ACCURMENT RESEARCH
Evaluations of Accurment AI
This research was conducted by a group of marketing scholars from Australian universities, including Curtin University, RMIT University, the University of Auckland, and the University of Queensland.
This research evaluated whether strategies generated by Accurment AI are perceived as more credible and more likely to be adopted than strategies produced by a generic large language model (LLM). Two experimental studies were conducted to test this, each focusing on different decision-making contexts and risk levels.
Study 1: Business Owners in Australia
The first study involved 270 Australian business owners (53% female, average age 49 years) recruited through a professional market research panel.
Participants were randomly assigned to evaluate one of two marketing plans for a fictitious insurance company. One plan was generated by a baseline AI model (GPT-4o), and the other by Accurment AI. Both plans were created using identical prompts to ensure consistency.
After reviewing their assigned plan, participants rated:
Results showed that participants rated Accurment’s plan as 9.4% more expert and 8.6% more likely to be adopted. This difference was statistically reliable, meaning it was very unlikely to have occurred by chance. A mediation analysis confirmed that perceived expertise explained the effect — Accurment AI’s credibility led directly to higher willingness to adopt its recommendations.
Study 2: Marketing and Sales Professionals in the US
To extend the findings, a second study was conducted with 404 marketing and sales professionals in the United States (53% female, average age 35 years) recruited through an online research panel.
This study introduced risk as a new factor. Participants evaluated marketing plans for a chocolate brand, either from the baseline AI or Accurment, under one of two conditions:
Measures mirrored those from Study 1, assessing perceived expertise and likelihood to adopt. The findings show that under low-risk conditions, the statistical difference between the two AI systems did not emerge. However and replicating the results in Study 1: Under high-risk conditions, participants rated Accurment’s plan as 9.9% more expert and 7.6% more likely to be adopted. This pattern suggests that when the consequences of a wrong decision matter, they turn to the source they perceive as more expert, credible, and reliable — in this case, Accurment.
Key Insight
Findings across both studies provide strong empirical proof that Accurment earns greater trust from decision-makers. Business owners and marketing professionals perceive Accurment AI’s strategies as more expert and more likely to be adopted than those from a generic AI model.
Importantly, this trust advantage is amplified under high-risk decision contexts, highlighting that Accurment’s value lies not only in producing sound strategies but in earning confidence when judgment and consequences matter most.
This research evaluated whether strategies generated by Accurment AI are perceived as more credible and more likely to be adopted than strategies produced by a generic large language model (LLM). Two experimental studies were conducted to test this, each focusing on different decision-making contexts and risk levels.
Study 1: Business Owners in Australia
The first study involved 270 Australian business owners (53% female, average age 49 years) recruited through a professional market research panel.
Participants were randomly assigned to evaluate one of two marketing plans for a fictitious insurance company. One plan was generated by a baseline AI model (GPT-4o), and the other by Accurment AI. Both plans were created using identical prompts to ensure consistency.
After reviewing their assigned plan, participants rated:
- Perceived expertise of the AI (e.g., “expert”, “knowledgeable”, “experienced”),
- Likelihood of adopting the recommendation.
Results showed that participants rated Accurment’s plan as 9.4% more expert and 8.6% more likely to be adopted. This difference was statistically reliable, meaning it was very unlikely to have occurred by chance. A mediation analysis confirmed that perceived expertise explained the effect — Accurment AI’s credibility led directly to higher willingness to adopt its recommendations.
Study 2: Marketing and Sales Professionals in the US
To extend the findings, a second study was conducted with 404 marketing and sales professionals in the United States (53% female, average age 35 years) recruited through an online research panel.
This study introduced risk as a new factor. Participants evaluated marketing plans for a chocolate brand, either from the baseline AI or Accurment, under one of two conditions:
- Low-risk decisions, where outcomes carried little consequence.
- High-risk decisions, where choosing the wrong strategy could harm brand reputation or cause financial loss.
Measures mirrored those from Study 1, assessing perceived expertise and likelihood to adopt. The findings show that under low-risk conditions, the statistical difference between the two AI systems did not emerge. However and replicating the results in Study 1: Under high-risk conditions, participants rated Accurment’s plan as 9.9% more expert and 7.6% more likely to be adopted. This pattern suggests that when the consequences of a wrong decision matter, they turn to the source they perceive as more expert, credible, and reliable — in this case, Accurment.
Key Insight
Findings across both studies provide strong empirical proof that Accurment earns greater trust from decision-makers. Business owners and marketing professionals perceive Accurment AI’s strategies as more expert and more likely to be adopted than those from a generic AI model.
Importantly, this trust advantage is amplified under high-risk decision contexts, highlighting that Accurment’s value lies not only in producing sound strategies but in earning confidence when judgment and consequences matter most.