Methodology
How a benchmark gets built, what the sample covers, and when each report is re-read.
1. Defining the frame
Every benchmark starts with a buying scenario, not a product list. "A 25-seat business on annual billing" or "150 linear feet of 6 ft cedar on flat ground" fixes the scope so prices are comparable. The scenario is printed at the top of the report.
2. Collecting prices
Software prices come from public price pages, captured and archived with the date read. Services prices come from written quotes collected against a single written scope, by metro. We never use vendor-supplied averages, and we never price from an NDA.
3. Normalising
Plans are matched on capability, not on tier name. Where a provider bundles something others sell separately, the add-on price is listed in its own column rather than folded into the headline number. Taxes and regulatory fees are excluded and quantified separately.
4. Reporting
We publish the median and the full range. Where the distribution is skewed, the report says so in the first paragraph. Counts are reported as "10 of 14" rather than as percentages when the sample is under 30.
Sample sizes and cadence
| Report↕ | Sample↕ | Collection window↕ | Cadence↕ | Next update↕ |
|---|---|---|---|---|
| Email marketing software pricing | 14 providers, 42 plans | 6 to 17 Oct 2026 | Quarterly | Feb 2027 |
| VoIP per-seat pricing | 12 providers, 36 tiers | 1 to 9 Oct 2026 | Quarterly | Jan 2027 |
| Fence installation cost | 312 quotes, 96 contractors | Jul to Sep 2026 | Twice a year | Apr 2027 |
Corrections
Send a correction to admin@buyerbenchmarks.com with the source. Confirmed errors are fixed within five business days, and the report carries a dated correction note from then on.
What we do not do
- No scores, no quadrants, no vendor rankings built from opinion.
- No prices under NDA, and no "contact us" numbers inferred from hearsay.
- No averages supplied by a vendor about its own category.