3C-AI — Understanding Competition When AI Shapes the Shortlist

6일 전

10분 분량

3C-AI is a strategy analysis model proposed by Jace Shim in AI 승자독식 (Raonbook Plus, 2026). It keeps the three elements of Kenichi Ohmae's 3C analysis — Company, Customer, and Competitor — and adds one more question: how does generative AI, acting as an information intermediary, change which alternatives a customer sees and the criteria on which they are compared? [1][2]

In one sentence: 3C-AI helps a company read competition question by question, because the rivals and the comparison criteria a customer meets in an AI-generated answer depend on the conditions written into that customer's question. [1]

3C-AI is not an official framework of Google, OpenAI, or any other AI provider. It is my proposed extension of a classic management tool for a market in which many buyers now meet a short, AI-assembled shortlist before they visit any company website. [1]

Why revisit the 3C analysis now?

In a strategy workshop with a company that sells booking software to small businesses, the management team listed five competitors: other booking applications with similar features. We then asked an AI assistant a question a real prospect might ask: "How should a solo instructor who takes about ten bookings a week manage reservations?"

The answer did not mention a single booking application. It recommended a free shared calendar, the booking feature inside a messaging app, and a simple online form. For that question, the company's real competitors were free tools and the option of not buying anything yet. The competitive map drawn from an industry classification and the map drawn from the customer's question were completely different.

The shift is not marginal. Pew Research Center analyzed the March 2025 browsing records of 900 U.S. adults and found that users clicked a traditional search result link in 8% of visits when an AI summary appeared, compared with 15% when no summary appeared. [3] Many customers now read an AI-selected comparison before they read a company's own explanation.

The original theory: Kenichi Ohmae's 3C analysis

The 3C analysis was articulated by the Japanese strategist Kenichi Ohmae in The Mind of the Strategist. [2] Ohmae argued that strategy should be built by looking at three actors together — the corporation, the customer, and the competition — which is why the 3C model is also called the "strategic triangle."

The strength of the 3C model lies in its simplicity. When a team places what it does well, what customers want, and what competitors offer on one page, the strategic gaps become visible. The space where customers have a need, competitors fall short, and the company performs well is the ground worth fighting for.

The model has remained the opening page of many strategy courses and consulting engagements for decades. New business reviews, brand repositioning projects, and executive workshops still begin with a 3C map. 3C-AI does not replace that map. It keeps all three elements and adds a question that the original model did not need to ask separately.

What the 3C model misses in the age of AI: an intermediary that rewrites

The 3C model was never unaware of intermediaries. Retailers, journalists, comparison websites, and word of mouth have always stood between companies and customers. Generative AI, however, differs from earlier intermediaries in one decisive way.

Earlier intermediaries mostly passed information along. A customer read a newspaper article or a comparison chart and drew their own conclusion. A generative AI system selects several sources, combines them, and writes a new answer, so the customer often reads a rewritten summary instead of the original material.

[Table 1] Traditional information intermediaries vs. generative AI intermediaries — why 3C-AI treats AI mediation as a separate lens

Dimension	Traditional intermediaries (media, comparison sites, retailers)	Generative AI intermediaries (AI assistants, AI search summaries)
How information is handled	Passes along or lists original material	Selects and combines sources, then rewrites them as a new answer
What the customer reads	Original articles, comparison charts, store shelves	A reconstructed summary answer
How alternatives are grouped	By industry or product category	By the conditions and constraints in the customer's question
Number of options shown	Dozens of products or about ten links	Often a handful of names
Competitors a company tends to miss	Few — rivals sit on the same shelf	Substitutes from other categories and the option of not buying

Source: Author's reconstruction based on Shim (2026), Chapter 4. The table describes typical patterns, not a measured rule for every AI service. [1]

The key row in Table 1 is the third one: the basis for grouping alternatives. A traditional intermediary placed products from the same category side by side. A generative AI system can assemble candidates according to the conditions a customer writes into the question — budget, team size, deadline, existing systems. That is why 3C-AI reviews AI mediation as a distinct part of the analysis.

