The above models are designed to capture most scenarios; overall, we found the results were mixed. We saw 17 total unique combinations of pricing and packaging models. Many companies use different strategies for different products and also use multiple pricing strategies and/or multiple packaging strategies. The most common combination was a per-seat pricing model with usage-based tiers, followed by a flat-fee pricing model with usage-based tiers. In the above classification, both of these strategies would fall under the Seat-based or Flat-fee Pricing with UBT category, which is the most common construct used in PLG pricing today.

The best PLG pricing strategy for most companies

The pricing and packaging models we defined in the previous section provide a solid and representative framework for PLG-friendly pricing and packaging. But as with many other aspects of an “as a Service” pricing strategy, the devil is in the details — and in many cases, the best strategy depends on the situation. Below, we provide some overarching guidance based on our study and data. However, it is important to consider how to adapt this guidance to your category, your business and, most importantly, your customers. Some thoughts on that follow.

As we noted in the previous section, “usage” in the context of a PLG commercial strategy impacts two areas: pricing and packaging. Determining the right PLG strategy means dissecting each of these elements in detail and then bringing them together into a holistic strategy.

We’ll start with what we think is easier: packaging. We believe that for nearly all SaaS companies, a UBT packaging strategy works best. However, there are some exceptions. If you sell infrastructure services like Amazon Web Services, you may not be introducing any type of tiering for your product. But for most companies across SaaS categories, UBT will work best.

What does a best-practice UBT packaging model look like? According to our data, for most companies it should include the following key elements:

• Between three and five total product editions, usually including a free plan, two to three self-serve plans, and an enterprise plan with custom pricing
• Editions are defined based on a combination of features (e.g., “SSO” or “admin portal”) and usage factors (e.g., “up to X widgets”).
• Three to five usage factors are employed to define editions; usage factors are consistent from plan to plan, and plans are clearly presented to show how upgrades enable additional usage (i.e., “Plan 1 – 100 widgets” versus “Plan 2 – 1,000 widgets”).
• The free plan and/or first paid plan may have more factors gating usage (approximately 10 to 15 on average) than other plans, which is done to place more fences on usage and promote faster upsell to paid plans.
• One of the factors is clearly signaled as the primary usage factor or “value metric” and presented as such on the page, often with a volume pricing drop-down menu or sliding scale calculator.
• Usage factors are usually “softly managed limits,” meaning vendors will monitor usage for repeated overages and then engage with customers to manage the upgrade; overage is only chargeable if the company employs a UBP strategy.

Pricing models for PLG are where things become a little murkier.

Our study suggested that just under 60% of leading PLG companies are using seat-based pricing models. This is a somewhat startling statistic, given that monetizing based on usage is a core PLG tenet (OpenView doesn’t say “Monetize based on users”) and that, generally speaking, per-seat pricing is a bit of a dirty word in SaaS pricing. In the LinkedIn chatter surrounding our PLG Pricing Analysis, Scott Logan, SVP of Marketing at Kronologic.ai, summarized it well by outlining that per-user pricing is “paying for access to value” whereas usage-based models are “paying for value.”

On the one hand, the best of the best PLG companies are in many cases using seat-based models, while on the other hand, industry wisdom is saying, “Don’t use seats.” Where does that leave you? Again, it comes down to the specifics of your category and the needs of your customers. But there are general frameworks that can be helpful to guide you.

Here’s a rule of thumb we like to use on seat models: If seats are truly your value metric, meaning each additional seat provides additive value benefits, then price per seat is right for you. If not, then consider another value metric and pricing metric. This post from Baremetrics summarizes this concept well. Let’s illustrate with a couple of examples:

• Lattice provides SaaS tools for employee engagement and performance management. This product is valuable only if everyone on your team or in your department at a minimum — or better, in your entire company — is using the platform. For Lattice, it makes complete sense to charge based on seats; otherwise, the value of the platform is being restricted.
• Databricks provides a big data warehousing and processing platform that runs on top of Amazon Web Services, Microsoft Azure and Google Cloud Platform environments. Databricks’ value is tied to how much data it manages independent of how many database administrators or other IT users are managing the platform. As such, users aren’t even mentioned on Databricks’ pricing page. Databricks charges based on a unit of processing capacity that it calls a Databricks Unit, or DBU. As usage of Databricks grows, DBUs increase and pricing increases.

