How to Compare Investment Platforms

Start by rejecting the single overall score

An overall score is a weighting decision presented as a fact. Whoever built it decided how much cost matters relative to product range, and that decision is where all the content is. If a site does not publish its weights, its ranking is unfalsifiable — you cannot check it, only believe it.

We publish ours on every ranking page, and different rankings use different weights on purpose. Our low-cost ranking weights cost at 55%; our retirement ranking weights account types at 38%. Same data, different question, different order.

Normalise the fee schedules before comparing them

Published fees are not stated in comparable units, and this is where most comparison tables quietly break:

• Some brokers price options per contract; others price per contract per side, which doubles the cost of a round trip
• Some price shares per share; others charge a flat fee, so the cheaper one depends entirely on order size
• Per-order minimums and per-order caps can invert the ranking at both ends of the size range
• Fee caps are sometimes per leg and sometimes per order — a four-leg spread pays the cap four times under one and once under the other

The only reliable method is to model your own realistic annual activity and price it under each schedule. That takes fifteen minutes and beats any table, including ours.

Separate verified facts from asserted ones

When you read a comparison, ask of every cell: could I check this? A commission is checkable — it appears on a published schedule. A regulatory status is checkable — it appears on BrokerCheck. A claim that a platform is 'intuitive' or that support is 'excellent' is not checkable, and if the site does not describe how it was tested, it is an assertion.

This is why our tables mark unverified fields explicitly instead of leaving them blank or filling them with an estimate. A blank cell reads as 'not offered'. An estimate reads as a fact. Neither is true.

Watch for the incentive structure

Most large broker comparison sites are paid per account opened. That is a legitimate business model and it does not automatically corrupt the content, but it creates a consistent gradient: brokers with affiliate programmes get covered; brokers without them get less attention regardless of quality.

The test is not whether a site has affiliate links. It is whether it covers firms that do not pay, and whether it discloses which is which at the row level rather than in a footer. Ask what happens to a broker's position if it ends its commercial agreement.

A checklist you can apply to any comparison

1. Are the criteria and their weights published?
2. Is each data point sourced, and is the source the provider or the regulator rather than another comparison site?
3. Is there a verification date, and is it recent?
4. Are unverified fields marked as unverified?
5. Are commercial relationships disclosed per broker, not just site-wide?
6. Does the site say what it did not test?