Most product leaders are measured on what they launch. The harder, higher-signal skill is knowing what to kill.

I have killed three product lines at a single company. A $40M program I was hired to ship. A VR training simulation with a dedicated international development team. An $8M partnership that was underdelivering. Each time, the capital freed up funded the products that actually shipped and grew revenue. Every kill decision followed the same evaluation framework.

Why is killing products so hard?

Because sunk cost is emotional, not rational. A $40M investment has organizational gravity. People built careers on it. Vendors are contracted. The board approved it. The business case was approved before you arrived.

The question is never “should we have started this?” The question is “knowing what we know now, would we start this today?” If the answer is no, the only variable is how fast you can redirect the capital.

What is the Three-Axis Kill Test?

Every product under evaluation gets tested on three independent axes. All three must pass for the product to justify continued investment. If any single axis fails, the product needs restructuring. If two or more fail, it needs to die.

Axis 1: Can it sustain itself on standalone revenue?

Does this product generate enough recurring revenue to justify its cost of operation, development, and support? Training products, platform add-ons, and ecosystem plays often get a pass on this question because someone believes they “drive adoption” of the core product. Test that belief with data. If the product has been live for 18+ months and standalone revenue doesn’t cover its fully-loaded cost, Axis 1 fails.

Axis 2: Does it create measurable pull-through on higher-margin products?

Some products earn their place by making the core product easier to sell. A training simulation that compresses the learning curve, a configurator that reduces pre-sales engineering, a free tier that drives paid conversion. But “we think it helps” is not evidence. Measure attach rates, deal velocity, and conversion lift with and without the product in the sales motion. If you cannot demonstrate a quantifiable lift on the products that actually generate margin, Axis 2 fails.

Axis 3: Is the modernization cost justified by the revenue trajectory?

Even a product with weak current revenue might deserve investment if the market is growing and the product can be modernized at a reasonable cost. But technology shifts can make this math impossible. A VR product built for tethered PC headsets when the market has moved to standalone devices requires a platform rewrite, not a feature update. Run the cost analysis: what does modernization require versus what revenue the product is actually on pace to generate? If the investment doesn’t pencil out on a three-year horizon, Axis 3 fails.

How do you build the case for a kill decision?

The analysis is the easy part. The hard part is organizational. Here is the sequence I have used across multiple kill decisions:

Document the methodology before sharing the conclusion. Present the three axes as an evaluation framework to leadership before revealing which product you’re evaluating. Get buy-in on the framework first. Then apply it. The conclusion follows from the shared methodology, not from your opinion.

Separate the people from the product. When a kill decision includes team reductions, particularly international teams, the communication must be precise. “Your role no longer exists” is categorically different from “your performance was insufficient.” The first is a strategic decision. The second is a judgment of the individual. Conflating them destroys trust across the entire organization.

Redirect the capital visibly. The kill is only half the story. The other half is where the freed capital goes. If leadership can see that the $40M killed program funded the flagship product that launched 18 months later, the narrative shifts from “Nick kills things” to “Nick reallocates capital to things that ship.”

When should you NOT use this framework?

The Three-Axis Kill Test is for products that have been in market long enough to generate data. It does not apply to products in discovery or early beta. Killing a product before it has had time to find product-market fit is a different failure mode, and a more common one. The framework assumes the product has had a fair chance and the data exists to evaluate it.

The takeaway

Portfolio discipline is the most underleveraged skill in product leadership. Every company has at least one product that is consuming capital, attention, and team energy without justifying its existence. The Three-Axis Kill Test gives you a structured, defensible methodology for identifying those products and freeing the resources for things that actually grow the business.

Three axes. All three must pass. If two fail, the product dies. The capital goes somewhere better.