There is one pricing model, however, that can work in nearly all scenarios: the flat-fee subscription model. Others may call this “tiered pricing” or use similar wording, but we try to differentiate tiering as a way of packaging services versus how services are priced. A flat-fee model is a pricing model in which the customer is charged a fee for a defined subscription on a monthly, annual or other term basis for a given tier or product.

Flat-fee models work well with usage-based tiering strategies because they align pricing directly to usage metrics but also provide safety to the vendor in terms of repeatable, predictable subscription revenues. There is risk in ensuring customers are consuming an appropriate plan for their needs, but otherwise these models scale well with usage as customers either will naturally need to upgrade for usage and/or pay usage-based overage fees for additional usage beyond their allotted amount, depending on how the pricing model is set up.

Our favorite example of flat-fee subscription pricing with UBT is Zapier. Zapier is often cited as a key success story in usage-based pricing, but the reality is that Zapier is a success story in UBT. Zapier offers five plans ranging from a free forever plan to an enterprise edition. Plans scale primarily based on features as well as a single value matric — tasks per month, where a “task” is defined as an automation action that is completed by Zapier (i.e., a contact that is created from a form). Customers can choose a set volume of tasks per month for their chosen edition level, ranging from 100 tasks per month for the free plans all the way up to 2 million tasks per month at the top end of the company plan. Here’s the beauty of this model for Zapier: While the pricing and usage scales by plan and by volume of tasks, Zapier doesn’t price “per task.” The company prices for a set volume of tasks that the customer is allotted in a monthly period, and regardless of whether that exact number of tasks is used or not, the customer pays the same subscription price. This creates predictable subscription revenue and puts the onus on Zapier’s customer success teams to ensure customers are using Zapier to the full extent that they paid for.

Flat-fee pricing can also make sense for products that scale based on users. Let’s take Snyk, for example, which is the developer of a security platform and another darling of the PLG world. Snyk’s free plan provides for unlimited developers but limits the number of tests per month that can be supported by the platform. This encourages companies to get all of their developers using Snyk and, ideally, experiencing the value of the product. Once companies switch to a paid plan, they are entitled to unlimited tests but for a limited number of developers. This effectively shifts the monetization metric from tests to developers (users). Snyk addresses this by using a flat-fee pricing strategy versus a per-seat pricing strategy. Pricing scales based on seats, but seats are sold in bundles starting at five developers and scaling up to 75 developers for self-service plans. The cost to the customer is based on the total number of developers for the plan they’ve chosen, regardless of whether that many developers are active or not during the monthly metering or annual metering period. This model works well if your product benefits from usage across a team or department.

Bringing it all together

If you aspire to employ a PLG pricing strategy, you are in good company if you design a UBT packaging strategy around three to five tiers and price your product based on seats or a flat fee. Flat-fee models work for most companies, even those that aim to position seats as a usage or value metric.

So that’s an effective starting point. But it’s not the be-all and end-all. Frameworks and averages can’t replace a deep understanding of your category and your ideal customers.

A tool like XaaS Pricing, and analyses like the one in this post and the PLG Pricing Analysis blog, can be conducted for your specific market category. This can help provide an overall understanding of norms in your sector and used to rough out an initial packaging and pricing model strategy that considers your market while being aligned to the unique value positioning and differentiation you seek to carve out for your product and each edition of that product.

The next step is customer research, which should be used to test packaging and pricing hypotheses, understand value, and finalize your PLG packaging and pricing strategy. This can include, depending on your time and budget, interviews and/or surveys with your existing customers and ideal customers. The research should be segmented to capture representative inputs from ideal customers of each plan, as well as input from the different geographic markets in which you sell, if applicable. The research should use an open-ended approach and aim to understand concepts such as willingness to pay, preference for different types of pricing models, and relative preference of features. This information will not give you the answers, but it will help you to test your initial hypothesis and capture data to make decisions on key elements such as usage factors, value metrics, pricing metrics, packaging and price levels.

In an upcoming post we’ll provide a template for conducting the type of research provided in our PLG Pricing Analysis at the category level. Become an Insider to receive that template in your inbox when it publishes, and don’t hesitate to contact us with any PLG pricing questions!