DIGIBAR AND DIGIBY - SURVEY RESULTS, ANALYSIS AND CONCLUSIONS

After designing websites for both Digiby and DigibarI uploaded content produced by ChatGPT, the AI Content Generator, and Grace, the Human Copywriter to each website. After, I created an online survey. 30 people amongst whom are fellow Cohorts, Lecturers, Design Industry Players, and friends took part in the survey. The main aim of the survey was to determine if text generated by Generative AI is as good or better than copy written by Human Copywriters. In the survey, I asked respondents to 1. Rate the quality of text content on Digibar & Digiby Website using the star rating scale (1 star being poor, 5 stars being excellent) and 2. Which website they think had text generated by ChatGPT (Generative AI). Find the results of the survey below.

SURVEY RESULTS - 1. QUALITY OF TEXT CONTENT

SURVEY RESULTS - 2. WEBSITE RESPONDENTS THINK WAS GENERATED BY CHATGPT

GOOGLE ANALYTICS RESULTS

Find Google Analytics of both Digiby and Digibar below.

DIGIBAR RESULTS

DIGIBY RESULTS

GRAMMARLY SUGGESTIONS FOR DIGIBAR

GRAMMARLY SUGGESTIONS FOR DIGIBY

TIME AND COST ANALYSIS

Generative AI : 2 Hours
Human Content Writer : 80 Hours
Generative AI : 15$
Human Content Writer : 1000$

For Digibar, which features content created by human copywriters, the results revealed that no respondents rated Digibar content as poor (1 star). This suggests that none of the participants found the human-generated content to be of the lowest quality. Only 2 participants (6.67%) gave a 2-star rating. This indicates a minimal dissatisfaction with the content, but the majority of respondents found it to be better than poor. 6 respondents (20%) rated Digibar content with 3 stars. While this is a significant portion, it still suggests that the majority of participants had a positive perception of the human-generated content. The most common rating for Digibar content was 4 stars, chosen by 11 respondents (36.67%). This suggests a strong overall positive sentiment towards the quality of content produced by human copywriters. An additional 11 participants (36.67%) gave the highest rating of 5 stars, indicating a substantial number of respondents highly valued the quality of the content on the Digibar website. The majority of respondents (73.34%) rated Digibar content as either very good (4 stars) or excellent (5 stars), with a relatively small percentage expressing some level of dissatisfaction (2 or 3 stars). For Digiby, which features content generated by ChatGPT, (Generative AI), the results revealed that Similar to Digibar, no respondents rated Digiby content as poor (1 star). It is fascinating, however, that there were no respondents who gave a 2-star rating to Digiby content as compared to 2 participants who gave  Digibar a 2-star rating. 3 participants (10%) rated Digiby content with 3 stars. 14 respondents (46.67%) gave Digiby content a 4-star rating, signifying a substantial positive sentiment towards the quality of the AI-generated content. An additional 13 participants (43.33%) rated Digiby content with 5 stars, indicating a strong overall positive perception. The majority of respondents (90%) rated Digiby content as either good (4 stars) or excellent (5 stars), with only a small percentage expressing some level of dissatisfaction (3 stars).The average rating for Digibar was 4.033 whereas the average rating for Digiby was 4.33. The results show that the respondents found the text on both websites to be of good quality. However, the Digiby website received slightly higher ratings, which suggests that the respondents found the text generated by Generative AI to be slightly better than the text written by human copywriters.For the second question in the survey, I asked which website respondents think was generated by ChatGPT. Out of the 30 respondents, 18 participants (60%) chose Digibar as text generated by ChatGPT. The result indicates a majority perception that Digibar, the website featuring content generated by human copywriters, is likely to be the source of the text generated by ChatGPT. Whereas 12 participants (40%) chose Digiby as text content generated by ChatGPT (Generative AI). The survey results revealed that a majority of participants (18 out of 30) incorrectly identified the human-generated content (Digibar) as being produced by Generative AI. Only 12 participants correctly identified Digiby as the website generated by ChatGPT. This finding suggests that individuals have difficulty distinguishing between human-written and Generative AI-generated text. Thus from the first question in the survey where respondents chose Digiby as the website with higher text quality most of the respondents were a bit biased in the second question in the survey by associating quality content with human creation, therefore incorrectly choosing Digibar as the website with content generated by ChatGPT. 

