Two in three consult AI before shopping. 

Detected Manipulations & Suggested Changes

Omission of key information

Leaving out important contextual or methodological details that are necessary to fully evaluate the claims.

The article repeatedly cites percentages from a survey (e.g., "two out of three respondents have used AI applications", "15% using AI applications on a daily basis", "42% occasionally during the week") but does not provide basic methodological information such as sample size, sampling method, demographics, geographic scope, or margin of error. It also states: "These percentages do shift, especially since a large percentage of respondents, over 40%, who have not taken any of the aforementioned actions, intend to do so in the future" without clarifying how intention was measured or how strong that intention is.

3 suggestions

Add survey methodology details, for example: "The IELKA survey was conducted in [month, year] with a sample of [N] adults in [country/region], using [online/phone/in-person] interviews. The sample was/was not weighted to be representative by age, gender, and region. The margin of error is approximately ±X percentage points."

Clarify the population: specify whether respondents are all supermarket shoppers, general consumers, or a specific subgroup (e.g., online shoppers).

Qualify the statement about future intentions, e.g.: "Among those who have not yet used AI for these actions, 40% said they are ‘likely’ or ‘very likely’ to do so in the future, according to self-reported intentions in the survey."

Framing effect / Mild bias in framing

Presenting information in a way that subtly emphasizes one interpretation or trend over others, influencing perception without changing the underlying facts.

1) Opening framing: "Up until a few years ago, a friend or relative would be consulted when it came to shopping. The friend was then replaced by search engines and price comparison platforms. Now the new shopping mate is called Claude, ChatGPT or Gemini..." This frames AI as a natural, almost inevitable successor to friends and search engines, suggesting a linear replacement trend rather than coexistence. 2) "Although still contacting store staff regarding information about a product seems like the best option, AI’s acceptance rates are significant as a possible solution." The phrase "seems like the best option" is an evaluative statement that is not directly supported by data in the article and subtly positions store staff as superior while still promoting AI as a strong alternative. 3) "...while based on specific criteria, it comes first, as it is faster, easily accessible and more affordable." This is favorable framing of AI without specifying which criteria were measured, how they were ranked, or providing the comparative numbers.

3 suggestions

Neutralize the opening framing, e.g.: "In the past, many shoppers consulted friends or relatives. Later, search engines and price comparison platforms became common tools. Recently, some consumers have begun using AI applications such as Claude, ChatGPT, or Gemini for shopping-related tasks."

Replace evaluative language like "seems like the best option" with data-based wording, e.g.: "A majority of respondents still prefer contacting store staff for product information, but AI is also considered a viable option by a significant minority."

Clarify the criteria where AI "comes first" by adding specifics: "In the survey’s evaluation of information sources, nutritionists were most trusted overall. However, on specific criteria such as speed, accessibility, and cost, a higher percentage of respondents rated AI more favorably than other options (e.g., X% cited AI as fastest vs. Y% for store staff)."

Oversimplification

Reducing a complex situation or trend to a simple narrative that may gloss over important nuances.

The narrative that "The friend was then replaced by search engines and price comparison platforms. Now the new shopping mate is called Claude, ChatGPT or Gemini" suggests a straightforward replacement sequence (friends → search engines → AI). In reality, these sources often coexist, and the article does not provide data showing that friends or search engines have been "replaced" rather than supplemented.

2 suggestions

Rephrase to acknowledge coexistence of sources, e.g.: "While many consumers still consult friends, relatives, and search engines, AI tools such as Claude, ChatGPT, or Gemini are increasingly being added to the mix of resources used for shopping decisions."

If the article wants to claim replacement, add supporting data: for example, comparative figures on how often people now consult friends vs. AI vs. search engines, and trends over time.

Appeal to authority (mild, implicit)

Using the reputation of an institution or event to lend weight to claims without providing sufficient underlying evidence.

The article notes that the survey was "recently presented at the 16th Food Retail Conference" and conducted by "the Research Institute for Retail Consumer Goods (IELKA)." While naming the source is good practice, the way it is presented may implicitly encourage readers to accept the findings as authoritative without methodological detail. This is a mild form of appeal to authority, especially since no limitations or caveats are mentioned.

2 suggestions

Balance the reference to authority with transparency, e.g.: "According to a survey by the Research Institute for Retail Consumer Goods (IELKA), presented at the 16th Food Retail Conference, [results]. The survey’s findings should be interpreted in light of its methodology and sample, described below."

Add a brief note on limitations, such as potential sampling biases or that the data reflect self-reported behavior and intentions.