TURKISH-LANGUAGE RESEARCH / RESEARCH ARTICLE

Are more comments on an AI advertisement good news?

Many comments on an advertisement do not necessarily mean people like it. This Turkish-language study examines YouTube comments on an AI-generated Coca-Cola advertisement and reports predominantly negative automated classifications. DijitalPi recommends reading the reasons behind engagement instead of treating its volume as success. Commenters do not represent every customer, and negative comments alone do not demonstrate falling sales.

ACADEMIC WORK REVIEWED

Yapay Zekâ ile Üretilen Reklamlara Yönelik Tüketici Tepkileri: Coca-Cola “Holidays Are Coming” Reklamının Duygu Analizi

ResearchersNevra Üçler, Defne Ertaş.

Kronotop İletişim Dergisi · 29 July 2026 · 3(2):407–433 · Source language: Turkish

Open the original publication Find on Google Scholar

ENGLISH REVIEW AND COMMENTARY: DIJITALPI

We explain the paper in plain English, discuss what it may mean for organisations and identify our own examples separately.

Read the explanation Go to 3 business examples

English review: 22 September 2026 · The source date appears in the citation above.

Editorial review draft

The source has been checked; the review still carries its original editorial status. DijitalPi did not conduct a customer experiment for this paper.

LET'S READ THE RESEARCH TOGETHER

We first explain the researchers' question, method and findings. We then discuss how to interpret the results, clearly separating DijitalPi's commentary from the source.

01 / WHAT DID THE RESEARCHERS WANT TO UNDERSTAND?

Do comment volume and tone tell the same story?

The study examines viewer comments on one AI advertisement.

Explaining the context · DijitalPi commentary

DijitalPi explanation: A comment is participation, not the same measure as satisfaction, purchasing or trust. State which of these your report can observe.

02 / HOW WAS THE RESEARCH CONDUCTED?

How did they test the question?

A RoBERTa-based model classifies 9,051 YouTube comments as positive, neutral or negative.

Understanding the method · DijitalPi commentary

DijitalPi explanation: Sentiment analysis labels the tone of text. Accuracy depends on the model and comment. People who remain silent are missing from the data.

03 / RESEARCH FINDINGS

Negative labels predominate.

The analysis classifies 58.4% of comments as negative.

Source: Original Turkish publication and source record ↗

04 / DIJITALPI'S EXPLANATION

How should we interpret these findings?

DijitalPi commentary: Add an explanation of why people commented alongside engagement totals.

Distinguish product problems from criticism of the film. They may require different work from the support and creative teams.

05 / CONCLUSION AND OPEN QUESTIONS

What did we learn, and what do we still not know?

Comment distribution is not the view of all consumers or a sales effect.

Our question: which reactions address the product and which address its presentation? Publication in Turkish does not mean the comments represent Türkiye.

Return to the researchers' original publication

Original Turkish publication and source record

DIJITALPI'S APPLICATION COMMENTARY

How can we use this in marketing?

We created these scenarios to make the topic concrete. They are not cases from the paper or measured client results.

EXAMPLE 01

Launch: a surge in comments

A new film receives many comments and the team wants to increase spending.

Do comments reflect interest or objections?

Open the recommendation for example 01

Read the reasons first.

Sample comments across different times. Separate reactions to the product, imagery, brand and AI use. Document selection and avoid presenting a few striking comments as the whole audience’s view.

EXAMPLE 02

Reporting: an automated sentiment score

A tool labels comments positive or negative.

Did it understand sarcasm?

Open the recommendation for example 02

Check a sample manually.

Inspect examples across positive, negative and uncertain labels. Record classification errors. Separate product complaints from criticism of the advertisement; the score is not an exact customer-satisfaction rate.

EXAMPLE 03

Brand team: inferring sales from comments

More negative comments are taken as evidence of lower sales.

Do order records support that claim?

Open the recommendation for example 03

Check behaviour separately.

Track orders, cancellations and support requests using separate definitions. Record price and stock changes during the period. Do not treat comment changes alone as the cause of a sales trend.

TRY IT WITH YOUR TEAM

Add reasons to the engagement report.

Separate comment volume, recurring opinions, manual checks and observable customer behaviour. This is a proposed DijitalPi worksheet, not a replication or proof of campaign impact.

Explore AI solutions

Source and ownership

Nevra Üçler, Defne Ertaş.
İstanbul Gelişim Üniversitesi; Giresun Üniversitesi.

Kronotop İletişim Dergisi · 29 July 2026 · 3(2):407–433. Open the original publication · Find on Google Scholar

The academic work belongs to the researchers named above. This page contains DijitalPi's explanatory review and original business examples; it is not a full translation of the paper.

← Return to all research articles