FM FM Software Check

Method

How we gather evidence, and what we refuse to do with it

A comparison site is only worth reading if you can see how it reached its conclusions and where it is weak. This page sets out the source rules, the four evidence states, the criteria we apply and the limits of what a page like this can tell you.

Source rules

  1. Vendor-published sources only

    Product pages, documentation, published pricing, help centres, release notes and case studies the vendor publishes itself. Not sales conversations, not conference hearsay, not another comparison site.

  2. Every claim carries a date

    Software changes. A capability documented in 2023 may be retired or renamed. Each reviewed criterion records the date it was checked, and the page shows the most recent review date at the top and bottom.

  3. No inference from adjectives

    Words like seamless, powerful and world-class carry no information. If a page says a platform is AI-powered, we look for what the AI does and how the vendor describes it working, or we record nothing.

  4. Gaps are published, not smoothed over

    Where a vendor publishes nothing on a criterion, the table says not published. That is a statement about their documentation, not about their software, and the page says so.

The four evidence states

StateMeaningWhat a buyer should do
Documented The vendor publishes the capability in a form a buyer can read, and we link to it. Verify it still works the way the page describes, in a demo.
Partly documented Something is published, but it is thin, undated, or describes the capability without the detail a facilities team needs. Ask for the missing detail in writing before shortlisting.
Not published We looked and found nothing on this criterion. Do not assume it is missing. Ask directly, and test it in the demo.
Not yet reviewed We have not checked this vendor against this criterion. Treat the row as empty, not as a negative.

The criteria, and why these eight

These are drawn from the failure modes that show up repeatedly in Gulf estates rather than from a feature checklist. A platform can be strong on all eight and still be wrong for you. A platform that is weak on any of the first four is a project risk in this market.

1. Asset import

Whether a messy register can be brought in without a data project you did not budget for.

2. Arabic and Hijri

Whether a mixed-language workforce can use it on the same record without a workaround.

3. Offline capture

Whether inspections survive basements, plant rooms and poor coverage.

4. Integrations

Whether BMS, access control and ERP connections already exist somewhere real.

5. ZATCA

Whether billing inside the platform can produce a compliant invoice.

6. Data residency

Where data sits, who can reach it, and what the transfer basis is.

7. Implementation

Whether there is a written plan with named milestones and an owner.

8. Support

Response times, escalation path, local presence and reference customers.

How the score is calculated

Method version 1.0, in force from 19 November 2026. The weights below are fixed before any vendor is assessed and are not adjusted afterwards. If a weight changes, the version number changes and every score is recalculated, so a published score can always be reproduced from this page.

CriterionWeightWhat we test
Arabic and RTL handling15Ask for the interface in Arabic and check a right-to-left screen, not a translated brochure.
Data residency and hosting disclosure10Ask where the database physically sits and get it in writing.
Local regulatory fit (ZATCA, PDPL)10Ask how invoices reach your finance system and how personal data is handled under PDPL.
Integration and open API15Ask for the API documentation and one live integration with a system you already run.
Mobile and offline working15Take the phone into a plant room with no signal and close a job.
AI or automation capability10Ask which decision the system makes without a human, and see the audit trail.
Implementation and migration guidance10Ask for the data import schema and a sample migration plan.
Pricing transparency15Ask for a number in writing before the third meeting.
Total100

The three bands

Each criterion is scored on three bands, and the band has to be supported by a page we can cite:

  • Documented The vendor publishes the capability explicitly, on a page we opened and dated. Band 2.
  • Partial The capability is implied, mentioned in passing, or claimed somewhere other than the product documentation. Band 1.
  • Not published We could not find it in vendor-published material. Band 0. This is not evidence of absence.

The arithmetic

The weighted total is the sum of each criterion's weight multiplied by its band, divided by two. A vendor documented on all eight criteria scores 100. Nothing else is added, and there are no discretionary points.

Coverage, and why some totals are withheld

Coverage is the share of the 100 weighted points where a vendor published something we could check, meaning any criterion scoring band 1 or 2. A vendor documented only on pricing and AI has a coverage of 25 percent, and a total built on that would describe their website, not their software.

So the rule is fixed: below 60 percent coverage we publish the evidence and withhold the total. Above it we publish the number and show the coverage beside it, because a score of 52 at 80 percent coverage means something different from 52 at 62 percent.

Ties and disputes

Where two vendors score the same, the one with higher coverage ranks first. Where coverage is also equal, they share the position and we say so. A vendor can dispute any band by sending a citable page or document; if it holds up, the band changes, the score changes, and the change is dated in the corrections log. We do not remove an entry because a vendor objects to it.

What the score is not

It is not a measure of product quality, fit for your estate, total cost, or support in practice. It measures published, checkable evidence. A thin website is punished here even when the software is sound, and a vendor with a strong documentation team is rewarded. That is a deliberate trade: it is the only way to publish a comparison a buyer can audit without taking our word for it.

What we refuse to do

No paid placement

No vendor can buy a position, a score or a mention. There is no sponsorship on this site and no affiliate arrangement behind any link.

No invented ratings

We do not publish star scores. A number out of five implies a weighting we cannot defend and a precision the evidence does not carry.

No undisclosed ownership

HITEK AI publishes this site and builds CAFMTEK. That is stated at the top of every page, not in a footnote, because a reader deciding where to spend six figures deserves to know who is talking.

No competitor bashing

Where a competitor publishes less than we do, the page records the gap. It does not turn the gap into an insult, and it does not speculate about why.

The limits of this approach

A documentation-based comparison tells you what a vendor has chosen to publish. It cannot tell you whether the software is pleasant to use, whether the support team answers the phone at 2am, or whether the implementation partner assigned to your project knows the building.

It also rewards vendors with mature documentation, which is a proxy for organisational maturity rather than a measure of product quality. A smaller vendor with a better product and a thinner website will look worse here than it deserves. We would rather say that plainly than pretend the method is neutral.

Use this site to build a shortlist and to prepare questions. Do not use it as a substitute for reference calls, a paid pilot, or your own data test.