You can’t go anywhere online without being at the mercy of algorithms. They rank websites on Google, give you recommendations on Netflix, and show you the most important photos at the top of your Instagram feed. Gone is the need for manual methods of curating and recommending products, services, and content.

Or is it?

As the business world becomes more digital, algorithms and automation are replacing humans in what were once very manual, often specialized tasks. For small businesses, the threat of automation replacing expertise looms heavily for those not prepared to introduce data-driven processes into their customer service offering.

However, automation is not for everyone. Small businesses* will thrive on the knowledge and experience of experts to cater to the customer of the future. They’ll be able to leverage these experts to cater to a niche audience with unique needs.

While midsize enterprises will benefit from algorithms to help bring their products to a larger market, small businesses will succeed by having real-life experts that can provide tailored services where algorithms and automation won’t cut it.

According to a recent Gartner report (available to clients), there are two types of business models that’ll benefit from algorithms and two that’ll fare better with experts.

These four business models are:

I’ll go through each business model below to show that algorithms don’t always trump experts, especially if you’re a small business.

Algorithms vs. Experts: Which do you need?
Scroll to the very bottom of the article for a summary chart outlining the questions you need to ask to see which one is better for your business.

*According to Gartner’s definition, small businesses are organizations that either have fewer than 100 employees or make less than $50 million in annual revenue. Midsize enterprises have between 100 and 999 employees or make more than $50 million but less than $1 billion in revenue.

Wait… what’s an algorithm?

Let me set the context real quick by defining exactly what an algorithm is.

In its most basic form, an algorithm is a set of instructions on how to perform a certain task. If you search for an Airbnb in Paris, you get a list of holiday apartments for rent in Paris. If you want an Airbnb in Paris in July, then you won’t see apartments that are fully booked in July.

Algorithms can be combined in a number of ways to deliver users with what they’re looking for (or what they didn’t even know they were looking for). As machine learning starts to kick in, these instructions can get more complex and more tailored to individual users, pulling in cues from tons of different sources to deliver the most valuable results.

According to Gartner, an algorithmic business then, is “the industrialized use of complex mathematical algorithms pivotal to driving improved business decisions or process automation for competitive differentiation.”

Customer data is an important source for improving algorithms to help deliver customers with a better experience— from ads and recommendations to bug fixes and user experience improvements.

Experts, on the other hand, will be at the center of a business where knowledge and human interaction must work together to be able to deliver a product or service effectively.

Whether it’s better to use an algorithm or an expert depends on your business model. Below, we’ve listed the four business models defined by Gartner, whether each should use an algorithm or an expert, and which size of business it’s best for.

Businesses that will use algorithms

Programmatic businesses

 Best for: Midsize enterprise 

“In the programmatic business world, algorithms displace experts, but the consumption experience remains one of individual ownership, or individual use of a product or service.” — Gartner

Programmatic businesses are those that use a ton of customer data to provide an individualized and tailored product or service to users. The product or service is improved using algorithms that take advantage of this data to make the offering more valuable to customers. Without it, the value would be significantly reduced.

Consider service outages as an example. If Netflix is down, the company will notice a change in customer usage that’ll help them spot the error. The number of customers it has, as well as historical data about previous technical issues, means Netflix can quickly spot and resolve the issue. It may even be able to use predictive maintenance to ensure that it doesn’t happen again.

Social media monitoring data also plays an important role here: It can search the web for mentions of an outage or hashtags such as #NetflixDown to spot how widespread the problem is.


Though it’s often areas of a business that are programmatic, rather than the entire organization, businesses will start using the programmatic model to run their entire operation. By 2030, Gartner predicts that 35 percent of all businesses will be programmatic businesses.

You should use this approach if:

  • You have a large amount of customer or user data (we’re talking millions of data points) which can be used to improve your product by using machine learning or AI technologies.
  • Product or service delivery can be automated, without the need for human interaction to allow for personalization.

Platform businesses

 Best for: Midsize enterprise 

“In the platform business world, algorithms displace experts and the consumption experience is one of shared access.” — Gartner

Similar to programmatic businesses, platform businesses make use of algorithms but in the form of a shared user experience. This can either be on a maker platform or an exchange platform, but both represent a central hub where users can share, contribute, or consume. The algorithms come into play to help make recommendations and ensure that users are getting a personalized experience.

