How fast is Product Filter with 100,000 products?
Every filter plugin says it is fast. WooBeWoo (WBW) publishes the measured numbers for Product Filter 4.0, with the machine and the date beside every number, and this page shows exactly what the script printed. The script, dev-tools/benchmarks/run.php, is in our source repository; the plugin ships the results in benchmarks/RESULTS.md.
Most shopper requests are served from the cache. Cold numbers are the worst case, right after the index changes.
Method
How we measure filter speed
- 01
Count query time
The time to compute the counts of every block for the empty query: what the shop page needs for its first paint. Served from the facet index.
- 02
Request time per combination
The server side time of a full filter request (product ids for one page, every count, the products HTML) for three typical combinations a shopper makes.
- 03
Median of 10, caches dropped
Every number is the median of 10 runs. The index version is bumped before each run, so no cached count is reused.
- 04
Machine and date beside the numbers
The script prints the machine it ran on and the date. A number without both is not published here.
Measured
Measured results (24 Oct 2026)
Measured on A 4 core test machine: Linux x86_64, PHP 8.3.6, MariaDB 10.11 with a 512 MB buffer pool, WordPress 7.1.3, WooCommerce 11.2.1, Storefront. Median of 10 runs. Cold means every cache off and every block recounted; cached is the same request from the cache., 24 October 2026, with Product Filter 4.0 and the facet index. Products is the catalog size; index rows is the size of the facet index for it.
| Catalog | Index rows | Count query (ms) | Request (ms): One attribute, cold | Request (ms): Size + color + in stock, cold | Request (ms): Category + price + sale + sort, cold | Request (ms): Any request, cached | Request (ms): Full page over HTTP | Request (ms): Legacy 3.x full page |
|---|---|---|---|---|---|---|---|---|
| 10,000 products | 62,297 | 144 | 155 | 283 | 190 | 10 to 15 | 217 to 222 | 256 to 584 |
| 50,000 products | 312,114 | 455 | 624 | 1,189 | 924 | 10 to 11 | - | - |
| 100,000 products | 624,717 | 1,440 | 1,459 | 2,247 | 1,822 | 10 to 18 | - | - |
How the script works
The script, dev-tools/benchmarks/run.php, is in our source repository; the plugin ships the results in benchmarks/RESULTS.md. The script picks the first 4.0 filter on the test store (or builds a temporary one), runs each measurement 10 times with the caches dropped, prints a Markdown table you can paste anywhere, and stores the result for the Speed and SEO screen. Nobody has to take our word for it.
Give every shopper three taps to the right product.
Product Filter Pro: $99 once, or $59 a year.