[In-house Ad Operations Survey] Only 5.6% of Companies Effectively Implement Ad PDCA. Factors Hindering Improvement and the Reality of AI Utilization Revealed

Irgurum Inc. (Headquarters: Kita-ku, Osaka, Representative Director: Susumu Iwata) conducted an internet survey targeting 516 executives and on-site personnel from companies managing advertising in-house, focusing on the reality of PDCA in ad operations and AI utilization.

This survey revealed that while many companies are working to improve their advertising, the improvement cycle is not functioning sufficiently in practice. Furthermore, a gap still exists between the expectations for AI and its actual utilization. These results suggest that common structural issues underlie the stagnation of PDCA and the slow progress of AI utilization.

The detailed survey results can be downloaded from the link below:

https://go.cm.ebis.ne.jp/insights/wp/inhouse_pdca/

**Survey Background**

In recent years, parallel operation of multiple objectives and multiple channels has become common in business companies' advertising operations, making the operating environment complex. This survey confirmed that companies, on average, have 2.2 types of objectives and operate 2.3 channels in parallel. Furthermore, tasks that consume the most time, such as creative production/improvement, ad placement/setting, and data aggregation/report creation, collectively account for 64.7% of operations, indicating a situation where it is difficult to secure sufficient time for analysis and strategy formulation. On the other hand, there is growing expectation for AI utilization in areas such as factor analysis and improvement proposals in ad operations. However, on-site, the improvement cycle itself is often difficult to implement due to tasks being reliant on specific individuals and fragmented data. Against this background, this survey aimed to clarify the reality of the improvement cycle in in-house ad operations and the current status and challenges of AI utilization.

**Survey Results**

**Only 5.6% of companies have an ad operations improvement cycle that is "fully functioning."**

Only 5.6% of companies responded that their ad operations improvement cycle (Plan → Do → Check → Act) was "fully functioning." In contrast, a total of 51.6% of companies responded that it was "not functioning well" or "hardly functioning at all," revealing that the ad operations improvement cycle is not adequately effective.

**Structural issues related to skills, time, and data hinder improvement.**

The top reasons why the ad operations improvement cycle is not functioning sufficiently were: "limited personnel capable of analysis" (23.4%), "no time for analysis" (21.9%), and "data is scattered and cannot be aggregated" (17.8%). This suggests that structural issues exist in terms of skills, time, and data, hindering improvement.

**A gap is observed between the expected role of AI and its actual areas of utilization.**

The top expected roles for AI in ad operations were "automatic analysis of performance data" (54.7%) and "insight generation" (50.0%). However, the actual tasks where AI is utilized are mainly "ad copy/creative creation" (31.8%), "report/presentation material creation" (31.0%), and "data aggregation/organization" (30.6%). This indicates a gap between the expected decision support areas and the current areas of utilization.

**Insights from Survey Results**

This survey highlights that while many companies have a high intention to engage in PDCA for ad operations, the reality is that the complexity of the operating environment is hindering improvement. Specifically, the top reason for lack of improvement being "insufficient personnel for analysis (23.4%)" reveals a "personalization structure" where advanced judgment tasks are concentrated on specific individuals, leading to resource strain and overall cycle stagnation. Furthermore, the fact that expectations for AI utilization are focused on "decision support" while current utilization remains in "operational support" suggests a lack of foundational infrastructure such as data preparation and process organization. For the continuous growth of in-house operations, it is urgent to build a system that can complete analysis to improvement as an organization, rather than relying on individual skills.

**Comment from Susumu Iwata, Representative Director, Irgurum Inc.**

"The results of this survey reflect the urgent challenge faced by the field: 'it cannot be managed by human power alone.' With the generalization of multi-channel, multi-objective operations and the current limit of operational workload, creating an 'environment' that seamlessly connects data comprehension, problem organization, factor analysis, and measure consideration, rather than ad-hoc improvements, is the key to success for in-house operations. Our company will support corporate in-house advertising operations by resolving these structural issues and developing an improvement environment that maximizes AI utilization."

**About the Survey Report**

This survey report provides insights into the in-house ad operations field.