Method
How we test and compare
Published in full, before results rather than after, so you can check whether we followed it.
The process
Scope the real requirement
Start from the decision a team is facing, not the category a vendor invented. If the honest answer is that nobody needs this category, that is the guide.
Publish the weights first
Scoring weights go up before testing starts. A weight cannot then be adjusted to justify a result we happen to prefer.
Test against the awkward cases
Hybrid identity, delegated admin, offboarding, and whatever breaks past a few hundred seats. Not the happy path in the vendor demo.
Price the renewal
First-year pricing is marketing. We compare year two at a stated seat count, including per-user creep.
Name who should not buy it
Every guide ends with a disqualifying case. A recommendation without one is an advert.
Default scoring weights
Individual guides may adjust these, and say so at the top when they do.
- Does the core job35%
- Coverage gaps25%
- Cost at renewal25%
- Admin overhead15%
Commission rate is not an input and never will be.
Conflicts of interest
No paid placements
No vendor has ever paid to appear here, to rank higher, or to have a competitor left out. There is no sponsored tier and there never will be.
Affiliate links change nothing
Some links earn a commission. The commission rate is never a ranking input, and products with no affiliate programme at all are recommended when they are the better answer.
Tested, not summarised
Every recommendation is based on running the product against the awkward cases: hybrid identity, delegated admin, offboarding, and whatever breaks past a few hundred seats.
Trade-offs stated
Each guide names what a product does badly and who should not buy it. A recommendation with no disqualifying case is an advert, not advice.
Found us getting something wrong? Corrections are published on the guide itself with the date, not quietly edited in. Tell us what we missed.