Most pricing in software is based on gut feelings rather than calculated data. This has created a huge gap in long-term growth revenue.
This assessment will destroy the three-tier default, breaking down the data behind a successful monetization model and revealing how leading companies use real billing models as opposed to generic pricing tips.
What You Should Know Now About SaaS Software Pricing
When searching for ways to structure saas software pricing, you will encounter surface tips on how to create a design for pricing on the market. Examples of common advice include the creation of three tiers of pricing, the inclusion of an annual billing option, and the emphasis on a mid-tier option.
What all of these pieces of advice have in common is that they fail to address the need for proper monetization.
Why Design Does Not Mean Pricing Strength
Creating design boxes on a website will not establish pricing strength. Pricing strength comes from aligning the product's usefulness to the intended result that the user wants to achieve by purchasing your product.
Why The Basic Model Fails
The advice is often to use a "Good, Better, Best" pricing model, but simply separating product features into three different price tiers does not address the issue of what the customer is willing to pay for your product.
The metric used to calculate the price you charge your customer (e.g., seats, API usage, or projects) must be correct; if it is wrong then your future growth revenue will be limited.
The best companies in B2B enterprise will have a net retention rate (NRR) greater than 120% while the best companies in PLG will have an NRR above 110%; an extreme example would be Snowflake reaching an NRR of 158% at its IPO. The way a company obtains these numbers is through using proper math to calculate what those cohorts would be and adjusting the pricing in a structured manner.
The 90-Day Test Cycle
An elite team no longer considers pricing to be a "huge" event that happens once every 12 months. Allowing an error in pricing to go unfixed for 12 months costs the company money and is risky. The current trend in revenue operations is for companies to do a 90-day price iteration cycle.
For each 90-day price iteration cycle, teams should gather real data on how customers use your product, conduct a Van Westendorp price sensitivity survey, run a fake door test, and conduct A/B price tests and other price tests. These teams shouldn't just measure total number of conversions; they should also measure revenue per visitor (RPV).
For instance, increasing your price by 30% may cause your overall number of users to decrease, but experimentation proves that if done correctly you can increase your overall conversion by 18%. Thus, those who can respond quickly using data can be successful in this new pricing environment.
Market Benchmarks for SaaS Software Pricing
To determine a price, you must understand the benchmark of your category when setting saas software pricing. It doesn't do you any good to rely on the overall market average; you'll just get misled.
Real Prices By Category
Each category has its own specific market price reality. The overall entry price (median) across the software market as a whole is $12-$15 per user per month; in the top quartile, the median entry price is $49; and in the bottom quartile, the entry price is $15.
The average entry prices of these broad categories mask the true differences between actual product categories. A CRM tool has a median entry price of roughly $22.60/user, project management software is about $11.21/user, and time-tracking tools are lower than both at around $8.71.
If you are launching a developer tool and pricing it the same as a time tracker, you are going to lose all your profits immediately.
Free Trials and Entry Levels
In addition, it is common practice to provide some type of free version or entry-level tier. Approximately 60% of tools tracked use some type of free entry point to help with early adoption.
Additionally, it has been shown that reducing the length of a free trial from seven days to three has also increased conversion rates for paid subscriptions between 22%.
How to Build Price Tiers
When constructing user tiers, it is important that the price differential between user tiers match the level of value being provided.
Based on the available data, the norm for a healthy tier structure is to have a price differential of at least 2.5 to 3.5 times the price of the basic user tier to the admin or advanced user tier.
Using Add-Ons for Growth
There is tremendous growth potential in add-ons. When a high-value feature is presented as a standalone add-on, rather than as an inherent part of a core tier, there is a documented benefit of 18% to 25% additional revenue on enterprise-level transactions generated from that add-on.
Add-ons typically range from $50-$200 per user per month. The fact that companies can charge additional revenue for add-ons allows those companies to capture revenue from power users without scaring the basic user away.
Usage Overages and SaaS Software Pricing for AI
The single largest driver of revenue growth from billing models based on usage is their ability to drive up ARR. This is evidenced by the fact that companies that transition to a usage-based billing model and away from the traditional per-seat model report 25% increases in ARR and 50% reductions in churn, annually.
The most effective usage-based billing architectures provide committed limits combined with variable overage charges. Top-performing companies achieve ratios of 60% committed revenue and 40% variable revenue on a long term basis. When users exceed their committed limits, the overage charge generally has a premium of 15% to 25% above the base rate.
Using different pricing strategies, including artificial intelligence features, creates completely different financial models. Companies charge a large percentage of a premium (between 60-85%) for AI tools due to the amount of server power required to support AI.
The current adoption rate of the premium AI tools (as a percentage of active users) is between 45-55%. However, to avoid losing money on heavy users, the compute-adjusted margins need to stay between 65-72%.
Billing Systems Must Support Your Strategy
If your infrastructure is weak, your company’s strategy will fail, as you can only charge based on how your billing system can measure and control the cost.
Tracking User Limits and Meters
An entitlement is a rule that indicates what your software permits a user to do. If your user purchases 100 API calls, they will have received an entitlement to use those API calls, and the entitlement system will count those calls and cut off access at 101 calls.
Many teams develop a very creative hybrid pricing strategy only to find that their custom-built code cannot significantly keep track of the usage that has occurred.
There are several tools available to manage these runtime rules, like Schematic or Autumn, which allow product teams to control access to features without having to write custom billing logic for every change.
Top Billing Providers for Software
The market depends on a specific set of infrastructure providers to transform pricing ideas into cashflow for the businesses. If you select the wrong tool, you will create significant operational inefficiencies.
