These days, consumers can sign up to almost anything on a rolling monthly basis — from music and streaming services, through to tailored vitamin subscriptions and pet food. It’s attractive because it’s easy — payments go out frictionlessly, with services continuing consistently, uninterrupted whilst allowing businesses to reap the benefits of passive loyalty.
The same is true in the B2B software world — or at least it has been. Over the last decade, subscription-based SaaS pricing has become the standard. Enterprises have moved away from expensive, monolithic vendors, instead choosing to build bespoke tech stacks that plug into a range of best-in-class solutions. Today, the average enterprise buys from over 900 SaaS vendors.
As integrations continue to stack up, some are thinking twice about pricing. Lately, there’s been a lot of talk suggesting that the era of the SaaS subscription model is over. Instead, pay-as-you-go and usage-based pricing are rapidly gaining favour. But is it really the silver bullet some in the SaaS world seem to think it is?
Why SaaS pricing works
Subscription-based SaaS pricing has worked so well to date because it offers stability and visibility. For obvious reasons, investors and the SaaS businesses they grow love it. Subscription models provide a level of predictability around cashflow, as well as a framework investors can use to gauge the quality and scalability of the customer base.
And some customers love it, too. Subscription models add value in the long term; being locked into contracts means customers can take advantage of continuous updates, rather than having to wait for annual ‘big bang’ releases. Plus, the cost control subscription models offer is attractive to the likes of banks, or others with rigorous compliance requirements.
For others, though, the inflexibility and upfront friction are off-putting, driving the evolution of Saas pricing. Instead, usage-based pricing models are becoming increasingly popular because they create value through the product itself.
What’s driving the transition to PAYG
SaaS became popular in the first place because it removed so many barriers to entry. Making resources available through the cloud made it easy for customers to adopt new products and services, bypassing arduous (and expensive) implementations.
Usage-based pricing models take that even further. Enterprises have hundreds of subscriptions already, and adding new ones can be a hassle: Subscriptions need to be justified to procurement and finance, which can be tricky if the RoI of a product only manifests in production and compounds with use. Instead, pay-as-you-go lets individuals get started easily, removing a lot of that anxiety. They can test the product to verify whether it’s going to bring value before they commit.
There’s a sense of fairness that comes with that. In some markets, like cloud infrastructure, DevOps and other developer tools, there’s a lot of variability in the underlying workflows. It can mean that the resource and capacity customers will need changes month to month. With pay-as-you-go pricing, they can scale their pricing up or down and be charged accordingly, rather than having to buy add-on packages at extra cost, or paying upfront for huge capacity they actually don’t need.
There’s value for vendors in customers’ sense of alignment with the product. The chances are that if they can use it freely and see the value it adds, they’re likely to scale it up and drive wider adoption inside their own business. A lot of vendors are taking advantage of this ‘virality’, favouring land and expand go-to-market strategies. And it’s working. Companies like Snowflake have seen massive success with the model, and practically all infrastructure resources are priced on consumption. Though the returns might be less predictable, they’re potentially more scalable; subscription models can leave a lot of money on the table.
Why PAYG isn’t a silver bullet
For SaaS businesses looking to capture more of the value they create, experimentation with payment models makes sense. But usage-based models aren’t a silver bullet, and won’t work for every business in every industry. And there are a few reasons for that.
For one thing, virality isn’t guaranteed. In actual fact, it’s a bit of a myth: Slack is really the only company that’s actually achieved it, possibly because of the network effects of having everyone across a business using it. Landing small means there is a big opportunity to expand, but it’s hard to get visibility on, as well as measure or control.
Usage models can create tension at the other end of the spectrum, too. Beyond a certain point, pay-as-you-go can become prohibitively expensive, especially in industries with extremely high workloads. Products which win loyalty by enabling cheap, easy access can actually end up so costly that they’re all-but-unusable. We’re already seeing this happen: data warehousing platforms are being challenged by new players who innovate on pricing, because it’s becoming prohibitively expensive for customers to run on consumption-based models.
Eventually, customers are going to expect some sort of ‘fair usage’ policy, or their loyalty will start to wane. Businesses that charge for usage aren’t necessarily invested in innovation, and customers know that. Product improvements that might reduce workloads aren’t in their interests: they make money from customers doing the same thing more often, not more intelligently.
What the future of SaaS pricing will look like
That we’re having these discussions today shows how much the market is advancing. There’s been an explosion of software and tools over the last five years — a byproduct of the move to cloud and a default AWS economy. The evolution of pricing models is simply a reflection of the evolution of the market itself.
That evolution is going to continue and as such, there’s no one way that new entrants to the market should look to structure their pricing. Different SaaS companies require different models and approaches. And there are a few ways to go with it: one is to get as much as you can up-front, with all employees immediately using the product. The other is to assume your product is super-viral, and rely on product-led growth. Somewhere in-between is land and expand, with a minimum number of active users.
To figure out the best approach, businesses need to think carefully about what value they add. Usage models give more room to grow, but SaaS businesses need to define clearly what that growth means — what the driver is that makes them sticky without becoming too expensive, too quickly. Paying per-use doesn’t work in every instance. It’s good for payments and fintech because they’re inherently transactional. It wouldn’t work so well for an email service, for instance: customers are unlikely to want to be charged for every additional email they send, or to see value from that.
Pricing, like product, is about market fit. It will (and should) change according to market, industry, and type of customer. It should change again once SaaS businesses reach scale. We’re already seeing a lot of experimentation. Many are finding they’ve started too cheap, and left too much money on the table with subscription models. Now, they’re switching to usage-based pricing to capture more of that value. Others that started with pay-as-you-go pricing are switching to subscription, to prevent themselves from being commoditised, and to bake visibility, stability and predictability into their growth.
There is no ‘right way’ to do it. There’ll be more experimentation over time, and the most effective models will build and iterate on what’s gone before. With the ecosystem in such a state of flux, it’s the perfect time for SaaS businesses to analyse different models and synthesise a new approach.