Try to find out how many European companies actually use a CRM. The first result cites a second article, which cites a vendor blog post, which cites a survey that is either unnamed, unpublished or a decade old. By the time you reach the bottom of that chain — if there is a bottom — the number has been repeated often enough to sound like a fact.
That pattern is why The Knowledge Engineers caught our attention. It is a new research publication covering the business software market, and its organising principle is stated in its own strapline: research you can check. Every figure it prints carries the dataset it came from, the reference period it covers and a confidence label. We spent an afternoon with it because the approach is close to the one we apply to our own directory, and because the gap it is trying to close is a real one.
What it publishes
The site has three parts. The first is a set of industry reports built from primary statistics rather than from other articles. At the time of writing there are ten, covering cloud adoption, artificial intelligence, e-commerce, ICT security, the skills shortage, digital marketing and corporate training, plus two analyses computed from the site's own vendor database. Together they hold several hundred numbered data points.
Each report reads like a statistics page, but with the plumbing left visible. A data point
does not simply state that 52.7% of European companies pay for cloud services; it names the
Eurostat dataset (isoc_cicce_use), the reference period (2025) and the date the
statistics office last published it. The reports are generated by a script that queries the
source directly, which means the figures can be reproduced by anyone willing to open the same
dataset with the same filters.
The second part is a market data index: every figure the site has published, listed once, with its source, period and confidence label, downloadable as CSV and JSON. There is also a set of country profiles — the same indicators cut per member state, each against the European average with the rank beside it. For anyone who has tried to assemble that comparison by hand from the Eurostat data browser, the time saving is obvious.
The third part is software advice: ranked category guides covering some two hundred products across twenty-five categories, from CRM and ERP to password managers, web hosting and point of sale. These are the pages that pay for the research, and the site says so.
The thing it refuses to do
Most comparison sites give each product a score out of ten, usually assembled from weighted sub-scores: features 40%, ease of use 30%, value 30%. It looks like measurement. It is not. Nobody outside the publisher can check how a product arrived at 8.4 rather than 7.9, and the weights themselves are a judgement presented as arithmetic.
The Knowledge Engineers does not publish scores. In their place the category pages carry counts that can be verified from the page itself: how many of the products publish a price, how many offer a usable free tier, how many of the vendors are established in Europe. It is a smaller claim, and a defensible one. On a site whose entire pitch is verifiability, inventing a decimal would be the first thing a sceptical reader would notice.
The same discipline shows in how the products are written up. Each entry states plainly where the product costs you time or money, who it is right for and — unusually — who it is wrong for. Reading the password manager guide, the entry for LastPass says out loud that the 2022 breach exposed encrypted vaults and unencrypted URLs, and that a security review has to address that rather than move past it. Most affiliate-funded comparison pages find a gentler formulation.
How it is funded, and why that matters
Research that hides its funding is marketing. The Knowledge Engineers states four revenue lines on its services page: commissioned and syndicated reports, data licensing, commissioned product analysis, and commission on some outbound links. Every category page carries a notice above the ranking saying that commercial relationships can affect which products are covered and where they appear, and that they never change what an entry says.
The commissioned analysis offer is the interesting one. A vendor can pay to have its own product taken apart, with the commissioner named on the page permanently, the right to correct factual errors but not judgements, and the option to shelve the piece before publication rather than receive a softened version. That is a more honest arrangement than the sponsored-content norm, where the label fades after a few months and the criticism was negotiated away before publication.
Whether readers reward that is another question. Publishing your own drawbacks is commercially inconvenient, and the sites that do it tend to be smaller than the sites that do not.
Who writes it
Five named researchers, each taking a different role on every page: one writes, a second edits, a third checks the claims against the source. Nobody edits or signs off their own work, and which role falls to whom is fixed per page, so a page keeps the same three names when it is rebuilt. The bylines carry photographs and link to individual profiles, and every page shows when it was published and when it was last reviewed.
That is a higher standard of attribution than most of this market manages. A great deal of software comparison content is published under a house name, or under a byline that turns out to belong to nobody in particular.
What we would want to see next
Three things. First, more sources: the reports lean heavily on Eurostat, which is excellent for enterprise ICT adoption and silent on plenty else. National statistics offices, sector regulators and company filings would broaden the base without weakening the standard.
Second, time series. Almost every figure on the site is a snapshot of the most recent survey wave. The interesting question is usually the direction, and the underlying datasets go back years.
Third, the honest caveat about a young publication: a method is a promise until it survives contact with a commercial conversation the publisher would rather not have. The structure is right. Whether it holds when a paying vendor dislikes a paragraph is something only time answers.
Why we are writing about it
European Purpose exists because where a company is established decides whose laws apply to the data it holds, and because that fact is usually buried in an appendix. The Knowledge Engineers is doing something adjacent: making the provenance of a number as visible as the number itself. Different subject, same instinct.
It is also a useful complement to what we do. Our directory answers where a tool comes from and what the European alternative is. Their reports answer how many companies actually use the category at all, and what they pay for it. Between the two, a buyer has both the shortlist and the context.
The site is at theknowledgeengineers.com. The reports are free to read and free to quote with attribution, and the underlying data index is downloadable, which is more than most publishers in this market offer.
The reports and the data index are free to read and free to quote with attribution at The Knowledge Engineers. For European alternatives to the tools they rank, browse our directory of European services.