Self-hosted software 2026: what the report measures
Self-hosted software in 2026: what the store report measures, where its figures come from, and how to read it without re-typing it.
This store publishes a report called the state of self-hosted software in 2026, and this page is about how to read it. Deliberately, no figure from it is copied here. A number typed into a second place is a number that can disagree with the first, and the report is generated from the price record rather than written by hand, so the live page is always the authority.
What can usefully be described is the method: what the report counts, what it refuses to count, and what its shape does and does not support.
What it measures
The report is built entirely from the store's own price record. Each row there is a vendor pricing figure read at the vendor, quoted verbatim, dated, and linked to the page it came from. The report reads those rows and does four things with them.
It counts. How many vendor figures exist in the record, how many of those carry a published monthly price, and how many publish none at all. That last group is the interesting one, and the report keeps it separate rather than dropping it.
It finds the middle. The median published monthly figure across the priced rows, plus the lowest and the highest. A median rather than an average, because a single enterprise tier would otherwise drag the answer somewhere unrepresentative.
It groups by job. Rows are gathered by the category of work they do -- the tools that book appointments together, the tools that keep a register together -- and each group shows how many vendors were read, the median monthly figure among them, and the one-time price of the product here that does the same job.
It ranks the gaps. For each row it takes the published monthly figure across a three-year horizon and subtracts the one-time price, then sorts. The largest gaps are the headline.
Everything on the page follows from those four operations on dated rows. There is no survey, no panel and no market sizing anywhere in it.
What it refuses to count
Unpublished prices. A vendor who gates pricing behind a sales call is counted as gated, not estimated. The report prints how many vendors sit in that group, which is itself a finding about this market.
Figures in another currency, converted. The record does not convert, and neither does anything built on it.
Anything without a source and a date. A row with no live source link does not enter the record and therefore cannot enter the report.
What the shape does not support
This is the part worth saying plainly, because a reader could over-extend the conclusions.
It is not a market study. The vendors in it are the ones this store has read because it sells something comparable. That is a deliberately non-random sample, and the report is a comparison of published prices rather than a description of an industry.
The three-year horizon is an illustration. It is one multiplication, chosen because it is a common planning period, and it assumes the subscription price never moves and the seat count never grows. Both assumptions are generous to the rented column.
The one-time column is not zero-cost. Hosting and your own time are real, and the store has a separate page about exactly that: what self-hosting actually costs.
A median is not your quote. Your price depends on your seat count and your tier, and the median tells you where the middle of the market's published pricing sits, not what you will pay.
How to use it
Read it for the shape rather than the totals: which categories have many vendors and which have few, how many refuse to publish, and where the distance between renting and owning is widest. Then go to the observatory for the individual row that applies to you, check its date, read the verbatim quote, and do the multiplication with your own numbers. The report is the overview; the record is the evidence.
The own-once side of any row is a product page -- Slotly and Rentara are two of them -- and each of those carries its own limits as plainly as its features.
Try the Slotly demo ↗Live, on sample data, no sign-up.
Questions people ask
Why does this page quote no figures from the report? Because the report is generated from the live price record. Copying a figure here would create a second place for it to be wrong, and the live page is the authority.
Is the vendor list representative of the whole market? No, and the report does not claim it is. It is the set of vendors this store has read because it sells something comparable, which makes it a comparison of published prices rather than a market study.
Where do the dates come from? From the price record. Every row was read at the vendor on a stated date, and the report carries the newest of those dates as its own.
Where to look next
The report is the state of self-hosted software in 2026, the evidence behind it is the observatory, and the honest account of the owned side is what self-hosting actually costs. Prices are on the observatory; read them on the date shown there.