How enterprise search giant Algolia embraced usage-based SaaS pricing

In the world of SaaS, there is an art and science to hitting the right pricing. While it can be tempting to adopt the popular “basic>>premium>>enterprise” subscription model that ushers in more features as the monthly pricing increases, there is a growing fanbase around usage-based pricing — that is, charging companies for what they’re actually using, rather than a flat monthly fee.

One company that has embraced usage-based pricing is Algolia, the billion-dollar search-as-a-service company that powers search functionality for companies including Atlassian, Slack, Staples, and Under Armour.

Algolia appointed Bernadette Nixon as its new CEO in May 2020, and one of her first tasks was to oversee the company’s transition from infrastructure-based pricing to pay-as-you-go, which went into action a couple of months later. This, effectively, meant that Algolia went all-in on usage-based pricing, though existing customers already on a flat fee were allowed to remain there.

Some two years on, VentureBeat checked in with Nixon to get the lowdown on the thinking behind Algolia’s pricing switch, and some of the lessons they can pass on.

Model example

Infrastructure-based pricing is when a customer essentially pays for all the behind-the-scenes technology required to power their app, while usage-based pricing is concerned solely with consumption. Much like how we pay for electricity, gas, and other domestic utilities based on how much we use, it can make more sense for companies to pay for what they’re actually using, rather than a set monthly figure that reflects the maximum amount of infrastructure a company might need in a given month or year.

“If your website has low traffic, that probably means the usage of our product is low too, so the cost [with usage-based pricing] reflects that,” Nixon told VentureBeat. “If traffic is high, then the cost goes up.”

It is worth noting that infrastructure-based pricing can sometimes align with its usage-based counterpart, for example when a company is selling the actual infrastructure, such as Amazon’s Elastic Compute Cloud (EC2). But for the most part, they are distinct models, each with their own unique set of challenges.

Logistically, it’s more straight forward to operate a subscription or flat-fee model — it’s easier to sell, market, manage, and understand. And when a company charges based on consumption, they must align all their billing operations, usage data, and commercial contract terms — and then be able to deliver real-time usage data to customers in a user-friendly format, showing how their consumption translates into spend.

“The most difficult part of a usage-based pricing model is having a customer really, truly ‘get it’ and understand how you are accurately accounting for your service,” Nixon explained. “From a product perspective, you want your customers to understand your pricing model, the same way they get their electricity bill. At the end of the day, the want is the same — you want visitors at your house, you don’t want to shut off the lights. And you want people to use your product and get the most out of it, but it’s a challenge to get that level of customer understanding for usage-based pricing.”

While Algolia designed its own usage-based pricing infrastructure in-house, not all companies have the resources to do that, which is why we’re starting to see a wealth of tools on the market that help companies manage all of this for them — the “data infrastructure” to handle complex pricing configurations, capture granular usage and cost data through system integrations, and then calculate bills in real time.

Pricing problems

In February alone, two separate platforms emerged from stealth to deliver the billing and data infrastructure required for companies to deploy and iterate usage-based pricing models — Metronome went to market with $35 million in funding, while M3ter launched with $17.5 million in its pocket.

One of M3ter’s launch partners was Paddle, a payments infrastructure company for SaaS companies. Paddle recently released a new report called Outliers: State of SaaS Growth, which was based on a survey of 180 SaaS companies and proprietary data from thousands of Paddle’s own customers. The report found that SaaS companies saw revenue grow by an average of 32% in 2021 — a 46 percentage-point drop on the previous year’s growth, suggesting a return to normality following a pandemic-driven peak in 2020.

However, Paddle’s data also highlighted so-called “outliers,” which are specific types of companies that performed particularly well during the broader slowdown. These include companies that have embraced new growth models, such as dynamic pricing — 40% of companies that change their pricing regularly reported a 25% higher increase in annual recurring revenue compared to those that didn’t. But what is perhaps most telling from the report, is that 20% of companies haven’t changed their pricing in the past five years.

And this highlights one of the reasons why Algolia shape-shifted its pricing back in 2020 — it listened, and adapted accordingly.

“Organizations pursuing an accelerated pace of innovation need flexibility — they must be able to adjust quickly to stay competitive, and eliminate technical debt,” Nixon said. “Businesses need to listen to their customers’ needs and preferences in order to make the smartest decisions regarding pricing models.”

Those pricing models don’t necessarily have to be usage-based, but they do have to be tailored to the business and industry in which they operate in, and what their customers are looking for. There isn’t a one-size-fits-all pricing model — it ultimately depends on the nature of a company’s business.

Algolia is far from the first company to adopt usage-based pricing — the likes of AWS, Twilio, and Snowflake are all firmly on the metered billing bandwagon. But to make usage-based pricing work, it helps if a company’s product or service can be broken down into small chunks or “units” which can determine what “usage” actually means — that could be a set per-message rate as Twilio might charge its customers for handling their communications infrastructure. For Algolia, it has defined a chargeable unit as 1,000 monthly search requests plus 1,000 records within Algolia’s indexes — and if either of those values exceeds 1,000, then a new unit begins.

But for a more nuanced SaaS product, such as a customer relationship management (CRM) tool that ships with many different features and functions, it could be more difficult to figure out a “unit” to charge against. In such cases, a monthly subscription may make more sense.

On top of all that, usage-based pricing can be confusing for would-be customers to work out, if they don’t know how many “API calls” they’ll use, for example. And for the SaaS vendor, their revenues are often less predictable under a consumption-based model.

Challenges

So there are many factors to consider, and while usage-based pricing certainly has its pluses, it might not be the answer for every type of SaaS. But when the decision has been made to pursue a usage-based consumption model, companies have to put the infrastructure in place to support it, and then explain to their customers how it all works. And that, according to Nixon, is one of the biggest challenges.

“We grow with our customers, but it requires a lot of upfront architecture to create a successful and sustainable usage-based pricing model,” Nixon explained. “It looks easy on paper — however, it’s a challenge to bring the automation and transparency to a customer, so they can easily understand. If you are looking to adopt a usage-based pricing model, you have to build it from the ground up. It needs to be developer-friendly — in our case — and easy to deploy, so you can realize the ROI (return on investment) quickly. There is a lot of behind-the-scenes work that goes into this, and it takes a lot of engineering and investment to do it the right way.”

Beyond pleasing customers by allowing them to pay for the features that they actually need and use, usage-based pricing helps companies see the direct value in the products that they’re paying for, given that there is a direct correlation between the features that they’ve deployed, and the price that they are charged. And that continues as the company launches new products — they can start to see the value on their investment immediately from a “revenue and quality of engagement perspective,” according to Nixon.

“If you are not using a usage-based pricing model and are relying on a flat fee, it’s a huge challenge to ascertain what is really driving customers,” Nixon said. “The knowledge you get from a usage-based pricing model drives your customer engagement, value and retention. When vendors offer a usage-based pricing model, it allows for a better customer experience and a greater, more immediate ROI.”