ANALYSIS OF BUSINESS PROCESS MODELING: A CASE OF PRIVATE TARGETED ADVERTISING AGENCY

АНАЛИЗ МОДЕЛИРОВАНИЯ БИЗНЕС-ПРОЦЕССОВ: НА ПРИМЕРЕ ЧАСТНОГО АГЕНТСТВА ЦЕЛЕВОЙ РЕКЛАМЫ

Оспанов Ислам Муратулы

магистрант по направлению «ИТ-менеджмент», магистрант, Казахстанско-британский технический университет,

Казахстан, Алматы

Д-р Авинаш Бигадеванахалли Мрутунджая

ассоциированный профессор школы информационных технологий и инженерии, кандидат наук, Казахстанско-британский технический университет,

Казахстан, Алматы

ABSTRACT

This research develops a hybrid Human-AI Business Process Modeling (BPM) framework for digital advertising agencies to optimize campaign management while maintaining regulatory compliance. A comparative analysis evaluated three BPM models (manual, AI-driven, and hybrid) using BPMN tools and real-world campaign data across performance metrics including targeting accuracy, efficiency, and compliance risk. The hybrid model achieved 96% targeting accuracy (vs. 65% manual, 92% AI-only), reduced manual workload by 30-50%, and maintained low compliance risk with optimal 15% human intervention. The Human-AI hybrid approach provides the most balanced solution, combining AI efficiency with strategic human oversight for scalable advertising process optimization.

АННОТАЦИЯ

Это исследование разрабатывает гибридную человеко-ИИ модель бизнес-процессов (BPM) для агентств цифровой рекламы с целью оптимизации управления кампаниями при сохранении соответствия нормативным требованиям.  Сравнительный анализ оценил три модели BPM (ручную, управляемую ИИ и гибридную) с использованием инструментов BPMN и данных реальных кампаний по таким показателям эффективности, как точность таргетинга, эффективность и риск соответствия.  Гибридная модель достигла 96% точности таргетинга (по сравнению с 65% для ручной и 92% только для ИИ), сократила ручную рабочую нагрузку на 30-50%, и поддерживала низкий риск соответствия при оптимальном 15% участии человека.  Гибридный подход "человек-ИИ" обеспечивает наиболее сбалансированное решение, сочетая эффективность ИИ со стратегическим человеческим контролем для масштабируемой оптимизации рекламных процессов.

Keywords: Business Process Modeling; Human-AI Collaboration; Digital Advertising; Process Automation; Regulatory Compliance

Ключевые слова: Моделирование бизнес-процессов; Сотрудничество человека и ИИ; Цифровая реклама; Автоматизация процессов; Соответствие нормативным требованиям.

INTRODUCTION

The digital advertising landscape is experiencing unprecedented transformation driven by artificial intelligence advancement and evolving regulatory requirements [3, c. 9]. Traditional manual business process modeling approaches struggle to meet modern demands for speed, accuracy, and compliance, while fully automated systems often lack the flexibility required for complex advertising decisions [1, c. 9].

Current industry practices fall into two problematic extremes: labor-intensive manual processes that limit scalability, or fully automated systems that sacrifice strategic oversight for efficiency. This dichotomy creates operational challenges including inefficient targeting, compliance risks, and difficulty measuring return on investment.

This research addresses these challenges by developing a hybrid BPM framework specifically designed for digital advertising contexts. The solution leverages AI for efficiency while preserving human judgment in critical areas such as strategic decisions, creative development, and compliance assurance.

1 Literature Review and Research Gaps

1.  Current State of BPM in Digital Advertising

Existing research on business process modeling in advertising contexts reveals several critical limitations [7, c. 9]. Studies focusing on cognitive load in BPM and digital tool integration often lack practical applications specific to advertising agencies. Privacy management strategies in digital advertising emphasize data protection importance but provide limited specific frameworks for managing privacy concerns within targeted advertising workflows [4, c. 9].

1.2 Identified Research Gaps

Our analysis reveals five primary gaps in current literature:

1. Limited Hybrid Model Research: Most studies examine either fully manual or completely automated systems, with insufficient exploration of practical hybrid models combining human expertise with AI capabilities [7, c. 9].
2. Lack of Performance Metrics: Current research inadequately quantifies the value added by human intervention in areas like creative decision-making and compliance verification.
3. Regulatory Compliance Challenges: Existing frameworks lack comprehensive solutions for maintaining regulatory compliance in AI-driven advertising systems without sacrificing efficiency [5, c. 9].
4. Scalability Concerns: Limited evidence exists regarding hybrid model performance when handling high-volume, high-velocity data typical of modern digital advertising campaigns [2, c. 9].
5. Implementation Practicalities: Insufficient research addresses organizational change aspects of transitioning to hybrid BPM models, including workforce reskilling and change management strategies.

2. Methodology

2.1 Research Framework

Our study employed a comparative mixed-methods approach to evaluate three distinct BPM models Manual BPM (traditional human-driven workflows), AI-driven BPM (fully automated processes), Human-AI Hybrid BPM (balanced approach).

2.2 Tools and Technologies

The research utilized Business Process Model and Notation (BPMN) diagramming tools for workflow visualization, data analytics platforms for campaign performance measurement, and privacy management software for regulatory compliance assessment [6, c. 9].

2.3 Performance Evaluation Criteria

Key performance indicators included process efficiency and decision speed, targeting accuracy and campaign effectiveness, error reduction and compliance risk management, cost optimization and resource allocation, worker satisfaction and adaptation challenges.

