DEEPER STUDY / Hellmann’s / 2014
WhatsCook
Live assistance connects a product with the user’s immediate cooking problem.
AI-assisted research and analysis for the SNS Creative Fieldbook
Sources reviewed 6 November 2026
The original campaign creators did not write or endorse this analysis.
Original campaign credits: Agency: Cubocc.
Recognition: No award claim is made in this study.
Launch / edition: 2014. Award years: No award claim.
The problem. The idea. The experience.
Original brief not established in these sources. Editorial reading: Live assistance connects a product with the user’s immediate cooking problem.
How the idea came to life
Hellmann’s launched WhatsCook in Brazil in May 2014, offering cooking help through WhatsApp after users registered their mobile number on a website. Contemporary reports describe a team of chefs answering questions and sending personalised recipes, photographs, recordings and short videos. Users could share a photograph of refrigerator contents to prompt a recipe suggestion. The service was a time-limited activity, with the launch reporting giving an early June end date.
How it reached people
Website registration followed by chef-supported WhatsApp messages, photographs, recordings and short videos.
Supporting sources: trade press → trade press →
Read the data in context.
The numbers below describe different kinds of outcomes. Their date, scope and reporting source are part of the result.
Execution and context verified; no isolated commercial or behavioural effect established.
Read the campaign story
Hellmann’s launched WhatsCook in Brazil in May 2014, offering cooking help through WhatsApp after users registered their mobile number on a website. Contemporary reports describe a team of chefs answering questions and sending personalised recipes, photographs, recordings and short videos. Users could share a photograph of refrigerator contents to prompt a recipe suggestion. The service was a time-limited activity, with the launch reporting giving an early June end date.
The editorial mechanism was assistance at the moment of use. Instead of asking people to admire a finished dish and later remember the brand, the service entered the practical gap between available ingredients and a meal. The brand’s role was connected to showing different culinary uses for mayonnaise. A conversation could respond to the user’s situation in a way that a static recipe page could not, at least within the service’s stated scope.
This was a human-supported service, not an artificial-intelligence chatbot, and the distinction matters. The experience depended on staffing, expertise and response capacity. The launch reports explicitly discuss simultaneous-user limits. That operational detail is part of the creative idea: promising individual attention changes what the team must deliver. A successful invitation without enough support could produce waiting, confusion or disappointment rather than useful brand contact.
The case is distinct from Hellmann’s Recipe Receipt campaign. Both connect food products with meal ideas, but the interaction here was a live messaging exchange, not a recipe generated on a retail receipt. The shared category should not erase the different customer moment or delivery system. One begins around a purchase record; this one begins with a person asking for help while planning or preparing food.
The sources establish the launch, channels and service design, not an independently isolated sales lift or a permanent customer-support offer. They also do not justify treating private kitchen photographs as unrestricted marketing assets. Any contemporary version would need clear consent, limited data retention and appropriate handling of food-safety or allergy questions. The useful lesson is to define a manageable assistance promise and make the real service capacity as carefully designed as the advertisement inviting people to use it.
Choose the moment
Help is more relevant when it meets a current task.
Budget the service
Personalised communication needs staffing and response limits.
Respect private context
Images and messages shared for help are not automatically promotional content.
A small test for another brief
Offer a short, clearly scoped product-help session.
Measure resolved questions and waiting time.
Stop if capacity fails or questions require specialist advice.
- Choose a low-risk question staff can answer accurately.
- Set visible hours, capacity and consent terms.
- Run a small opt-in session and record only non-sensitive service issues.
Sources & further reading.
These are educational analyses of external campaigns, not SNS client projects. Source-backed campaign details, editorial interpretation and proposed tests are distinguished. Brand and campaign-entry reports are not independent causal audits.
- Meio & Mensagem: Hellmann’s WhatsApp recipes →trade press · Supports: Cubocc credit, chef-supported service and time-limited operation.
- Panrotas: WhatsCook launch →trade press · Supports: May 2014 registration, media formats and refrigerator-photo interaction.