What is 3C-AI? The definition

3C-AI (Shim, 2026, Chapter 4): a strategy analysis model that examines the relationship among Company, Customer, and Competitor while also reviewing how AI's selection of information and construction of answers affect the strengths and alternatives that reach the customer. [1]

In 3C-AI, AI is not a fourth "C" and not a fourth competitor. It is an intermediary that selects and arranges information among the other three. It helps decide which candidates enter the same arena and on which criteria they are compared. The company's task is to read how that arena is being set up and to make sure its real strengths can be explained accurately within it.

This framing matters for practice. If AI were treated as a competitor, the natural response would be to "beat" it. Treated as an intermediary, the right response is to understand what it can find, what it compares, and which conditions it keeps or drops.

How does 3C-AI differ from the classic 3C analysis?

[Table 2] Classic 3C questions and the questions 3C-AI adds — four perspectives

Perspective	Classic 3C question (Ohmae)	Question added by 3C-AI (Shim)
Company	What are our strengths and weaknesses?	Are those strengths visible in public information? Does a strength change meaning when its conditions are dropped?
Customer	Who are our customers and what do they want?	What conditions and constraints — budget, time, staff, existing systems — do customers put into their AI questions? What do they still check after reading the answer?
Competitor	Who competes with us in our market and what do they do?	Which alternatives does AI compare with us for a given question, including substitutes and the option of not buying?
Intermediary (AI)	(No separate element)	On which criteria does the answer compare options, and which conditions does it retain or omit?

Source: Classic 3C questions after Ohmae (1982). Added questions from Shim (2026), Chapter 4. [1][2]

The most important row in Table 2 is the last one. The classic model has no separate element for the intermediary, while 3C-AI names the way an answer frames a comparison as an object of analysis. The other three rows inherit Ohmae's structure and add one question each that has become necessary in the era of AI answers.

How to read each of the four perspectives

Company: the company we know versus the company AI can find.

Executives and websites often describe different strengths. Management may say, "We have never missed a delivery date," while the public website says only "the highest quality." An AI system can work only from accessible information, so the gap between those two lists is often the first piece of evidence of what Shim calls the citation gap — the difference between a firm's actual competitiveness and the competitiveness conveyed in AI answers. [1]

Customer: the constraints hidden in the question.

In front of an AI assistant, customers describe not only their industry and size but also their constraints: "We have no accountant, most of our customers pay from abroad, and we need this running by next month." Constraints usually fall into four groups — budget, time, people, and existing systems. Because answers reflect those constraints, customers in the same segment can receive very different shortlists.

Competitor: the question sets the boundary of competition.

Customers do not ask questions by reading an industry classification; they try to solve a problem. AI answers therefore surface substitutes from other categories and non-purchase alternatives such as "wait until demand is proven." 3C-AI asks teams to record both the competitors in their industry and the competitors defined by each question.

Intermediary: read the arena, but do not claim to know the engine.

If a brand appears repeatedly in observed answers, that alone does not prove the AI system prefers it. 3C-AI is not a tool for guessing the internal workings of an AI model. It is a tool for recording the shape of the comparison and the gaps in the explanation as they appear in actual answers.

The question defines the boundary of competition

[Table 3] Industry competitors vs. alternatives defined by the customer's question — booking software (hypothetical example)

Customer question	Competitors by industry classification	Alternatives an AI answer may compare
A booking, payment, and instructor-management system for a multi-location Pilates studio	Booking applications with similar features	Similar booking applications; industry-specific all-in-one platforms
How should a solo instructor with about ten bookings a week manage reservations?	Booking applications with similar features	Free shared calendars; booking features in messaging apps; online forms
How can I reduce no-shows?	Booking applications with similar features	Text-message reminder services; deposits; changes to cancellation policy
We open next month — should we adopt a booking system now?	Booking applications with similar features	Waiting to confirm demand (non-purchase alternative); free trials

Source: Illustrative example based on Shim (2026), Chapter 4. These are hypothetical scenarios, not recorded AI responses. [1]

The middle column of Table 3 is identical in every row, while the right-hand column changes with each question. The same company competes against entirely different alternatives as the question changes. For that reason, a 3C-AI competitive map is not one page; it is one page per important customer question.

