Survey Report: Structural Dysfunction in M&A Matching Platforms and the Irreplaceability of Human Mediation
Chapter 1: The Wave of Digitalization and the Setback of M&A Platforms: Overview
The succession problem among Japanese SMEs has long been a social challenge. For business owners struggling with a lack of successors, M&A has become an effective means to avoid closure and pass on employment and technology to the next generation. Against this backdrop, numerous M&A matching platforms (hereafter, platforms) emerged to efficiently connect buyers and sellers through technology. However, many of these platforms have faced harsh reality—low closing rates, deteriorating information quality, and even withdrawal from the business itself—contrary to their initial ideal of “democratizing M&A.”
The essence of platform failure lies not in mere UI deficiencies or immature matching algorithms. It stems from the design philosophy itself: attempting to overcome the high walls of “information asymmetry” and “emotional decision-making” inherent in M&A transactions through digital systems alone. Particularly notable is the failure to capture “tacit knowledge”—management’s true intentions and on-the-ground nuances—while overemphasizing “explicit knowledge” such as financial data.
This report structurally analyzes why existing platforms have fallen into dysfunction, centered on the client’s hypotheses that “truly valuable information is only shared with people” and “AI has inherent limitations in interviewing.” The primary objective is to clarify the role technology should play in M&A—a domain requiring highly specialized expertise and relational trust—and the boundary of human elements that cannot be replaced.
Chapter 2: Structural Analysis of Existing Matching System Dysfunction
2.1 Lemon Market Dynamics and Information Quality Deterioration
The economics concept of “lemon markets” is highly effective in explaining the plight of current M&A platforms. In markets with strong information asymmetry, high-quality sellers (peaches) withdraw from the market fearing information leakage or inappropriate valuation, leaving only low-quality deals (lemons) on the platform.
Many platforms set low barriers to entry to increase registered deals. Yet this “ease of entry” fosters “ease of exit” and degrades information quality. Non-disclosure memoranda created by sellers without professional advisor involvement often fail to adequately describe the most critical risk factors and core strengths for buyers. As a result, buyers incur enormous research costs to find “winners” among the vast array of deals, ultimately becoming skeptical of platform-mediated transactions—a negative spiral.
2.2 Limitations of Professional Support and Increased Risk
Most platform-type services specialize in providing the “venue” for matching, with limited support for subsequent due diligence (DD), negotiation, and PMI (post-merger integration). M&A is not complete at closing; value is created only when integration succeeds. Many failure cases lack this understanding.
In fact, a majority of M&A participants face trouble, with backgrounds including negotiation deficiencies and insufficient investigation of buyer credibility. In simplified platform-mediated transactions, the “information screening” and “seller education” that intermediaries should provide are skipped, making financial risks (such as overlooked window dressing or discovery of off-balance-sheet debt) and potential compliance violations more likely to materialize.
| Impact | Description |
|---|---|
| Synergy shortfall | Expected synergies are not realized, making investment recovery difficult. |
| Goodwill impairment | Large accounting losses when acquisition price exceeds fair value. |
| Scandal/risk discovery | Off-balance-sheet debt or legal issues surface post-acquisition, damaging corporate image. |
| Key person departure | Talented personnel leave, undermining business continuity and competitiveness. |
2.3 Success Fee Model and the Trap of “Closing First”
The incentive structure of platform operators also contributes to failure. With a business model where fees are generated only upon closing, operators tend to prioritize “many closings” over “high-quality matching.” This leads to cases where M&A is inappropriate for the timing or where matching with unsuitable buyers is forced through.
Cases where sellers over-concede to buyers, leading to internal dissatisfaction erupting and negotiations breaking down, or post-closing internal conflict, are tragedies caused by forced matching. M&A becoming an end in itself rather than serving strategic intent is another factor that determines failure.
Chapter 3: Information’s Human Nature: Why “Valuable Information” Does Not Flow Digitally
3.1 Tacit Knowledge and the Importance of Non-Verbal Information
In M&A, much of the information that truly drives value does not appear in financial statements or business plans—“explicit knowledge.” Charisma of management, employee loyalty, long-term trust with specific partners, or reputation in the community—such “tacit knowledge” often constitutes a company’s true competitive advantage.
This information is extremely difficult to digitize and structure. One cannot enter “depth of trust between CEO and employees” in a platform’s input form, and management would not honestly record such information on a system viewable by unspecified parties. Such delicate information is “person-specific” and is disclosed only gradually, when the person is convinced the counterpart is trustworthy, through face-to-face dialogue.
