Product-Led
AI & SaaS

Product-led pricing is designed for a buying journey in which the product itself plays the leading role in acquisition, conversion and expansion.

Customers typically have greater ability to discover, evaluate and purchase without a salesperson, so pricing needs to be easy to understand and packaging needs to create intuitive paths from initial adoption to paid and expanded use.

Sales-led pricing is designed for a more human-mediated buying process and can generally accommodate greater complexity, negotiation and customization.

Many businesses ultimately combine the two: product-led acquisition for some customers and sales-assisted or enterprise motions for others.

Start by understanding why the right users are not converting.

Low conversion can result from attracting the wrong users, insufficient value realization, weak differentiation between free and paid plans, poor upgrade triggers, or a freemium model that doesn't fit the product.

A good freemium model holds two things true at once, not as a trade-off. The free tier needs to be genuinely good for some users, because that creates advocacy and a self-propagating pipeline. It also needs to be insufficient for the users you actually want to monetize, because that creates the upgrade path.

Making the free product more generous isn't automatically right, and making it less generous isn't automatically right either. The real question is whether both conditions hold for the specific users on each side of that line.

Behavioral data and customer research show which side is broken: free users not finding enough value to advocate, or target users not hitting a wall that justifies upgrading.

It may well be, particularly if customers receive very different levels of value while paying essentially the same amount.

A flat model can limit monetization when there is no mechanism for customers' spend to increase as their needs, usage or value increase. It can also make it difficult to serve customers with materially different willingness to pay.

The answer is not automatically usage-based pricing. Depending on the product, the better solution might involve packages, feature differentiation, a scaling price metric, add-ons, usage components or a combination of these.

The right design creates meaningful differentiation while preserving the simplicity that makes a product-led model work.

These are different commercial mechanisms and solve different problems.

Freemium can work well when ongoing free use helps drive adoption and a meaningful subset of customers develops needs that justify upgrading. A free trial is often better when customers can experience sufficient value within a relatively short period. Usage-based pricing is appropriate when a measurable form of consumption has a strong relationship with customer value and creates a sensible way for spend to scale.

They can also be combined.

The decision should reflect the product's value journey, customer behavior, economics and growth strategy and should be validated through the most appropriate combination of customer research, behavioral data and experimentation.

Seat-based pricing is usually the wrong model for AI features, not just a weaker one.

Seats fail on two dimensions at once. They don't track value, and they don't grow with it. A user generating ten times more output pays the same as one who barely uses the feature. As AI increases productivity, the number of human users required can shrink even as value delivered grows.

The task is to find the metric that actually tracks value: actions, workflows, outcomes, consumption or another measurable unit. Evaluate for value alignment first, then customer acceptance and operational feasibility.

Cost to serve is a real constraint, but it belongs at the portfolio level, not as a per-transaction pricing input. Pricing to cost recovery is a common mistake, and it produces a metric disconnected from the value customers receive.

If the model changes, migration matters as much as design. Existing customers need a credible path from seat economics to the new model without bill shock.

We begin by understanding the objectives, product, customer journey and economics of the current model.

Product and behavioral data can be particularly valuable in PLG because it shows how customers actually use the product, where conversion occurs, how usage develops and where different customer groups behave differently.

We combine that internal evidence with appropriate customer research and competitive analysis, then evaluate potential packaging, price metrics, free-versus-paid boundaries and price levels.

Where live experimentation is practical, it can add another layer of validation. The specific research methods depend on the decision being made rather than following one fixed PLG methodology.

Better PLG monetization can increase revenue through higher conversion, stronger expansion, better value capture from high-value users or an improved revenue model.

For example, one customer service platform projected more than $10 million in incremental revenue after redesigning its monetization model around a usage metric, while customer research supported materially higher price levels for new AI capabilities.

The precise opportunity varies considerably by product and starting point. The aim is better monetization that does not undermine the adoption and low-friction customer experience that make product-led growth attractive in the first place.

Start by defining what “off” means.

The problem could be weak free-to-paid conversion, poor expansion, excessive churn, insufficient monetization of heavy users, an inappropriate price metric, ineffective packages or simply price levels that have fallen behind the value of the product.

Usage, conversion, churn and customer data can help identify where the economics are breaking down. Internal stakeholder perspectives and external customer research can then explain why.

Only once the problem is understood should the company decide whether the answer lies in price levels, packaging, metrics, free-versus-paid design or a broader change to the monetization model.