How Heavy Are Recipe Websites? We Measured 50 of Them (2026 Study)
Published July 2026 · Data collected 20 July 2026
Quick answer: the median recipe page transfers 3.3 MB, makes 120 network requests, and contacts 27 third-party hosts — mostly ad and tracking infrastructure — and the ingredients don't appear until a median of 3,925 pixels down the page, about four and a half screens of scrolling. We measured this on the 50 recipe websites that home cooks convert most often with Drizzlelemons, in a real browser, with no ad blocker. The full method and every number are below, and you are welcome to cite them.
What we measured, and how
A note on who is asking: we build Drizzlelemons, an ad-free recipe converter — so yes, we have a horse in this race. That is exactly why everything here is reproducible: one page per site, loaded the same way, measured with the same tool, and reported with medians rather than cherry-picked extremes.
- The sites: the recipe domains our users convert most (from our own July 2026 conversion data) — BBC Good Food, AllRecipes, NYT Cooking, RecipeTin Eats, Delish, Serious Eats, Sally's Baking Addiction, and 47 more. One popular recipe page per site.
- The browser: headless Chromium at 1,280 × 900, UK location, no ad blocker, no cookie banners dismissed — the page exactly as a first-time visitor receives it.
- The load: initial load plus an eight-second dwell with a full scroll to the bottom and back, so lazy-loaded ads and videos count, because a real cook scrolls.
- The numbers: transferred bytes and request counts straight from the Chrome DevTools Protocol; third-party hosts are unique hostnames contacted that don't belong to the site itself; ingredient depth is the pixel position of the first visible "Ingredients" heading or block.
- Exclusions: of 54 pages attempted, two failed to load in the harness and two served bot-check placeholders instead of the page; they are excluded, leaving 50 valid measurements.
The headline numbers
| Metric (per recipe page) | Median of 50 sites | Worst measured |
|---|---|---|
| Data transferred | 3.3 MB | 43.2 MB (Delish) |
| Network requests | 120 | 512 |
| Third-party hosts contacted | 27 | 67 |
| External scripts | 36 | — |
| Iframes (mostly ad slots) | 7 | — |
| Page height | 17,009 px (~19 screens) | 55,628 px |
| First "Ingredients" position | 3,925 px down (~4.4 screens) | 36,472 px down |
For context, the median page on the entire web transfers about 2.2 MB on mobile according to the HTTP Archive's Web Almanac — so a typical recipe page is roughly half again heavier than the median of all web pages, to deliver what is ultimately a list of ingredients and some steps.
The heaviest recipe pages we measured
| Site | Transferred | Requests | Third-party hosts |
|---|---|---|---|
| Delish | 43.2 MB | 323 | 21 |
| Gozney (recipes section) | 40.6 MB | 512 | 67 |
| The Pioneer Woman | 9.9 MB | 269 | 24 |
| Epicurious | 8.2 MB | 98 | 14 |
| Gourmet Traveller | 8.0 MB | 249 | 64 |
| Sally's Baking Addiction | 6.0 MB | 123 | 39 |
A 43 MB recipe page costs a meaningful chunk of a capped mobile data plan — for one recipe. Load ten recipe pages like the median one while planning a week of dinners and you have transferred about 33 MB — most of it advertising infrastructure rather than food. Do the same on a site at the heavy end and ten pages can burn through more than 400 MB.
How far down is the actual recipe?
On the 45 pages where our harness could locate a visible ingredients block, it started a median of 3,925 px from the top — about 4.4 screens on a laptop. On 20 of those 45 pages, the ingredients began more than 5,000 px down. The deepest we measured: 36,472 px (Alexandra Cooks' sourdough guide), nearly 41 screens of story, photos, and ad slots before the first ingredient. This is the experience the "jump to recipe" button exists to apologise for — and it is why the recipe URL trick resonates with so many cooks.
The same recipe, converted
Measured with the identical method, a converted recipe page on Drizzlelemons transferred 727 KB, made 63 requests, contacted 1 third-party host (our analytics), contains zero iframes, and puts the ingredients 544 px from the top — on the first screen. That is 4.5× less data and 27× fewer third-party hosts than the median recipe page. We are obviously not a neutral party here, but the method is identical and the gap is not subtle.
Why are recipe pages like this?
Not because food bloggers are villains. Display advertising is how independent recipe developers get paid, and ad networks pay per impression — which rewards long pages, many ad slots, and heavy retargeting scripts. The 27 third-party hosts on the median page are mostly ad exchanges, header-bidding wrappers, consent platforms, and analytics vendors, each adding requests and JavaScript. The result is a structural problem, not a moral one: the reader's experience and the writer's income currently pull in opposite directions. (Our longer essay on this: why online recipes are so annoying.)
Use this data
The numbers, method, and site list above are free to cite in articles, videos, or research with attribution — link to this page so readers can check the methodology. If you would like the raw per-site measurements, email contact@drizzlelemons.com and we will send the JSON. We plan to re-run the study periodically; numbers are from a single crawl on 20 July 2026 from a UK connection, and individual pages will vary by visit, region, and consent choices.
Frequently asked questions
Why are recipe websites so slow?
Because the median recipe page loads 3.3 MB across 120 requests and runs roughly 36 external scripts, contacting 27 third-party hosts — mostly ad and tracking infrastructure (July 2026 measurement). The recipe itself is a few kilobytes of text; nearly everything else is advertising infrastructure, which is also why pages feel janky while ads load and reflow around the content.
Why do recipe websites have so many ads?
Display ads are the main way independent recipe sites earn money, and networks pay per impression — so longer pages with more ad slots earn more. That incentive produces the long introductions, the 19-screen page heights, and the ingredients starting four screens down that we measured in this study.
How were these measurements taken?
Headless Chromium at 1,280 × 900 with no ad blocker, loading one popular recipe page from each of the 50 most-converted sites in our data, dwelling eight seconds with a full scroll so lazy-loaded ads count, and recording transfer sizes and requests via the Chrome DevTools Protocol on 20 July 2026.
Can I cite these numbers?
Yes — the data and methodology are free to use with attribution and a link to this study. For the raw per-site JSON, email contact@drizzlelemons.com.
Want the clean version of any recipe? Paste the URL into the free recipe converter — no ads, no popups, ingredients on the first screen.