How We Decide What to Charge For

The Line Between Free and Paid

Every time you ship something new, someone on the team asks: "Should this be free or paid?" The question sounds simple. It never is.

Get it wrong in one direction and you train users to expect everything for free. Get it wrong in the other and you gate the thing that would have convinced them to stay. Most founders treat this as a gut call. It doesn't have to be.

Here's the rubric we use. It's not perfect, but it's repeatable — and repeatable beats perfect when you're making this call every few weeks.

Two Questions Before Anything Else

Before you debate pricing tiers, answer these:

1. Does this feature help someone decide if the product is for them? If yes, it probably needs to be free. Gating evaluation features is like locking the front door of a store and asking people to pay before they browse.

2. Does this feature deliver value only after someone has already committed? If yes, it can carry a price. People already inside your product, getting results, will pay for more results. They won't pay for the privilege of figuring out whether you're useful.

Most features fall cleanly into one bucket or the other. The trouble is the ones that don't.

The Trust Layer: What Should Always Be Free

Some things exist to build trust. They reduce risk for the person evaluating you. They answer "Can I rely on this?" before money changes hands.

Trust-layer features:

• Visibility into how your product works (status pages, documentation, transparent logs)
• Basic functionality that lets someone experience the core loop
• Anything that helps a user leave if they want to (data export, account deletion)

That last one deserves emphasis. Making it easy to leave is not a revenue problem. It's a trust accelerator. The easier it is to walk away, the less risky it feels to walk in.

If you gate trust-layer features, you don't make more money. You make fewer customers.

The Value Layer: What Earns a Price Tag

Once someone trusts the product enough to commit, certain features amplify the value they're already getting. These are fair game for pricing.

Good candidates for paid features:

• Scale — more volume, more users, more storage
• Collaboration — letting teams work together inside the product
• Depth — advanced configuration, richer data, finer control
• Support — faster response, dedicated help, onboarding assistance

Notice the pattern. Paid features multiply something the customer already has. They don't create the initial value. They extend it.

A useful test: if a brand-new user would get nothing from this feature because they haven't built anything yet, it belongs in the paid tier. If a brand-new user needs it to understand whether the product is worth their time, it doesn't.

The Gray Zone

Some features sit on the border. They help with evaluation and deliver ongoing value. A reporting dashboard, for example — it helps prospects see what the product can track, and it helps paying customers make decisions every week.

For these, we use a simple split: show the shape for free, charge for the depth.

Let someone see that reporting exists. Let them poke at a summary view. Reserve the detailed breakdowns, exports, and historical trends for paid plans. They get enough to evaluate. They pay for enough to operate.

This isn't a trick. It's honest. You're saying: "Here's what this does. If it matters to your business, here's what it costs." Most people respect that.

When You're Wrong

You will misprice features. Everyone does. The signal to watch for: if free users aren't converting and they tell you the product is great, you've probably given away too much of the value layer. If prospects bounce before engaging, you've probably gated too much of the trust layer.

Neither mistake is fatal if you catch it early. The expensive mistake is having no framework at all and making every call from scratch — because then you can't diagnose patterns, just individual decisions.

The Rubric on One Card

When you ship something new:

1. Is it a trust feature or a value feature? Place it accordingly.
2. If it's both, show the shape for free and charge for the depth.
3. Watch conversion and engagement for 30 days.
4. If the numbers disagree with your placement, move the feature. Don't defend the original call.

No pricing consultant needed. No A/B test requiring traffic you don't have. Two buckets, one rule for the gray zone, and a commitment to adjust when reality talks back.

The Deeper Principle

Pricing decisions are trust decisions. Every gate you put up says something about what you value. Gate the wrong thing and the message is: "We don't trust you enough to let you evaluate us." Gate the right thing and the message is: "We've already proven ourselves. This is where the relationship grows."

Draw the line where trust ends and value multiplies. Hold it until the data says otherwise.