3 Framework Implementation

3.1 Manual BPM Model Characteristics

The traditional manual approach involves human-controlled campaign setup, audience selection, budget allocation, compliance verification, and performance monitoring. Key challenges include slow decision-making processes, high dependency on human expertise, and increased compliance violation risks due to human error.

3.2 AI-Driven BPM Model Features

The fully automated model replaces human intervention with machine learning algorithms for campaign optimization, budget allocation, and compliance verification. While improving efficiency, this approach presents risks, including a lack of human oversight, potential decision-making biases, ethical concerns in automated content generation, and challenges in adapting to rapidly changing regulations [4, c. 9].

3.3 Human-AI Hybrid BPM Framework

Our proposed hybrid framework integrates AI automation with strategic human validation checkpoints. AI assists in ad targeting, budget optimization, and performance analysis, while human experts oversee compliance and strategic alignment. This approach combines AI processing speed with human judgment for critical decisions.

3.4 Framework Advantages

The hybrid model offers several key benefits, like AI accelerates repetitive processes while humans provide oversight. It reduces compliance risks by allowing human validation of AI decisions. The approach also increases adaptability in advertising strategies by combining AI insights with human expertise, ensuring both efficiency and regulatory compliance.

4 Results and Discussion

4.1 Performance Comparison Results

Empirical testing revealed significant performance differences across the three models. Manual BPM Performance, as decision speed is slow, targeting accuracy is 65%, compliance risk is high, human intervention is 100%, cost reduction is 5%, and worker satisfaction is low. AI-Driven BPM Performance, as decision speed is fast, targeting accuracy is 92%, compliance risk is medium, human intervention is 0%, cost reduction is 45%, and worker satisfaction is medium. Human-AI Hybrid BPM Performance, as decision speed is optimized, targeting accuracy is 96%, compliance risk is low, human intervention is 15%, cost reduction is 35%, and worker satisfaction is high.

4.2 Key Findings

The hybrid model demonstrates superior performance across most metrics, achieving the highest targeting accuracy (96%) while maintaining low compliance risk. The strategic 15% human intervention provides necessary oversight without significantly impacting efficiency gains.

4.3 Human-AI Collaboration Benefits

Research findings indicate several measurable benefits from human-AI collaboration. Enhanced Productivity, as task completion improved by 30% with AI-assisted workflows. Also error Reduction, as AI-driven content generation reduced manual input mistakes by 40%. Next is improved Decision-Making, as AI provided superior targeting suggestions, improving advertisement engagement rates. And worker Satisfaction, as employees reported reduced workload stress while maintaining focus on strategic tasks

4.4 Implementation Challenges

Despite advantages, several challenges require attention. AI Bias Management, as AI systems occasionally reinforced biases from training data, scalability Considerations, as AI performed well in targeted campaigns but required additional tuning for broader strategies. Over-Reliance Prevention, as employees sometimes trusted AI outputs without critical evaluation, highlighting human oversight necessity

5 Future Research Directions

Several areas merit further investigation. Development of hybrid AI decision models with enhanced human validation loops, industry-specific AI tuning for advertising processes, automated workflow optimization systems that adapt based on real-time campaign performance and integration of reinforcement learning for improved AI adaptability.

CONCLUSION

This research demonstrates that a Human-AI Hybrid BPM model presents the most balanced approach for digital advertising agencies, ensuring efficiency, compliance, and adaptability while maintaining worker satisfaction. The framework addresses critical gaps in existing literature by providing practical solutions for integrating AI automation with necessary human oversight.

Key contributions include empirical evidence supporting hybrid model superiority over purely manual or automated approaches, practical framework for balancing automation efficiency with regulatory compliance, measurable performance improvements in targeting accuracy and operational efficiency and guidelines for managing human-AI collaboration in advertising workflows.

The findings underscore the necessity of adapting BPM strategies to specific advertising industry needs, ensuring both technological advancement and operational effectiveness. Future implementations should focus on industry-specific customization and continuous optimization of human-AI collaboration mechanisms.

Organizations considering BPM transformation should prioritize hybrid approaches that leverage AI capabilities while preserving essential human judgment in strategic decision-making, compliance verification, and creative processes. This balanced approach offers sustainable solutions for navigating the evolving digital advertising landscape.

Список литературы:

1. Anderson P., and Brown R., An experimental investigation of BPMN-based corporate communications modeling // Business Process Management Journal – 2022. Vol. 28 № 4 - P. 701–719.
2. Cavallo A., Cosenz F., and Noto G, Business model scaling and growth hacking in digital entrepreneurship // Journal of Small Business Management - 2 – 2023. Vol 62 № 4 - P. 2058–2085.
3. Coffin J., Asking Questions of AI Advertising: A Maieutic Approach // Journal of Advertising – 2022. Vol. 51 № 5 - P. 608–623.
4. Harris P., and Lewis M., Advertising and privacy: an overview of past research and a research agenda // Journal of Marketing Communications – 2022. Vol 28 № 4 - P. 345–362.
5. Simone A., Silvia B., Davide P., Data-driven digital advertising: benefits and risks of online behavioral advertising International // Journal of Retail and Distribution Management – 2020. Vol. 48 № 11 - P. 1205–1223.
6. Lopes T. and Guerreiro S., Assessing business process models: a literature review on techniques for BPMN testing and formal verification // Business Process Management Journal – 2023. Vol. 29 № 8 - P. 133-162.
7. Uskenbayeva R.K., Moldagulova A.N., Satybaldiyeva R.Zh., Bektemyssova G.U., Kalpeeva Zh.B. Methodology for modeling hybrid administrative business processes // Herald of the Kazakh-British technical university – 2019. Vol. 16 № 3 - P. 318-323.