What Home Cooks Really Want

Executive Summary

Recipe websites have not lost relevance - they have lost continuity.

In a blind survey of 354 active home cooks with no awareness of Allspice, we observed a consistent pattern: people still discover recipes on websites - but they frequently stop using the page once cooking begins. The issue is not trust, quality, or even competition from AI. The issue is execution friction.

Most users approach recipe pages with a clear goal: complete a cooking task. Yet the page they arrive on is optimized for reading, not cooking. As a result, the cooking session often moves somewhere else - into screenshots, notes, videos, memory, or occasionally AI tools.

This is measurable:

• 72.9% immediately bypass the article to reach ingredients or steps
• 80.5% reconstruct the recipe off-page while cooking
• 70.6% actively avoid recipe sites because they are frustrating to cook from
• Only 4.8% actually replace recipes with AI

When cooking becomes easier, behavior reverses:

• 80.8% would use the page during cooking with guided mode
• 73.2% would cook from the site more often
• 70.1% would return to the same publisher
• 65.8% would try more recipes

The implication is straightforward: Users are not abandoning recipe websites - they are abandoning difficult cooking sessions.

(Margin of error ±4%, n=354)

About This Research

Participants were recruited through a third-party panel and screened to ensure they cook regularly at home and actively use online recipes.

This random sample represents average home cooks:

• 66.4% cook daily
• 74.4% cooked at least 3 online recipes in the past month
• 87.6% use search engines to find recipes

We measured behavior, not opinions:

• How people navigate recipe pages
• What causes frustration
• Where the cooking session moves
• Whether AI is replacing sites
• How improved cooking UX changes loyalty

All percentages refer to the quality-filtered dataset (n=354).

Discovery vs. Cooking Are Two Different Jobs

Users arrive on recipe sites through traditional discovery channels - primarily search. But once they decide to cook, their behavior changes immediately.

The first action on a recipe page is rarely reading:

• 48.6% scroll directly to ingredients
• 24.3% click "jump to recipe"
• Only 11.3% read most of the article

Nearly three-quarters of users attempt to bypass the content layer immediately. During cooking, attention narrows to task execution.

Why Users Avoid Recipe Websites

A large majority (70.6%) report often or sometimes avoiding recipe websites specifically because they are frustrating to cook from.

The reasons are highly consistent and overwhelmingly execution-related:

Notably absent: recipe quality.

Even among users who say they use recipe sites less than before, the leading cause is worsening site user experience experience (53.6%), exceeding popularization of AI tools (21.4%).

The pattern is consistent across open-ended responses: people struggle to execute recipes, not to find them.

What Happens Instead of Cooking on the Page

When execution becomes difficult, the cooking session migrates. The most common off-page workarounds for recipe viewing:

• Screenshot ingredients or steps - 56.7%
• Find a video version of the same dish - 23.9%
• Copy to notes - 22.6%
• Find another site - 22.3%
• Print - 16.6%
• Ask AI - 15.0%

The important implication: the publisher often loses the cooking session even when they win the search click.

The Actual Role of AI

AI is present, but it is not the primary replacement.

Only:

• 10.2% regularly leave for AI
• 4.8% primarily cook using AI instead of the page

When users do consult AI:

• 24.3% fully or mostly continue cooking or discovering recipes using the AI platform
• 28.6% go back and forth between AI and the recipe
• 47.1% ask a quick question and return

A much greater percentage of users, 57%, would use an AI assistant if built into the recipe page.

What Changes Behavior

We tested two other improvements: a concise overview and guided cooking mode.

Overview: Decision Confidence

Users primarily evaluate recipes based on:

• Ingredients - 88.4%
• Steps - 66.1%

A clear summary improves commitment before saving or cooking. 81.5% rated a top-of-page search engine-style overview useful (mean 5.42 / 7). Interestingly 44% of respondents would view an AI-generated summary more positively than a human-written summary; only 5% would view an AI-generated more negatively than a human-written summary.

Guided Mode: Execution Confidence

If the recipe page offered a guided step-by-step mode (keeps your place, timers, per-step ingredients):

• 80.8% would use the page during cooking
• 73.2% would cook from the site more often
• 70.1% would be more likely to return to the site
• 65.8% would be more likely to try more recipes from the site

The strongest response comes from the highest-risk users - those who currently leave the page to cook.

Loyalty Is Driven by Success, Not Email

Modern recipe publishing relies heavily on email subscriptions to maintain repeat readership. However, the data suggests email is compensating for a loss of on-page continuity rather than creating loyalty itself. Email subscriptions correlate weakly with loyalty:

• Only 4.6% say subscribing made them more loyal
• 71.5% uncomfortable with hashed email sharing after newsletter subscription
• The majority of users subscribe after trusting a site, not before

In contrast, successful cooking directly predicts return behavior:

• 70.1% return
• 70.1% bookmark
• 65.8% try more recipes

This indicates loyalty is outcome-driven rather than marketing-driven.

Cooking Spans Many Sites - Retention Does Not

Home cooks assemble meals across many sources. In the survey:

• 87.6% get recipes from Google search
• 51.1% from social media
• 47.2% from YouTube
• 50.8% from recipe websites
• 34.7% from cookbooks

No single site naturally becomes the user's permanent hub. However, most saving features are built as if it should. When a recipe is saved locally on one website, the user must remember to return specifically to that site later. In practice, they instead create portable versions of the recipe:

• 56.7% screenshot recipes
• 22.6% copy into notes
• 22.3% search for the same recipe elsewhere
• 23.9% switch to video versions
• 16.6% print

Overall, 80.5% report sometimes or always reconstructing recipes outside the page. We tested a shared toolkit allowing readers to save, organize, and cook recipes across participating sites instead of exporting them.

Results:

• 69.8% likely to use it
• 67.8% more likely to return to sites offering it
• 68.4% overall return lift
• 57% would use built-in assistance while cooking

When continuity exists, future cooking sessions begin inside the recipe ecosystem instead of in screenshots, notes apps, or AI tools.

Interpretation

Across the dataset, the same behavioral sequence appears repeatedly:

1. Users discover recipes on websites
2. They attempt to cook from the page
3. Friction pushes execution elsewhere
4. When execution becomes easy, they return

The competition is not search engines or AI tools.

The competition is the moment cooking begins.

Recipe websites currently win discovery. They frequently lose execution.

Conclusion

Home cooks have not abandoned recipe websites. They have adapted around them.

The majority still prefer to cook on the original page - but only when the page behaves like a cooking tool rather than an article.

Improving the execution experience does not just improve usability. It measurably increases cooking frequency, return visits, and long-term loyalty.

For publishers, this represents a clear opportunity:

Whoever owns the cooking session owns the relationship.

Study Details