One can have the rating from people of certain age and gender, compare the attitude of the target audience to various versions of design advertising materials and see which of them have the highest emotional response. This knowledge may be helpful in the startup of a successful advertising campaign and minimize the risks of inefficiency.

The possibility to know the users involvement (the extent of their focusing on an advertising material) helps identifying the most efficient posters in terms of information about the company’s values, as well as the elements which kept the users’ attention.

By adding the new identified elements contributing to the users involvement the communication materials may be corrected.

Let’ consider specific cases from Globus IT practice and see how it can be implemented in real life.

Goals:

• Evaluate the image-building posters designed for placement in the clients’ offices.
• Make a poster layout study to correct the design before the posters are placed in the company’s branches.

Tasks:

1. Define which posters are the most eye-catching.
2. Figure out the most attractive things in the posters and which emotions they bring out when looking at them.
3. Define the reasons of positive and negative ratings.

Method 1. Online focus-group

• 12 persons, individual persons.
• 2 age groups:

— junior group – 25-35 years old;
— senior group – 35-55 years old.

• 8 clients from the company, 4 clients from third parties

Qualitative method:

• The Customer could monitor the live discussion of posters between the clients/non-clients.
• Users’ emotional response was observed when viewing certain topics.
• Users’ perception and understanding of the company’s concepts and narratives were clearly seen.

Method 2. Computer vision based online analysis

Quantitative method:

• Conclusions made by the focus group were tested on a wider audience.
• Users’ emotional response and involvement during viewing were analyzed.
• Posters most relevant in terms of the Customer’s goals were identified.

Focus-group:

• 60 persons, individual persons
• 3 age groups:

— junior group — 25–30 years;
— middle-age group — 31–40 years;
— senior group — 41–55 years.

Algorithm:

1. Engage a target audience group with a profile matching the future posters “viewers”
2. Show the posters to the target audience online and film their faces through the web-camera.
3. Scan the facial gesture (coordinates of the viewer’s sight may be scanned – use the eye-tracking technology).
4. Define the TA emotional response and involvement (interest) on the video.
5. Make a report:

— posters ranking by level and type of emotional response;
— posters ranking by the level of interest;
— identification of differences in the viewers’ response with the breakdown by: gender, age, client or non-client
— a heatmap and viewers sight scan pattern may be built.

Results
1. Pace:
— 1 day – AI-based analytics;
— 2–3 days – gathering of a focus group.
2. Reliability. The most attractive posters were selected and ranked. AI analysis data matched with the ratings made by the focus group.
3. Saving. Cutting of costs — production of content which is really interesting for the Client.

Artifact-report:
• detailed analysis of the posters rating;
• response and quotations of the respondents;
• emotional response and involvement plots;
• advice on the areas to be improved.

Let's have an insight into Adreview solution which was the basis for the study.
Adreview — is a service from Globus IT team designed to evaluate and forecast the effectiveness of the computer vision based advertising campaign.

Parameters to be evaluated:
Video and banners advertising

What is effectiveness?
This includes indicators of two metrics:
• VCR for video advertising = ''video completion rate''
• CTR for banner advertising = ''click-through rate''

What is used for evaluation:

Based on emotional response of 6 emotions (joy, surprise, sadness, disgust, fear, rage), as well as their intensity and «sign», i.e. indicating if these emotions are positive or negative;
Involvement = degree of the viewer's concentration on the demonstrated content;
Eye-tracking data: heat map + viewer's sight scan pattern.

Tasks to be solved:
— How to attract and hold attention.
— What are the indications that the image/video has failed
— How to cut the marketing costs
— How to make the testing process faster compared to the rivals.

Adreview application areas
— Large business: banks, telecom, retail sales, pharmacology, developers, technology services (FoodTech, EdTech, RideTech и пр.)
— Small business: Market place sellers (Ozon, Wildberries, Yandex)

“Mimic is an emotion indicator. Same emotions are expressed as similar mimic patterns, i.e. a combination of muscles motion that can be classified as 6 base combinations — says Ahmet Kuralkan, Country Manager Globus IT. — Any visual imagery triggers human emotions, and advertising is not an exception. This emotional response clearly correlates to the advertising campaign effectiveness”.

The effectiveness means variation in the target metrics:
• for images — click-through rate (CTR);
• for video — video completion rate (VCR).
Both metrics influence the final action: a purchase, a phone call, a signup and etc.

“By using computer vision to define the viewer’s emotional response to an image/video in the focus-group — face expression, mimic, emotions, we may estimate the results of the advertising campaign" — says Ahmet Kuralkan.

Eventually, all posters triggered an emotion of joy that was prevailing among the viewers. Also, analysis allowed identifying the differences in perception among various age groups.

Some interesting results were demonstrated in perception of posters among the bank client/non-client group. The non-clients occurred to have even a more positive response to the posters which means that posters may have positive effect on the decision to change the bank.

“These studies allowed evaluating the advertising layouts quickly and properly and defining which of them can clearly demonstrate the Bank values for various segments of the target audience, — summarizes Ahmet Kuralkan. — Due to the use of the qualitative neuromarketing study, conclusions made by the focus-groups for a short-time period were tested on a wider audience and at a deeper level of neurophysical reactions. As a result, the most suitable visual patterns were identified, activating the TA, raising efficiency of the appropriate communication strategy based on these patterns”.

We have analyzed the fly-pages of the company selling cotton flower bunches.

The study had the following task:
• compare appeal of the Client’s fly-page with the fly-pages of the rivals;
• figure out the top-3 “anchor” elements of the fly-page.

Input data:
• 8 fly-pages from 8 different sellers at the marketplace
• 1 fly-page of our Client
• 40 viewers:
— gender: female
— age: 25–40 / 41–55+ years old

Output:

1. top-3 “anchors" are as follows: discount, price, customer reviews;
2. on the photo — focus on the product itself: “bowls distract the viewers’ attention;
3. infographics is a helpful tool.

As was requested by the Customer of the ad video we have analyzed the viewers’ emotional response to a social advertisement against hazardous and banned cultural trend spread among the young people to know how efficient this advertisement will be for its target audience.

Input data:
51 viewers:
— gender: male/female
— age: 18–25 / 26–40 years old
— education: high school/other

Task — find the differences in perception of information depending on the following:
— level of education
— region
— age
— gender

Questions for data labeling:
1. Did you get the idea of the advertisement video? (yes/no — 51/0)
2. Have you ever thought about the talking point discussed in this video? (yes/no — 38/14)
3. Do you agree with the message in this ad video? (yes/no — 48/4)
4. Will you recommend this ad video to other people? (yes/no — 44/8)

Output:
• the extent of the viewers involvement clearly differs depending on the level of education, age and region
• there's also a distinct difference depending on the degree of immersion in the topic of the advertisement video.

Conclusion:
If the target audience is young people without a college degree this video advertisement shall be corrected.

Ahmet Kuralkan, Country Manager Globus IT:
"Due to a draft evaluation using a neuro-model of advertising messages, creativity, design and videos and by selecting optimal components for each advertising campaign the efficiency of marketing activities may be increased.
Testing can be carried out for any target group, assessing the involvement and shaping the designed image and perception of the brand.
This will allow not only making an efficient advertisement maximally attractive for the target audience but also cutting the costs due to removal of non-efficient advertising campaigns."