For Google Analytics, in the last 7 days, Digibar recorded 1 new user and 11 event counts. In Google Analytics, an event is a user interaction or occurrence on your website or app that you want to track. Events can be anything from clicking a button to watching a video to completing a purchase. By tracking events, you can gain valuable insights into user behavior and measure the success of specific actions on your site or app. The event count is the total number of times an event has been triggered during a specified time period. Digibar had 5 page views. Digiby, in contrast, exhibited a more robust performance over the same period. With 9 new users and a notable 45 event count. This showcases a higher level of user engagement compared to Digibar. The visitor distribution is diverse, with 7 from the United States, 1 from Canada, and 1 from Ghana. Digiby had 18 page views. Digiby outperforms Digibar in terms of new users and event count, indicating a potential advantage in user attraction and interaction through AI-generated content. The geographic distribution of Digiby’s visitors is also more diverse, spanning across North America and Africa. This suggests that AI-generated content may have a broader international appeal compared to human-generated content, which is crucial for websites aiming to reach a global audience. The findings of this analysis suggest that Generative AI-produced content may be an effective alternative to Human-Written content for websites. 

As part of my text generation review, I  checked the website design page content produced by both ChatGPT – the AI generator, and Grace the Human Content writer for grammatical errors. I used  Grammarly, a software that reviews spelling, grammar, punctuation, clarity, engagement, and delivery mistakes in English texts, detects plagiarism, and suggests replacements for the identified errors. For Digibar, Grammarly made 36 grammar suggestions. The higher number of suggestions suggests that there may be more grammatical errors or areas for improvement in human-generated content. For Digiby, Grammarly made 11 grammar suggestions. The relatively low number of suggestions indicates a commendable grammatical accuracy. The results suggest that Digiby demonstrates commendable grammatical accuracy, while Digibar, despite its human touch, may benefit from additional editing.

Website content including the Homepage, About Us Page, Main Service Page, Website Design Service Page, UI/UX Design Service Page, Copywriting Page, Search Engine Optimization Page, Contact Us Page, and 2 Blog Posts, generated by ChatGPT (Generative AI) exhibited a remarkable efficiency, requiring only 2 hours for completion. This rapid pace is a testament to AI systems’ automated and parallel processing capabilities. Generative AI operates by ingesting vast data and swiftly synthesizing coherent and contextually relevant content. On the contrary, the human-driven approach by Writer Grace demanded a significantly larger time investment of 10 Days. Now to remove any bias I asked Grace and she used 8 hours per day for 10 days, hence the  80 hours. This discrepancy is attributed to the inherently iterative and time-intensive nature of human creativity. Grace, like many skilled writers, engages in thoughtful ideation, research, and revisions to deliver a polished piece. The cost associated with ChatGPT which I used to produce content for the Digiby website was entirely free, but to remove any bias, I added the premium charge of 15$ per month in my analysis.  Generative AI is significantly lower than that of a human writer.  This cost-effectiveness is a result of the one-time investment in the AI system, with subsequent content generation incurring minimal additional expenses. The cost associated with human-driven content creation is notably higher. Grace’s services, valued at $1000, highlight the financial commitment required to employ a skilled human writer. This cost encompasses not only the writer’s time but also their expertise, creativity, and subjective interpretation of the given task. According to the survey results, the content generated by ChatGPT (Generative AI) is marginally superior to that produced by Grace, (Human Copywriter). ChatGPT’s ability to analyze patterns, understand context, and produce grammatically sound content contributes to its high-quality output. While the human touch in content creation often adds a unique flair and creativity, the survey results suggest that the content produced by Grace is slightly edged out by ChatGPT (Generative AI). This finding may be surprising given the common perception that human-generated content is inherently superior as can be seen in my second questionnaire when a majority of respondents incorrectly chose Digibar as the website with content produced by ChatGPT. However, it underscores the evolving capabilities of Generative AI in understanding and emulating human-like writing.

2 thoughts on “Survey Results & Google Analytics Results”

  1. Good research and interesting perspective! The role of AI cannot be overemphasized.
    Great research job done. Thanks for giving insight into these

  2. Incredible work. Obviously AI has a lot of work ahead for improvement but with the help of human direction

Comments are closed.