Comix Inc. Releases Free 'GPT Image 2.0 + Codex Utilization Know-how Collection' to Reduce Production Costs, Begins Consultation for Implementation Support for SMEs

In small and medium-sized enterprises (SMEs), increased design outsourcing costs, lack of production speed, and the absence of personnel capable of effectively utilizing AI within the company are factors delaying the improvement of web strategies.

Meanwhile, by combining image generation AI with automated code generation, practical applications are beginning to spread, enabling the rapid advancement of landing page creation, banner design, screen design, and parts of website improvement.

Comix Inc. (Headquarters: Maruyama-cho, Shibuya-ku, Tokyo; Representative Director: Akihiro Suzuki) is making public for free its 'GPT Image 2.0 + Codex Utilization Know-how Collection,' which Representative Director Akihiro Suzuki compiled through in-house operations and client support. Concurrently, the company is starting to accept consultations for implementation support tailored for SMEs.

## Background and Challenges

In SME web marketing, the frequency of updating deliverables directly linked to sales, such as LPs, ad banners, sales materials, diagrams, and service website improvements, determines results. However, outsourcing costs and verification man-hours are incurred for each production, and it is not uncommon for budgets and time to run out before multiple improvement proposals can be tested.

According to a survey by the Tokyo Chamber of Commerce and Industry, 'cost burden' was the most common challenge (31.9%) for digital shift and DX in SMEs, with 'lack of personnel capable of leading the initiative' also reaching 31.0%. Furthermore, 77.9% of companies that advanced their digital shift felt results, with 'operational efficiency improvement' being the most frequently cited effect at 81.0%.

In this situation, generative AI is becoming not just an image creation tool, but a practical foundation that supports everything from organizing production requirements, analyzing reference images, generating multiple proposals, implementation, and verification. On the other hand, many companies fail to achieve expected results due to common pitfalls such as not providing reference images, not organizing brand rules, giving vague instructions like 'make it look nice,' or making judgments based on only one proposal.

This know-how collection is a practical resource for companies in the early stages of consideration to organize 'where to start' and 'how much to leave to AI and where human judgment is needed.'

## Offerings

1) Free Public Release of 'GPT Image 2.0 + Codex Utilization Know-how Collection'

Comix Inc. will systematize the insights confirmed through its in-house operations and client support into six fields of practical know-how, which will be publicly released and provided for free.

- Free Public Document: GPT Image 2.0 + Codex Utilization Know-how Collection
- Document URL: To be replaced after release

Main contents of the document (excerpt):

- Production flow for designing and implementing a web app from a single screenshot
- Application methods for landing page creation, ad banners, diagrams, screen design, and web components
- Approach to reducing design outsourcing costs from 500,000 yen/month to approximately 30,000 yen/month
- Operational design to maintain quality and consistency by having AI incorporate brand guidelines
- Common pitfalls in implementation and a 30-day small-scale trial roadmap

2) Consultation Acceptance for 'GPT Image 2.0 + Codex Implementation Support Package'

In addition to the document release, the company will begin accepting consultations for an implementation support package for companies aiming to internalize production operations and improve improvement speed. The support is designed with the premise that even companies lacking dedicated IT personnel or in-house designers can verify effects through small-scale trials.

- Environment Setup: Organizing tools, permissions, production flow, and security policies
- Template Preparation: Creating brand guidelines, prompts, and review standards
- Practical Support: Executing verification of landing page or banner production, measuring effects, and formalizing operational rules

Click here for free consultation on AI utilization

By answering a free diagnostic questionnaire, you can register for an online free consultation. The document will be provided during the online consultation.

## Features and Strengths

- Design based on back-casting from management challenges: Designed not just for image generation, but with the goal of improving outsourcing costs, production time, number of iterations, and ad verification speed.
- Systematization of 20 pieces of practical know-how: Organized into six fields: production flow, cost structure, quality control, development style, human resources/organization, and risk management.
- Operation with human judgment retained: Assumes a system where humans are responsible for brand, expression, rights confirmation, and final quality judgment, rather than completely entrusting everything to AI.
- Premise of a small start: Initially, start with redesigning 20 types of SNS banners or one existing landing page, and compare quality, cost, and time required with outsourcing.

## Target Users and Use Cases

Target Users

- SME managers: Those who want to reduce design outsourcing costs and increase the frequency of web strategy improvements
- Marketing managers: Those who want to speed up the creation and verification of landing pages, ad banners, and SNS materials
- Designers/Engineers: Those who want to strengthen their instruction ability, judgment, and quality control skills in the AI era

Example Use Cases

- Ad Improvement: Generate multiple ad banner proposals and increase the rotation speed of comparative tests
- Site Improvement: Visualize improvement proposals in a short time by referring to competitor landing pages or existing pages