powered by different kinds of technology. Some of the ways in which websites and services create personalized experiences include:
- IP address. Many websites and services use your location to tailor your experience. For example, search engines can use your location to deliver personalized results for general queries like “tacos” or “shoes,” highlighting stores that sell those items near you. How do they know where you are? Your internet service provider (ISP) sends your IP address, which is tied to a geolocation, to every website you visit. An IP address is kind of like a zip code, in that it provides your general location.
- GPS. If you’re using a GPS-equipped device like a smartphone, apps may be able to access an even more precise location, thanks to location services.
- Machine learning. Machine learning is a branch of artificial intelligence (AI) that allows computers and software to improve their performance based on new data. Machine learning has been a boon to personalization, allowing computers to quickly learn about users and personalize their experiences in real-time. It’s used to rank search results, show product recommendations, and even choose what song you listen to next.
4 Examples of personalization
Thanks to personalization, every internet user experiences the web in a slightly different way, without necessarily realizing it. In 2000, Jeff Bezos, founder of Amazon, said, “If we want to have 20 million customers, then we want to have 20 million ‘stores.”’ Tony Jebara, VP of Machine Learning at Spotify, echoed this sentiment almost two decades later, saying, “There isn’t one Spotify. Really, there are 230 million Spotifys, one for each user.”
But is that a good thing? Personalization can certainly save you time and energy, by presenting the media and products that are most relevant to you. But personalization also creates echo chambers, which are increasingly concerning. Sridhar Ramaswamy, cofounder of Neeva, believes that the key to personalization is agency. “Agency means that you are in charge of the personalization, and it's relatively easy for you to turn it off, if you choose,” he explains.
Since personalization isn’t always transparent, here are some of the ways in which your day-to-day experience of the internet is personalized just for you.
1. Personalized content recommendations
Personalized content recommendations are tech companies’ version of a book or movie recommendation from a friend. But instead of coming from a human friend, these recommendations come from algorithms.
- Your viewing history, including how long you spent watching the content
- Your ratings history: whether you gave something a thumbs up or thumbs down
- The time of day
- The type of device you’re using (computer, TV, smartphone)
- What other people with similar tastes (called “taste communities”) have watched
Netflix even uses what it knows about you to show the most relatable stills in its preview images. Each movie or TV show has several preview options, and your viewing habits will dictate which one shows up on your screen. For example, if you watch a lot of rom-coms, the preview image for the TV show Outer Banks might be a couple embracing, whereas if you gravitate towards sports movies, you’ll see surfers running with their boards.
Recommendations engines drive an incredible amount of streaming activity. In 2015, Netflix’s recommendations accounted for 80% of streaming, and the company valued the impact of personalization and recommendation at $1 billion.
Music streaming. Spotify uses machine learning algorithms to scan its 70 million songs to find ones suited to your tastes. When opening the app, every user is greeted by a unique landing page featuring personalized music recommendations, which are updated in real-time based on factors such as:
- The music and podcasts you listen to
- Your likelihood to accept recommendations
- How long you listen to different tracks (If you skip a track within 30 seconds, Spotify assumes you didn’t like it.)
- Demographic information like your location, age, and gender
In addition to your personalized landing page, Spotify also offers personalized and curated playlists which, in 2017, accounted for 31% of listeners' streaming time. Unlike Netflix, Spotify’s recommendations feature a mix of both music you’ll probably like, and some that is more of a stretch. According to Spotify, in 2020, 68% of listeners discovered new music thanks to Spotify’s algorithms.
Social media. Social media feeds aren’t usually presented in chronological order—they’re prioritized by algorithms that track your behavior. Additionally, many types of social media feature "explore" or "discover" sections where you can find new accounts to follow, based on the content you’ve already interacted with. For example, when you “pin” an image on Pinterest, you’ll start seeing recommendations for similar images, and the more you pin, the more Pinterest can improve its recommendations for you. Exposing users to personalized, relevant content is at the crux of Pinterest’s utility, but it also raises concerns about echo chambers or “filter bubbles,” where you only see what you want to see.
2. Personalized product recommendations
Product recommendations from retailers are just like content recommendations, but the difference is you may end up buying something.
Amazon shoppers, for example, can view a digital storefront personalized based on their shopping history and demographics, and estimated shipping times based on their location, which helps narrow down the massive amount of products available from the website. These product recommendations could contribute to the speed in which purchases are made: 28% of Amazon purchases are made in three minutes or less, and half of all purchases take less than 15 minutes.
3. Personalized search results
If you and a friend type the same terms into the same search engine, you might get different results. Why is that? Most modern search engines use what they know about you to tailor your results.
- Location. Using your IP address or GPS, search engines can easily identify your location and use that information to provide you with local weather results when you search “weather” or local bars when you search “cocktails.”
- Search history. Your search history can help your search engine know that different searches may be related—so, if you recently searched for information about soccer, and then looked up “Barcelona,” your search engine would show results about the soccer team, not the city.
4. Personalized advertising.
It's easy to see the benefits of personalization, but personalization without knowledge or consent can feel intrusive, especially in the case of personalized ads. Also known as behavioral ad targeting, these advertisements use behavioral data points to serve ads that you are more likely to click on. One particularly invasive subset of targeted advertising is retargeting, which involves tracking your actions across the web to show you ads for products that you have previously viewed or placed in a shopping cart.
Personalized ads are a clear win for advertisers, resulting in higher click-through rates and conversion rates. But for internet users, the benefits aren’t as clear. Personalized ads can feel creepy, and they have created backlash to the point that websites now must get consent to use ads personalization, per the European General Data Protection Regulation (GDPR).
Neeva is the world’s first private, ad-free search engine, committed to showing you the best result for every search. We will never sell or share your data with anyone, especially advertisers. Neeva is currently in an early testing phase. You can sign up to be on our waitlist at neeva.com/signup.