Two arenas: competition in products and competition in explanations

[Table 4] Product competition vs. explanation competition — the two arenas 3C-AI examines together

Dimension	Product competition	Explanation competition
When it happens	After the customer uses the product	Before the customer chooses
What decides the outcome	Quality, price, performance, service	Public evidence, clarity of conditions, third-party assessment
Main stage	Real-world use	AI answers, comparison articles, search results
What winning brings	Repeat purchase, referrals, loyalty	Inclusion on the shortlist, a first conversation
Winning only one arena	"Great, but nobody knows it"	"Well known, but disappointing"

Source: Shim (2026), Chapter 4. [1]

The last row of Table 4 is the warning built into 3C-AI. A company that wins only the explanation contest will eventually disappoint customers, and their reviews become new material for AI answers. An explanation should be a mirror that reflects the product accurately, not a shadow larger than the product.

How to apply 3C-AI in five steps

1. Select the questions. From sales conversations and support records, choose the five customer questions that matter most to revenue.

2. Ask and record. Put those questions — without your company name — to two or three AI services. Record the candidates named, the comparison criteria, the conditions mentioned, the sources cited, and the date.

3. Classify the competitors. Sort every candidate into three groups: industry competitors, substitutes from other categories, and non-purchase alternatives.

4. Read the comparison criteria. Note whether the answer compared options on price, ease of use, a specific feature, customer size, or something else.

5. Check your evidence. For each criterion, confirm whether public evidence supports your strength. Move every empty cell to next quarter's work list.

The worksheet is not a one-off exercise. AI services and customer questions keep changing, so fill it in again each quarter with the same questions and read the trend rather than a single snapshot.

Propositions and limits of 3C-AI

[Table 5] The two propositions of 3C-AI and their managerial implications

Proposition	Statement	Implication for management
Proposition 1	As the constraints in customer questions change, the alternatives and explanatory evidence a company must review may also change.	Update the competitive analysis for each important question, not once a year.
Proposition 2	When the application conditions of a public strength are unclear, the gap between actual competitiveness and the comparison conveyed to customers may widen.	Publish strengths together with their conditions so they can be conveyed accurately.

Source: Shim (2026), Chapter 4. [1]

Both propositions end with "may." 3C-AI does not claim that AI answers always widen the scope of competition, and it is not a formula for calculating the probability that an engine will select a brand. In markets where long-standing relationships or offline experience decide purchases, the influence of AI mediation may be limited.

Frequently asked questions

Is 3C-AI the same as a 4C analysis?

No. In 3C-AI, AI is not a fourth "C." It is analyzed as the intermediary that conveys the relationship among Company, Customer, and Competitor to the customer.

Should a company treat every brand an AI answer recommends as a competitor?

Record it as a candidate to observe, then confirm it with customer records and repeated observation over several days. AI answers vary by service, wording, and time.

Which companies benefit most from 3C-AI?

Companies in categories where customers ask long, specific questions; companies seeking new or international customers; and companies that compete with free tools or substitutes.

What comes after 3C-AI?

After drawing the competitive map, use VRIO-D to check which resources and public evidence you can bring to that competition, then use STQP to decide which questions to focus on.

References and provenance

The 3C-AI model, its definition, and its propositions are attributed to Jace Shim's Korean-language book AI 승자독식 (Raonbook Plus, 2026), Chapter 4. The established theory and research listed below provide background; they do not constitute independent validation of Shim's extension.

[1] Shim, J. (심준규). (2026). AI 승자독식 [Korean-language book; title translated descriptively as "AI Winner-Takes-All"]. 라온북플러스 (Raonbook Plus).

[2] Ohmae, K. (1982). The mind of the strategist: The art of Japanese business. McGraw-Hill.

[3] Pew Research Center. (2025, July 22). Google users are less likely to click on links when an AI summary appears in the results.

[4] Aggarwal, P., Murahari, V., Rajpurohit, T., Kalyan, A., Narasimhan, K., & Deshpande, A. (2024). GEO: Generative engine optimization. Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining.

The comparison tables are explanatory reconstructions, not reproductions of tables from the cited works. Hypothetical examples are illustrations, not observed AI responses or measured business outcomes.

How to cite this article

Shim, J. (2026, October 16). 3C-AI — Understanding Competition When AI Shapes the Shortlist. TSCB Group — INSIGHT B. https://www.tscbgroup.com/single-post/3c-ai-model-en

When citing the original proposal rather than this web explanation, cite the Korean-language book AI 승자독식 (Raonbook Plus, 2026). The English wording of its title in brackets is descriptive, not a claim about a separately published English edition.