How We Test

Our mission at is simple: provide thorough and unbiased reviews of online brokers.

For the 12th best online brokers review published in January 2022, a total of 3,075 data points were collected over three months and used to score brokers. This makes the site home to the largest independent database on the web covering the online broker industry.

We also take pride in’s data quality. Every broker that participates in the annual review is afforded the opportunity to complete an in-depth data profile and is offered executive time for an annual update meeting. We then audit each data point to ensure its accuracy.

This rigorous data validation process yields an error rate of less than .001% each year, providing site visitors quality data they can trust.

Testing Devices how we test - testing devices iPhone tablet

All testing is conducted on the devices listed below, using a hardwired, 50 MB Internet connection. Mobile device testing is conducted using an 802.11n wireless WiFi connection; 4G connections are used when WiFi is not available. Unless noted, all websites and web-based platforms are tested using the latest version of the Google Chrome browser.

Smartphone (Apple): iPhone XS with the most current iOS.

Smartphone (Android): Samsung Galaxy S9+, 6.2" 4K Super AMOLED (2960x1440) 64-bit Octa-Core Snapdragon 835 Processor 2.7GHz, 6GB RAM 6.2" with the most current (Android 10) operating system.

Tablet: iPad Air 2 with the most current iOS.

Desktop 1: Windows 10 Pro, 32 GB RAM, 64-bit Operating System, Intel Core i7-9700 CPU @ 3.00 GHz with Radeon RX 550 Video Card and 100GB of SSD Flash Storage.

Desktop 2: Dell XPS 8930 desktop with three ASUS 27″ LED Monitors, Intel Core i7-8700 Processor at 3.2 GHz, Nvidia GeForce GTX 1070, and 16 GB RAM. Windows 10 Pro operating system.

Laptop: MacBook Pro 13”, 1.2 GHz Intel Core m5, Memory 8 GB LPDDR3 RAM, Intel HD 515 1536 MB Graphics card, 500 GB SSD Flash storage with the most current macOS operating system.

Review Categories, Variables

For the 2022 Review, all U.S. equity broker participants were assessed on 205 different variables across nine areas: Commissions & Fees, Research, Platforms & Tools, Mobile Trading Apps, Offering of Investments, Education, Ease of Use, Banking Services, and Overall.

The variable counts of each category were as follows: 

  • Commissions & Fees: 23
  • Offering of Investments: 32
  • Platforms & Tools: 44
  • Research: 47
  • Mobile Trading: 29
  • Education: 14
  • Ease of Use: 1
  • Customer Service: 9
  • Banking Services: 6

Overall Score Calculation

Our eight primary categories are factored in to compile the Overall ranking for each broker. The weightings of each category were as follows: 

  • Commissions & Fees: 10%
  • Offering of Investments: 10%
  • Platforms & Tools: 10%
  • Research: 20%
  • Mobile Trading: 20%
  • Education: 10%
  • Ease of Use: 10%
  • Customer Service: 10%

Awards testing devices mobile apps presented “Best in Class” awards to brokers across all eight main areas as well as additional categories Beginners, Options Trading, Futures Trading, Active Trading, IRA Accounts, and Banking Services. A “Best in Class” designation means finishing in the top five brokers for that category.

The following individual Industry awards were also distributed:

  • #1 Innovation
  • #1 International Trading
  • #1 Investor App
  • #1 Trader App
  • #1 Desktop Platform
  • #1 Web Trading Platform
  • #1 Platform Technology
  • #1 Client Dashboard
  • #1 Overall Client Experience
  • #1 Investor Community
  • #1 New Tool
  • #1 ESG Investing
  • #1 Professional Trading
  • #1 Order Execution
  • #1 Crypto Technology
  • #1 Fractional Shares
  • #1 Broker Podcast
  • #1 Phone Support

Category Scoring Methodology

While most categories are scored using our default method of combining Yes / No points (Variable score) with a 1 - 10 Net Promoter Score (Opinion score), scoring does vary, as shown below.

