-- Increasing need for advanced analytics, including statistical models, AI, and machine learning --

XICA has conducted a "survey on measuring advertising effectiveness among corporate advertising and promotion personnel" as the second installment of its original research and study report in the offline advertising field (*1), and is now announcing the results.

About this survey

Recent advances in digital marketing technology have led to a dramatic increase in the number of internet advertising (*2) and offline advertising media. As the number of media options for advertising has increased and advertising strategies have become more complex, "how to select the optimal media to invest advertising budgets in" has become the most important issue in companies' advertising and promotion activities. For this reason, in recent years, data utilization and analysis methods for determining optimal allocation of advertising budgets that transcend the distinction between internet advertising and offline advertising have been attracting attention.

However, a survey conducted by Yano Research Institute in 2016 (*3) revealed that, "while there is recognition of the need for comprehensive data analysis in corporate advertising and promotional activities that takes into account internet advertising, offline advertising, and external influencing factors (stock prices, competitor movements, etc.), not many companies are actually putting this into practice."

In this survey, we will conduct a questionnaire survey about changes in the state of companies' efforts toward data utilization and analysis as of 2016, approximately two years after the 2 survey, and analyze and report the results.

*1: Offline advertising refers to advertising that does not involve the Internet, and includes television commercials, radio, newspapers, magazines, flyers, etc.
*2: Internet advertising refers to advertising placed on the internet, and includes listing ads, video ads, email ads, and SNS ads.
*3: Survey on the use of data to measure advertising effectiveness (2016) | Yano Research Institute Ltd. | July 2016, 7 (https://www.yano.co.jp/press/press.php/001549）

Summary of survey results

In 2016, few advertising and promotion professionals were actually implementing comprehensive analysis of online and offline advertising, despite the need for it. However, two years have passed, and now many advertising and promotion professionals have begun to implement it, with those who do not are in the minority. However, in many cases, analysis methods are limited to visualizing data using Excel or BI tools, and only 2% of advertising and promotion professionals are implementing advanced analytical methods such as statistical models, AI, and machine learning, and using visualized data to solve problems such as optimally allocating advertising budgets.

On the other hand, the most common answer (39.0%) was that they would like to try this in the future. This suggests that although there is a high demand for advanced analysis using technologies such as statistical models, AI, and machine learning in the comprehensive analysis of internet advertising, offline advertising, and external influencing factors, this has not yet become mainstream (is not yet in practice).

Additionally, the most frequently cited challenge in undertaking analysis was "difficulty in collecting offline and external data," followed by "lack of internal resources (manpower)" and "lack of analytical knowledge (no one who can perform analysis, etc.)."

These survey results suggest that in order for advertising and promotion personnel to be able to put analysis into practice more, solutions that make it easier to obtain offline and external data, reduce the amount of time required for analysis, and make analysis easier are needed.

Summary of survey results

1. Status of comprehensive analysis

We investigated what kind of data analysis advertising managers conduct to measure the effectiveness of their ads, and obtained the following results.

Regarding integrated analysis of online and offline advertising, in the 2016 survey, those who "want to try it in the future" exceeded those who "are currently working on it," but in this survey, those who "are currently working on it (30.2%)" exceeded those who "want to try it in the future (18.9%)" (red box in Figure 1).

In 2016, while there was awareness of the challenges involved in integrated analysis of online and offline advertising, few advertising and promotion professionals were actually able to put it into practice. However, two years have passed, and now many advertising and promotion professionals have begun to put it into practice, with those who have not yet started to do so being in the minority.

On the other hand, when it comes to comprehensive analysis that includes external influencing factors (seasonal factors, competitive situation, etc.) in addition to online and offline advertising, the number of people who "want to do this in the future (2016%)" exceeded the number of people who "are currently doing this (15.1%)" (green box in Figure 34.0), just like in the 1 survey, and the difference in response rate was the largest at 18.9 points (yellow box in Figure 2).

It appears that, just like in 2016, the need for comprehensive analysis that includes external influencing factors is still recognized, but for one reason or another, not many companies are able to put it into practice. (We will discuss the factors that make it difficult to undertake the analysis later.)

2. Analytical methods used

We surveyed advertising and promotion personnel who conduct data analysis to find out what analytical methods they use to conduct their analysis, and obtained the following results.

For all three of the responses, "Making decisions based on the previous fiscal year by referring to data such as past advertising amounts," "Aggregating collected data," and "Visualizing collected data using Excel or BI tools," the number of people who "are currently working on this" exceeded the number of people who "want to work on this in the future" (Figure 3), suggesting that data visualization using Excel or BI tools is already a commonly used data analysis method.

On the other hand, when it comes to "quantifying advertising effectiveness using technologies such as statistical models, AI, and machine learning, and simulating optimal budget allocation," only 4.9% of respondents said they were "currently using" this (red frame in Figure 3), and the difference with "want to use this in the future (39.0%)" was 34.1 points, the largest result (green frame in Figure 4).

The results of this survey indicate that while there is a high demand for advanced analysis using technologies such as statistical models, AI, and machine learning, this has not yet become widespread.

3. Challenges in conducting analysis

We investigated the challenges that arise when conducting data analysis to measure advertising effectiveness, and obtained the following results.

When asked about challenges in undertaking analysis, the most commonly cited answer was "difficulty in collecting offline data or external data," followed by "lack of internal resources (manpower)" and "lack of analytical knowledge (no one who can perform analysis, etc.)" (red frame in Figure 5).

The results of this survey suggest that, in order for marketers to be able to put analysis into practice more, they need solutions that make it easier to obtain offline and external data, reduce the amount of work required for analysis, and make analysis easier.

4. Why should we conduct an integrated analysis?

Below are the results of our survey on why it is important to conduct a comprehensive analysis of online advertising, offline advertising, external influence factors, etc. across categories.

Survey Outline

Investigation period:June 2018
Survey participants:106 people who meet the following criteria
・Employees of companies that place ads both online and offline*2
- Those who have worked in offline advertising at the company for the past year or currently work in that company
Research method:Web survey