Video Ad Analysis Pro (DPro) Succeeds in Complete Automation of SNS Ad Operations via AI Agent Integration──Achieving Single-Day ROAS of 1400%
KASHIKA successfully automated SNS ad operations 24/7 by linking its video ad analysis tool DPro with an AI agent. It achieved an astonishing single-day ROAS of 1400% by eliminating operational downtime and continuously optimizing creatives and bidding.
📋 Article Processing Timeline
- 📰 Published: April 24, 2026 at 23:07
- 🔍 Collected: April 24, 2026 at 14:31
- 🤖 AI Analyzed: April 24, 2026 at 23:37 (9h 5m after Collected)
KASHIKA Inc. (Headquarters location and representative name omitted) announces that it has succeeded in fully automating SNS ad operations by utilizing its AI video ad analysis tool 'Video Ad Analysis Pro (DPro)' ( https://kashika-20mile.com/dpro/ ). In a demonstration experiment during internal operations, it achieved a high result of 1400% for single-day ROAS (Return on Ad Spend).
■ Background of the Initiative
The digital advertising market continues to expand year by year, and SNS ads, in particular, occupy an important position in corporate marketing strategies. However, operating SNS ads presents unique difficulties.
On SNS platforms, user interests and trends change in real time. It is not uncommon for ad creatives that garnered high engagement until yesterday to experience a significant drop in response rates today. To cope with such fluctuations, it is necessary to continuously rotate diverse operations at high speed, including constant monitoring of performance data, detecting trend changes, rapid replacement of creatives, and adjustment of bidding strategies.
In traditional operational structures, the majority of these tasks relied on human resources. The workflow involves ad operation personnel periodically checking dashboards, detecting performance changes, and manually replacing creatives or adjusting delivery settings. Generally, there is a time lag of several hours to half a day from data confirmation to execution, during which SNS trends continue to shift.
Furthermore, when sudden trend changes occur during late nights or holidays, a challenge was that the response would be delayed until the next business day. Even though SNS user behavior occurs 24 hours a day, 365 days a year, ad operations tend to be restricted to business hours, and this 'operational downtime' was a major factor in missed opportunities.
■ Mechanism of Complete Automation
To resolve this issue, KASHIKA constructed a fully automated operation system that forms a trinity by linking DPro's API, an AI agent, and the ad platform's API.
This system operates largely in three phases.
Phase 1: 'Analysis'
DPro's API collects and analyzes real-time data such as SNS ad performance, user engagement changes, competitor ad movements, and overall platform trend shifts. Utilizing DPro's strength in multifaceted video ad analysis technology, it quantitatively grasps which elements of a creative contribute to results.
Phase 2: 'Judgment and Optimization'
The AI agent receives the analysis results from DPro and automatically judges whether the current ad performance is optimal. If it detects signs of declining performance, the AI agent automatically generates proposals for creative replacement, targeting adjustments, and bid strategy changes.
Phase 3: 'Execution'
Optimization measures formulated by the AI agent are instantly reflected in the delivery settings via the ad platform's API. Creative replacements, target audience adjustments, and budget allocation changes are executed in real time without human intervention.
By continuously cycling through these three phases, immediate ad operations linked to SNS movements are realized on a 24-hour basis.
[*Please insert a screenshot of the 'DPro analysis dashboard screen' here. A graph of ROAS trends, a log of the AI agent's automatic adjustments, or a conceptual diagram of the system linkage is ideal. If possible, post a graph showing improved performance before and after automatic optimization.]
■ Achieved Results
As a result of fully operating this system in internal SNS ad operations, a single-day ROAS of 1400% was achieved. This means that 14 times the amount spent on advertising was generated in sales in just one day. This is a level that was difficult to reach with traditional manual operations, and it can be said that real-time optimization by the AI agent accurately captured the wave of trends.
In addition, we succeeded in practically reducing human-hours spent on ad operations to zero. Because the monitoring and adjustment tasks that ad operation personnel previously spent several hours a day on have been completely automated, personnel can now focus on upstream strategy formulation and creative planning.
Furthermore, through constant 24/7/365 optimization, we can now instantly respond to trend fluctuations occurring during late nights and holidays. It is now possible to reliably capture opportunities during the 'operational downtime' that had previously been missed.
■ Future Outlook
At KASHIKA
■ Background of the Initiative
The digital advertising market continues to expand year by year, and SNS ads, in particular, occupy an important position in corporate marketing strategies. However, operating SNS ads presents unique difficulties.
On SNS platforms, user interests and trends change in real time. It is not uncommon for ad creatives that garnered high engagement until yesterday to experience a significant drop in response rates today. To cope with such fluctuations, it is necessary to continuously rotate diverse operations at high speed, including constant monitoring of performance data, detecting trend changes, rapid replacement of creatives, and adjustment of bidding strategies.
In traditional operational structures, the majority of these tasks relied on human resources. The workflow involves ad operation personnel periodically checking dashboards, detecting performance changes, and manually replacing creatives or adjusting delivery settings. Generally, there is a time lag of several hours to half a day from data confirmation to execution, during which SNS trends continue to shift.
Furthermore, when sudden trend changes occur during late nights or holidays, a challenge was that the response would be delayed until the next business day. Even though SNS user behavior occurs 24 hours a day, 365 days a year, ad operations tend to be restricted to business hours, and this 'operational downtime' was a major factor in missed opportunities.
■ Mechanism of Complete Automation
To resolve this issue, KASHIKA constructed a fully automated operation system that forms a trinity by linking DPro's API, an AI agent, and the ad platform's API.
This system operates largely in three phases.
Phase 1: 'Analysis'
DPro's API collects and analyzes real-time data such as SNS ad performance, user engagement changes, competitor ad movements, and overall platform trend shifts. Utilizing DPro's strength in multifaceted video ad analysis technology, it quantitatively grasps which elements of a creative contribute to results.
Phase 2: 'Judgment and Optimization'
The AI agent receives the analysis results from DPro and automatically judges whether the current ad performance is optimal. If it detects signs of declining performance, the AI agent automatically generates proposals for creative replacement, targeting adjustments, and bid strategy changes.
Phase 3: 'Execution'
Optimization measures formulated by the AI agent are instantly reflected in the delivery settings via the ad platform's API. Creative replacements, target audience adjustments, and budget allocation changes are executed in real time without human intervention.
By continuously cycling through these three phases, immediate ad operations linked to SNS movements are realized on a 24-hour basis.
[*Please insert a screenshot of the 'DPro analysis dashboard screen' here. A graph of ROAS trends, a log of the AI agent's automatic adjustments, or a conceptual diagram of the system linkage is ideal. If possible, post a graph showing improved performance before and after automatic optimization.]
■ Achieved Results
As a result of fully operating this system in internal SNS ad operations, a single-day ROAS of 1400% was achieved. This means that 14 times the amount spent on advertising was generated in sales in just one day. This is a level that was difficult to reach with traditional manual operations, and it can be said that real-time optimization by the AI agent accurately captured the wave of trends.
In addition, we succeeded in practically reducing human-hours spent on ad operations to zero. Because the monitoring and adjustment tasks that ad operation personnel previously spent several hours a day on have been completely automated, personnel can now focus on upstream strategy formulation and creative planning.
Furthermore, through constant 24/7/365 optimization, we can now instantly respond to trend fluctuations occurring during late nights and holidays. It is now possible to reliably capture opportunities during the 'operational downtime' that had previously been missed.
■ Future Outlook
At KASHIKA