1. We define the buyer before we test
Every category names a specific buyer. For hosting it's "a small business running its own site or a small store on a shared plan." The pick is the best choice for that buyer, not in the abstract. A different buyer gets a different answer, and we say so on the page.
2. We build the scorecard before we look at products
We list the criteria that matter to that buyer and assign each a weight that adds up to 100%. The weights are locked before scoring begins. For small-business hosting:
| Criterion | Weight | What we measure |
|---|---|---|
| Total cost of ownership | 25% | Real cost over a 4-year term, including renewal rates |
| Ease of control panel | 25% | Timed task tests with a first-time user |
| Real-world speed | 20% | TTFB and load time on a stock WordPress install, multiple regions |
| Support | 15% | Response time and answer quality across 5 real tickets |
| Included features | 10% | SSL, email, backups, CDN, migration: included vs paid add-on |
| Uptime | 5% | Monitored over the full test window |
3. We test the whole field
We buy the plans ourselves, at retail, on the same billing terms a reader would. We do not use press accounts, and vendors do not know which of their plans we hold.
4. We score, weight, and name one winner
Each criterion is scored 0 to 10, multiplied by its weight, and summed. The highest total wins. If two products finish within 0.3 points we break the tie in favour of the lower-risk choice for the buyer (usually price or ease) and explain the call.
5. We publish the runners-up
Every pick lists the second and third place and the specific reason each lost. If your priorities differ from our weighting, the runners-up section tells you where to go instead.
6. We date it and revisit it
Each pick carries a "last tested" date. We re-check a category at least once a year, or sooner if a major product change or price change lands. When the winner changes, the old pick stays visible in the changelog.
Rules we hold ourselves to
- The scorecard is set before testing and never adjusted to change the outcome.
- Affiliate relationships never influence a score. We link to products we rank low too.
- A sponsor can never be the pick in their own category. Sponsorship buys visibility, never a verdict.
- If we can't test something properly, we don't rank it.