YouTube is the perfect example of a maker platform: It has content created by users, for users. The platform itself is just a means to an end (engaging with video content) but helps users by making recommendations for things to watch.

Exchange platforms are similar in that they offer a place where users can buy or sell, as well as store and share. In this case, however, it’s more about consumption than creation.

Amazon and eBay are both exchange platforms, which basically serve as the intermediary where people go to buy to and sell products. In both scenarios, user interaction in the form of ratings, reviews, and views play a role in how (and how much) content gets viewed.

Though platform businesses are notoriously difficult to launch, Gartner predicts that by 2030, platform businesses will encompass 23 percent of all business models compared to only 10 percent today.

You should use this approach if:

  • You feel comfortable experimenting with new technologies such as AI and IoT to help improve your offering.
  • You have a new or creative approach to establishing strong and sustainable relationships with customers.

Businesses that won’t need algorithms

Subscription businesses

 Best for: Small business, midsize enterprise 

“In the subscription business world, experts rule with no use of algorithms and the consumption experience is one of shared access.” — Gartner

Subscription businesses rely on access instead of ownership. A central access point can provide thousands (or millions) of users with the same output, but they’re consumed in a different way based on a user’s preferences. The kicker is that experts are needed at the center of the experience to provide value to users.

Spotify, for example, let’s you listen to songs that you don’t actually own and allows you to save them as a collection of music on your personal account. Cloud software is another example, where you can access the tool and all of your data online but don’t actually own a physical copy of the software.

If you consider there is some automation involved when it comes to recommendations, the crux is the experts behind the automation. Sure, Spotify can offer users its Discover Weekly playlist, but without music curators defining genres or classifying artists, the algorithms themselves would be useless. While entertainment is the most common example, other businesses such as food delivery, news, or even makeup, can fit the same subscription model.

Gartner notes that only about 10 percent of businesses are “access” instead of “ownership”, but it predicts that by 2030, that number will be closer to 33 percent.

You should use this approach if:

  • Customers look to your company as experts in the field.
  • There’s a distinct reason why it’s more valuable for your customers to access your product instead of “own” it.

Artisan businesses

 Best for: Small businesses 

“In the artisan business world, experts hold sway with no use of algorithms and the consumption experience is individualized.” — Gartner

Artisan businesses are those that provide a specialized product or service using human interaction to give customers the best experience. It requires a real person to listen to a customer and then tailor the product or service to their needs. People seek out artisan businesses because of a their reputation for personalization, expertise, and authenticity.

This is where small businesses have the biggest chance to thrive against large, “impersonal,” algorithm-driven enterprises. In fact, algorithms aren’t even an option for small businesses. Fewer customers and greater individualized needs means more disparate and less useful data.

A good example would be local businesses such as food trucks and Etsy sellers, or speciality services such as wedding planners and custom furniture designers, where expertise is valued over automation or a quick resolution. Building a reputation with marketing and social media is especially important here.

Currently sitting at 7 percent, Gartner predicts that the number of artisan businesses will go up to 10 percent by 2030.

You should use this approach if:

  • You offer a niche product or service and a strong brand identity that thrives on craftsmanship and authenticity.
  • You don’t have enough data to make valuable use of it in aggregate.

Catering to the customer of the future

Beyond choosing whether you’re going to use a business model based on algorithms or expertise, it’s more about reading the customer of the future and being able to cater to what they’re looking for.

  • If they’re expecting automation and quick resolutions, algorithms will be indispensable.
  • If they’re seeking knowledge and advice, experts will be a better bet.

As with the platform and subscription models, algorithms and expertise can work together to perfect the customer experience.

Consider Spotify as an example. The music streaming service delivers users a weekly playlist, curated to each individual based on an array of algorithms and data points. These include things like the type of music you listen to, the type of music that others with similar listening patterns listen to, and playlists with similar songs on them. The result is a weekly list of 30 songs that—in theory—caters to your tastes. And for the most part, it’s quite successful.


For some, however, it doesn’t always hit the mark.

What this means for the algorithm versus expert debate is that, yes, algorithms can do a fine job of delivering something tailored to users. But at the center of it is the expert that’s able to classify and categorize music accordingly.

Will algorithms or experts benefit your business?

If you’re still having trouble deciding how far you should go toward automation in service delivery, ask yourself these questions:

Chart: Algorithms vs. Experts


The data used in this article can be found here (full text available to Gartner clients).

Share This

Share this post with your friends!