Algorithms: What You Need to Know and How to Beat Them in 2020

You hear the term algorithm when it comes to social media or search frequently. But what exactly is an algorithm?

What Is An Algorithm?

In its simplest form, it’s a set of rules or instructions that tells computers or software what to do.

When coupled with Machine Learning, it can be powerful. Algorithms are the reason your Facebook feed looks different than your friends. By analyzing past behavior and engagement, the social network tries to anticipate the things you’ll enjoy seeing the most and serve them up.

Algorithms are what powers Netflix’s recommendation engine that is responsible for 75% of all viewing on the streaming video service. It’s what powers Amazon’s powerful recommended product display that encompasses more than a third of all consumer purchases on the online shopping platform.

How Do Algorithms Work?

When you search – whether on Google or within a social platform – the algorithms go to work. They take the words you type and compare it to billions of other interactions to look for patterns to predict what you want and then serve up the most relevant answers it can. Algorithms have subroutines that examine your intent, rank web pages or social profiles by quality and usability, and determines context before delivering your results. There are hundreds of ranking factors the search engine sorts through as it decides what to show – all in a fraction of a second.

It uses these ranking signals to find the content that matches your search intent as much as possible. That’s why it’s critical to understand these ranking signals and design content that plays nice with algorithms. It’s not enough to create great content, it has to be done in a way that is algorithm-friendly.

“Just because you post it, doesn’t mean people actually see it,” -Taylor Gaines, Owner & CEO of Social Gaines.

How Algorithms Work On Each Platform

Each platform’s algorithm will filter content in a different manner. However, they all have the same goal in mind: surface content you will find most valuable. The better the algorithm does the job, the more likely you are to engage with the content and the platform. 

Facebook

There has been a major shift in how Facebook’s algorithm works in the past year. CEO Mark Zuckerberg decided to prioritize meaningful conversations from family, friends, or groups. Closer relationships will matter most. If someone has a personal connection with a person or brand, the likelihood increases their posts will be seen.

While the currency on Facebook had been "likes," that’s changed as well. Facebook’s algorithm will prioritize interactions such as shares, comments, and reactions over likes and click-throughs. It also takes into account time spent on content (engagement), content type, quality, and relevance to individual users based on past interactions.

Facebook prioritizes content that is more likely to encourage engagement.

1.    Video can increase engagement by as much as 50%

2.   Posts with images perform better than posts without images

3.    Posts without media are typically ranked lower than those with video or images

Facebook ranks content based on four factors: Inventory, ranking signals, predictions, and relevancy score.

Adam Mosseri, former VP of Facebook’s New Feed, explained that Facebook looks first at the available inventory to choose from. It then assesses the ranking signals for each piece of content to predict performance.  

“Predictions are things like how often are you likely to comment or share a story,” Mosseri said.  “Or how likely you are to hide or report a story.”  

The algorithm favors more current content, but if you post too many things too closely together, you may undercut your chances of having multiple posts distributed.

All of these predictive analytics roll up into a relevancy score. These scores act in competition with other posts in the inventory. The post that rank the highest are shown more often. 

Twitter

More than 6,000 tweets are added to Twitter every second. That’s 500 million a day or roughly 200 billion tweets a year. How does Twitter know which ones you want to see? Algorithms.

Twitter’s algorithm values recency, engagement, and location. It places a value on how many followers an account has and folds that into the algorithm when choosing the content. More so than other platforms, it prioritizes rich media such as video, images, GIFs, and polls. Tweets with video get 10X the engagement. GIFs get 55% more. Use rich media to beat the Twitter algorithm. 

Instagram

Instagram’s algorithm priorities relationships, interest, and timeliness. Engagement by you and others plays a significant role.

Instagram feeds don’t just show content based on whom you follow. Its algorithm tries to determine whom you like and what you like to see. Instagram’s algorithm will factor your past behavior interacting with similar content.

When content reaches critical mass on Instagram, it can go viral quickly. High engagement triggers broader distribution which can, in turn, lead to even more likes.

LinkedIn

Pete Davies, Senior Director of Product Development at LinkedIn, describes its algorithms’ goal this way: “People you know, talking about things you care about.”

Each time the app is opened, the algorithm checks for recent posts made by your connections. The closer the level of connection you have, the more likely you will see something from that person. It will sort through people, pages, groups, and hashtags you follow. It considers which content you have clicked on, liked, and interacted with in the past. Engagement with your content is key in increasing your visibility on LinkedIn.

Pinterest

Pinterest doesn’t get the mainstream media attention it used to, but more than 250 million people use the online platform every month. It’s still growing, too. More than 5 million new people signed up last year and it reaches 84% of US women aged 25-54.

Its algorithm sorts through more than 175 billion pinned items users have followed. It searches accounts and boards you follow and ranks account authority, domain quality, pin quality, and interactions from each source. More so than other platforms, it focuses heavily on recency. Recent activities such as searches, repins, and engagement play a significant role in what’s shown.

Visual image recognition is also a big factor in Pinterest's success, Google has even been known to copy this feature. How it works is “users can isolate items in pins and search them visually to find similar objects and shop them online in the Pinterest app.” This technology is so advanced that users can now take photos or upload their own images to search for products. Pinterest is often disregarded as just a social media platform, but really it is one of our most commonly used search engines of our time. 

How is GDPR going to Impact the Algorithms?

Social networks and online platforms are built on knowing highly personal information about individuals to serve up relevant content. That can be in direct conflict with recent data privacy laws, such as GDPR (General Data Protection Regulations) in the European Union, CCPA (California Consumer Privacy Act), and pending legislation all over the world. Algorithms will be forced to change to comply with these new rules or their owners will face substantial fines and penalties.

These changes mean algorithms may have less access to the type of information used to serve targeted ads and relevant content. If you want to make sure your content gets seen, you need to invest in paid placement, engaging content and invest in building an authentic audience to drive your virality score. 

How to Beat the algorithm in 2020

The secret sauce to beating the algorithm is simple - build an authentic audience, and deliver value-added content… then put some $$$ bills behind it to amplify your reach to capture new audience members and repeat the cycle. If you have questions about how to build your social road map in 2020 - let’s talk!