  • Commissions & Fees - Custom
  • Platforms & Tools - Yes/No with Opinion
  • Research - Yes/No with Opinion
  • Mobile Trading - Yes/No with Opinion
  • Offering of Investments - Yes/No only
  • Education - Yes/No with Opinion
  • Ease of Use - Opinion only

Each individual category score is also translated into a star rating. Specifically:

  • 95% - 100% = 5 Stars
  • 85% - 94.99% = 4.5 Stars
  • 75% - 84.99% = 4 Stars
  • 65% - 74.99% = 3.5 Stars
  • 55% - 64.99% = 3 Stars
  • 45% - 54.99% = 2.5 Stars
  • 35% - 44.99% = 2 Stars
  • 25% - 34.99% = 1.5 Stars
  • 15% - 24.99% = 1 Stars
  • 5% - 14.99% = 0.5 Stars
  • 0% - 4.99% = 0 Stars

How We Score: Yes/No with Opinion

First, points are awarded for each “Yes” answer. The number of points awarded for each "Yes" varies by variable depending on its significance. Next, the total points earned from “Yes” answers is divided by the total points possible to calculate a final Yes/No score, also known as a Variable score.

Example: Broker XYZ has 49 of 57 features for category Platform and Tools and earns 73 out of a possible 79 total points (remember, points allotted to each individual variable varies). Thus, Broker XYZ earns a Variable score of 92.41% for category Platform and Tools.

Next, the editorial team gives an opinion score to each broker for the category. Opinion scores range from 1 – 10, with one being “very bad” and ten being “very good”. Half points are used, e.g. "9.5", so scoring isn't restricted to just whole numbers. Note: While the editorial team does take into consideration the Variable score, the Opinion score is more focused on the editorial view of how well the broker performs overall as compared to the industry standard and its closest competitors.

Example: Broker XYZ does a great job with its Mobile offering, exceeding the industry standard. Not only are a variety of tools and research available, but they are rich with depth and also easy to use. The editorial team awards Broker XYZ a score of "9" for its Opinion score, which equates to 90%.

Finally, to determine a final percentage score for the category, the Variable score and Opinion score are independently weighted and added together. This score is then also translated to a star rating. Weightings vary by category depending on the complexity of the category, e.g. 70% Variable, 30% Opinion.

Example: Broker XYZ earned a Variable score of 93.33% and a Opinion score of 100%. For category Research, the Variable score is 60% of the total and the Opinion score is 40%. Thus, Broker XYZ earns a final score of 96% for Research, which is calculated as "(0.9333*.6)+(1*.4)". Translated into a star rating, Broker XYZ earns 5 Stars for Research.

Commissions & Fees

First, multiple scenarios are assessed for equity trades, options trades, and margin rates. For example, the cost of placing a 500-share market order or the cost of placing a two-legged one-contract options spread, or the margin rate for a $50,000 balance. Brokers are awarded points depending on how low cost they are relative to their peers.

Next, multiple monthly pricing scenarios are assessed for different investor types. For example, the net monthly cost to place one stock trade (500 shares per), representative of a casual investor, or the monthly cost to place 10 stock trades (500 shares per) and 5 options trades (10 contracts per), representative of an active trader. Brokers are awarded points depending on how low cost they are relative to their peers.

Thereafter, all points are tallied to determine a total Scenarios score. Separately, scoring is then conducted to assess Order Execution quality, which applies an Opinion methodology.

Order execution quality dial comparison

Finally, to determine a final percentage score for the category, the Scenarios score and the Order Execution score are independently weighted and added together. Note: While there is currently no universal metric to assess order execution quality, the significance of execution quality is as important as ever given the reliance on PFOF as a source of revenue for brokers that charge $0 for trades. Read our guide to understanding order execution quality.

This final score is then also translated to a star rating.

Customer Service

How we test customer service data graphic

For the past seven years, has partnered with customer experience research group Confero to conduct phone tests from locations across the United States.

Confero agents used one of 10 unique scenario templates provided by (on average, three questions are asked) to assess each broker. To ensure fairness as well as gather the highest quality data, each agent used one single template to test all brokers.

To score each test, three separate buckets are assessed: 1. Time to connect 2. The ability of each broker representative to professionally (and thoroughly) answer each question and 3. A Net Promoter Score for the overall support experience. Response times are 15% of each test score, representative assessments are 35%, and Net Promoter Scores make up the remaining 50%.

Once all individual test scores were compiled, they were combined to determine an overall score for Customer Service. This final score is then also translated to a star rating.

Read Next