We captured pricing for 131 project management tools. 29% of plans have no public price.

Between 10 July and 12 August 2026 we read the pricing pages of 131 project and portfolio management products and stored every price we found: the amount, the currency, the billing period, the commitment, the seat band, the source URL, and the date we captured it.

663 individual prices, each one traceable back to the vendor page it came from.

This post is what that exercise turned up. Not a ranking, and not a recommendation. Just the shape of the data, including the parts that are missing.

Nearly a third of plans don't publish a price at all

Across those 131 products there are 407 plan tiers. 290 of them publish a number. 117 do not, which is 29% of the plans on the market.

That is not a data-collection failure. Those pages exist, they describe the tier in detail, they list its features, and where the price should be they say "Contact sales". The information is deliberately withheld.

It clusters where you would expect. The enterprise tier is very often the one without a number, which means the plan a large buyer is most likely to end up on is the plan they can research least. You can compare the Starter tiers of forty tools in an afternoon. You cannot compare their Enterprise tiers at all without forty sales calls.

Worth saying plainly: quote-only is a legitimate way to sell software. Complex deployments genuinely do need scoping. But it is worth knowing how much of the market it covers before you assume a shortlist can be built from public information.

What a seat actually costs

Filtering to plans sold per user, per month, on a straightforward per-seat basis (not banded, not usage-metered), leaves 226 rates.

101 of the 226, or 45%, are under $15 a seat.

The spread is the interesting part. A 66x range between the cheapest and the dearest seat, in what is nominally one category, tells you the category is not one category. A $1.49 seat and a $99 seat are not competing for the same buyer, and a comparison that puts them in the same table without saying so is not comparing anything.

The numbers that nearly went into this post

While preparing it, our own analysis script reported the most expensive plan in the catalog at $22,317 per user per month.

The real figure is $224.46. It covers a band of "1-3 concurrent users": it is the total for the band, not a per-seat rate. The row said so, in its own description, right next to the number. The script had read the unit code and ignored the sentence beside it.

We caught it because the number was absurd enough to check. The lesson we took is not "check absurd numbers". It is that a price is only meaningful with the text around it. A figure lifted out of its row, away from its band and its commitment and its billing period, is not a smaller truth. It is frequently just wrong.

This is most of the work. Not fetching pages, but reading them carefully enough that the number means what you think it means.

Why we built a catalog instead of a ranking

There is no shortage of "best project management software" lists. They rank tools against a scoreboard the reader did not set, using weights the reader cannot see, and the ordering is the same for a 5-person studio and a 500-person engineering org.

The weights are the actual decision. Whether Gantt charts matter more than time tracking is a fact about your team, not about the software. A ranking that hides its weights has answered a different question than the one you asked and presented it as an answer to yours.

So OpenDecision holds the catalog and lets you set the weights. You describe what you actually need, you see and adjust the weighting, and you get an ordered shortlist with the reasoning attached. Every price in it carries the vendor page it came from and the date it was read, so you can check any number without taking our word for it.

It is free, and there is no account required to start.

What we will not claim

We publish pricing and what each plan includes. We do not publish feature-by-feature scores, user reviews, or star ratings, because we do not hold that data and generating it would make the rest untrustworthy by association.

Where a vendor does not publish a price, we say so rather than estimating one. A modelled price is a made-up price with better manners.

---

If you want to look at the underlying data, the catalog is at opendecision.com/products: 131 products, every plan, every price with its source and capture date.

Figures in this post are from the catalog as captured between 10 July and 12 August 2026 and will drift as vendors change their pricing.