Stripe Billing: Stripe billing is the default engine for pay-as-you-go and standard tiered pricing models. Stripe has gained a wide usage due to its developer-friendly tools, but due to the complexity of contracts, many teams discover that they quickly outgrow Stripe.
Chargebee & Maxio: Chargebee and Maxio target mid-level and enterprise-level companies that require detailed management of recurring subscriptions and enforce strict revenue recognition policies.
Metronome and Orb: Usage recording and complex application program interfaces (APIs) are often relied upon as sources of revenue for products; therefore, these tools do not track contract-based metrics but instead enable sales teams to create custom contract terms related to large customers, without disrupting the main billing system of the product.
Paddle: For global companies selling on a global basis, Paddle assists with being your merchant of record by automatically handling global tax compliance and localised pricing.
Evendeals and Lago: With Evendeals, there are geographic pricing laws in place to prevent users from using virtual private networks to purchase poorly priced plans. Lago offers an open-source option for teams who require complete control over how they bill for service usage.
Defending Your Prices Against Smart Buyers
Buyers no longer have to speculate whether they are obtaining a good offering; they know precisely what other customers of a comparable kind have been paying and use this knowledge against you.
The Threat of Software Spend Trackers
Enterprise buyers have now access to software spend benchmarking platforms, such as Varisource, Vertice, Zylo, CloudEagle, Tropic, Spendflo, and Vendr which gather aggregate billions of transactions worth of data from tens of thousands of vendors.
When the buyer is preparing to negotiate for renewal, they have complete knowledge of your median discount. They know exactly what you charged a similar customer last week.
If your sales team tries to defend a position contrary to the data of Vendr or Tropic, the buyer will call their bluff; you cannot afford to continue to hide behind uninformative "Contact Us" buttons.
Stopping Sales Discounts
To protect the profit margin against armed buyers, the revenue leadership of a company must enforce discount governance. The manner in which the median market deal is discounted continues to destroy value within the industry; bottom quartile performers will discount an average of 64% off the total value of their deals.
This has led to significant devaluation of both enterprise contracts and associated revenues. On average, enterprise contracts are larger than other types of contracts.
A company with less than 100 user licenses will typically contract for an average of $47,000 per year, while a company with more than 1,000 user licenses can expect a deal to be worth approximately $890,000. For example, when enterprise customers enter into three-year service agreements, they are often entitled to deep discounts; in most cases, these discounts equate to approximately 23% off of the total contract value.
To address the continual problem of losing revenue by offering discounts as a means of closing deals quickly, many leading organizations implement approval matrices that prohibit their sales representatives from giving away some or all of a deal's revenue without prior approval from a senior finance executive should the discount exceed the set limit.
Final Thoughts on SaaS Software Pricing
The final conclusion regarding monetization architecture is that saas software pricing is a function of the operations of a company, not a marketing function. Therefore, a company will not achieve a sustainable level of growth by simply placing three arbitrary tiers of pricing on their homepage with the top-tier pricing concealed.
In order for an organization to be successful, the exact value of the pricing must be matched with the desired outcome/objective of the customer/user. In addition, a firm must implement complex billing systems (such as Metronome or Chargebee) to accurately track usage.
Pricing must also be viewed as an ongoing cycle of testing over a 90-day period, in which a firm's pricing margins will be continually defended against customers who have prior knowledge of the maximum amount of discounts they may receive. Companies that successfully master this analytical process will take the market from those organizations that simply use gimmicks and continue to lose their revenue.
How to Stop Losing Software Revenue
System Limits That Stop Hybrid Billing
Most legacy billing systems were designed to bill credit cards once a month for a fixed amount. When a business implements a hybrid model of a base and variable fee based upon a customer's level of usage, the legacy billing systems cannot adequately track usage daily.
If the usage cannot be measured, the business cannot bill accordingly. This leaves the business locked into flat rates and allows high-power users to utilize an unlimited amount of server resources and pay for none of it.
Signs Your Prices Do Not Match Customer Value
The most obvious indicator of a misaligned value metric is high initial conversion rates followed by a flat-line growth expansion and a high churn rate. If customers can easily buy into the basic plan but do not upgrade, this indicates that the customer is receiving all the value they need from the base plan and is being held back from crossing through the paywall.
Companies that are high performers have a strong correlation (approximately 80%) between the pricing metric and the measurable outcome achieved by the customer. If the metric being measured is "number of logins," customers will gamify the system to minimize their logins, which will ultimately destroy future growth.
Why Testing Prices Often is Better Than Yearly Changes
A single large price change usually causes shock to the customer base, with sudden spikes in cancellations and negative feedback. In addition, since a price change takes several months to put into place, the data used to make the price-change decision is most likely outdated by the time the new price goes live.
Continuous test cycles allow businesses to take advantage of new cohorts and validate the willingness-to-pay using live traffic with real dollars. By continuously testing smaller price adjustments, the business reduces the overall risk and increases total revenues at a faster rate.
How to Price New AI Tools Without Losing Users
Since AI utilities create significant computing costs, businesses will be unable to provide AI capabilities free within an existing price tier. The best practice is to price AI capabilities as a separate module and receive a premium (60% to 85%) price, as opposed to increasing the base price for all.
The data indicates that by keeping the base price constant and charging a premium specifically for AI usage, the vendor protects the core user base. Therefore, the heavy computing costs for the AI tools will only be incurred by the 45% to 55% of users who actually utilize and find value in the AI tools.
