The two biases that decide most selections
Familiarity with the current system and enthusiasm for the newest option pull in opposite directions and are both about comfort rather than fit.
Two forces distort software selections more than any analytical failure. One is attachment to what exists. The other is attraction to what is new. Most evaluation groups contain both, and the outcome often reflects which is better represented rather than which product is better.
Incumbent bias
The current system has one enormous advantage that appears in no comparison: everyone knows how to use it, including its faults. Its problems are familiar and therefore feel smaller than the unknown problems of a replacement. When comparing claims around mouse-jiggler detection, mouse jiggler detection software gives one vendor’s framing to test against documentation, demos and references.
It shows up as requirements that describe the existing tool, as scepticism applied asymmetrically — the incumbent's flaws are known limitations while a candidate's flaws are red flags — and as an unstated preference for the option that requires nobody to learn anything.
If a candidate would be rejected for a limitation, ask whether the incumbent has the same one. Frequently it does, and it has been tolerated for years.
Novelty bias
The opposite pull, usually from whoever is most technical or most frustrated. The newest product looks better because it is designed against current expectations, and its weaknesses are less visible because fewer people have hit them.
It shows up as enthusiasm for a young product with a small customer base, discounting of maturity and support, and an assumption that gaps will be filled soon because the roadmap says so.
Both are managed the same way
Neither bias is removed by naming it. What works is process: criteria fixed before candidates are seen, the incumbent included as a formal option scored on the same sheet, and structured trials that expose both the new product's rough edges and the current system's real cost.
Including the status quo as a scored candidate is the most effective single measure, because it forces its disadvantages to be written down alongside everyone else's.
Watch for the sunk-cost argument
'We have invested too much in this to change' is a statement about the past. The money is gone regardless of what happens next, and the only relevant question is which option costs less from today.
The version that is legitimate is different and worth distinguishing: switching cost. Migration, retraining and disruption are real future costs and belong in the comparison. Past investment does not.
Bring in someone with no history
Ask someone uninvolved — another department, an adviser, a peer at a similar organisation — to review the shortlist and the reasoning for an hour.
They will ask the obvious questions that everyone inside the process stopped asking weeks ago. It is the cheapest correction available and it is skipped almost universally.
For cloud-service due diligence, NCSC cloud guidance provides an independent reference point.