Company
How We Test and Research
Last updated
Every CodeNexon guide is built on one of two kinds of evidence: published information that we check and date, or hands-on tests that we run ourselves. This page explains both, so you can judge how much weight to give any page on the site.
Two kinds of guide
| Researched guide | Tested review | |
|---|---|---|
| Based on | Vendor pricing pages, documentation and published standards | Our own accounts, measurements and logs |
| How to recognize it | A note near the top beginning "How this guide was researched" | A note stating the plan, region, tools and dates of the tests |
| What it can tell you | What a product costs, what it includes and how options compare on paper | How a product performed when we used it |
| What it cannot tell you | How fast or reliable a product is in practice | How it will perform on a different plan or in a different region |
We never describe a product as tested unless we ran the test.
How researched guides are made
- Start from the primary source. Prices, limits and requirements come from the vendor's own page or the relevant standards body, not from other review sites.
- Record the date. Each guide states the day its figures were checked.
- Note the currency and region. Vendors show different prices in different countries. Where we could not see US dollar pricing directly, the guide says so and marks the figure as approximate.
- Do the arithmetic in the open. Multi-year totals, percentage increases and per-sale fees are calculated from the published rates, with the inputs shown.
- Link everything. The Sources section at the end of each guide lists the pages used.
- Say what is missing. If a detail could not be confirmed, the guide says so.
The test protocol for hosting reviews
When we review a hosting plan hands-on, these are the tests it goes through.
| Test | How it is run | Tools | Frequency |
|---|---|---|---|
| Speed | The same standard test site is deployed on each plan. Load time and server response are measured from several regions, and the median is reported | WebPageTest, GTmetrix, Lighthouse | Three or more runs on different days |
| Uptime | Each test site is monitored continuously and every incident is logged with a timestamp | An uptime monitor such as UptimeRobot | At least 30 days before a verdict |
| Load handling | Modest, controlled load is applied to see how response time changes, within the provider's acceptable-use terms | k6, Loader.io | Once per plan |
| Support | At least three requests on different topics: basic, technical and billing. Response time and outcome are recorded | A support log | Per review |
| Ease of use | Key tasks are timed: sign-up, installing WordPress, creating and restoring a backup, pointing a domain, setting up SSL | Stopwatch and screenshots | Per review |
| Pricing | First-term price, renewal price, required term and add-ons are recorded with a dated screenshot | Screenshot archive | Monthly for key posts |
| Features and limits | Claims about backups, staging, CDN, email, storage and refunds are checked in the dashboard and documentation | Dashboard and vendor docs | Per review |
The test protocol for software reviews
- Define three to five real tasks a reader would do, such as sending a campaign, importing a list or creating an invoice.
- Complete them on a real trial or paid plan, recording time taken and any friction or errors.
- Check plan limits, integrations, export options and data portability.
- Test support where there is a free way to do so.
- Record pricing for each tier with the date.
- Record what the product does badly. Every review includes real downsides.
How products are scored
Tested products in the same category are scored with the same rubric. Each area is scored and weighted, for a total out of 100.
| Area | Hosting weight | Software weight |
|---|---|---|
| Performance or core capability | 25 | 25 |
| Reliability | 20 | 10 |
| Support | 15 | 10 |
| Features and limits | 15 | 20 |
| Pricing and value, including renewal | 15 | 20 |
| Ease of use | 10 | 15 |
| Total | 100 | 100 |
Scores are given only to products we have tested. Researched guides do not carry scores.
Keeping results current
- Prices on our most-read guides are rechecked monthly.
- Reviews and comparisons are reviewed in full every quarter.
- Speed and support tests are repeated every 6 to 12 months.
- When a vendor changes its pricing or features materially, the affected guides are updated and the change is listed in the guide's update history.
What does not influence a result
No company can pay for a score, a ranking or a mention. Our funding and our rules for commercial relationships are set out on the Disclosure page.
Tell us if we got something wrong
If a figure is out of date or a step did not work for you, write to info@codenexon.com with the page address. Corrections are handled as described in the Editorial Policy.