3.2 Black-Boxed Know-How of Transferring Management
Especially in SMEs, decision processes and critical know-how often exist only in the management’s head. The risk of black-boxed items—matters that only specific managers or staff understand—without formal documentation or regulations is noted as an impact on transfer recipients.
Platforms lack mechanisms to capture this “invisible information.” Management may not be able to articulate what “special know-how” they possess. Skilled advisors unravel such black boxes through casual conversation with management and site visits, but digital systems lack the ability to extract information beyond pre-set items.
3.3 Confidential Information Management and Fear of Leakage
M&A consideration is the most confidential matter for management. Information leakage can fuel employee anxiety, damage credit with business partners, and give competitors an opening. Management therefore selects “where to share information” with extreme care.
Platforms tout system security, but management’s psychological barrier remains high. When selling one’s life’s work—the company—the “weighty decision” is supported not by faceless systems but by “humans” with accumulated trust: accountants and local bankers with long relationships. The reason valuable deals remain in offline networks rather than on platforms lies in this difference in “trust accumulation.”
Chapter 4: AI Interviewing Limitations and the Quality of “Questions”: Lack of Psychological Safety
4.1 The Essential Cause of AI Being “Poor at Asking”
As the client noted, AI interviewing is often evaluated as “poor at asking.” This stems not from language processing accuracy but from a lack of “contextual understanding” and “emotional empathy” in communication.
M&A interviewing is a process of “co-creation,” not “extraction.” Management is not merely returning accurate answers to questions. They are uncertain about their decision, feeling anxiety, and seeking someone to help them organize. The impersonal question “Why do you want to sell?” from AI sounds to management like a rude intrusion that ignores their “years of hardship since founding” and “heart-wrenching feelings.”
4.2 Interviewing Techniques and AI’s Functional Deficiency
Comparing professional interviewing techniques with AI reveals a clear gap.
| Element | Professional Advisor | AI/System |
|---|---|---|
| Ice-breaking | Build trust through regional topics, hobbies, industry discussion. | Start data input questions immediately after greeting. |
| Pacing | Create rapport by matching the other’s speech, volume, breathing. | Proceed at system’s fixed pace. |
| Use of silence | Wait for the other to organize thoughts, draw out true feelings. | Treat “no input” as “timeout.” |
| Deepening (Why) | Dig into motives and emotions behind surface answers. | Process pre-set questions in sequence. |
| Empathy expression | Show understanding through acknowledgment and summarization. | Confirm only “input received.” |
AI may be able to form hypotheses about “what they want to solve” or “what demands they will make,” but it lacks the flexibility to adjust question intensity or angle based on the atmosphere or management’s expression.
4.3 Psychological Bias and Rejection Response
Humans have “status quo bias” and “loss aversion bias,” feeling instinctive resistance to new systems and change. AI interviewing may bring the “benefit” of efficiency but is often perceived by users as “loss”—pain from “not being understood correctly” or “having one’s work and life treated lightly.”
When AI implementation projects prioritize technology over user empathy for feelings and anxiety, bias accelerates and can lead to rejection of critical information. Management does not “lie” to AI but may judge there is no need to volunteer “the truth (especially inconvenient truths).”
Chapter 5: Essential Examination of M&A Failures Through Examples
5.1 Major Losses from “Information Oversight” by Large Companies
Beyond the matching platform context, large-company M&A failures involving human experts also abound in information quality. In Toshiba’s Westinghouse acquisition, the buyer failed to detect improper accounting and massive nuclear business losses, ultimately incurring losses of up to approximately 700 billion yen. This shows that even highly skilled humans err when information transparency is not ensured.
LIXIL’s acquisition of China’s Joyou is another tragedy of black-boxed information. Joyou concealed its debt overhang and improper accounting, which even parent company Grohe failed to grasp. Countering such “intentional information concealment” requires not mere data comparison but human intuition to sense anomalies on the ground.
5.2 Site Discord and PMI Failure
Even when M&A reaches agreement in numbers, the business cannot run if the “people” on the ground do not move. In DeNA’s curation site acquisition, copyright infringement and inappropriate content production emerged post-acquisition, leading to closure of all sites. This stemmed from failing to grasp the company’s “way of working” and “ethics”—the most critical information—before acquisition.
Kirin Holdings’ failed acquisition of a Brazilian company also has “information quality” at its root—not just market research inadequacy but inability to adapt to local market conditions and culture. The result of selling for 77 billion yen after acquiring for 300 billion yen symbolizes how hollow the pre-acquisition valuation was.
5.3 Primary Cause of Failure: Hollowing of Due Diligence
“Due diligence (DD) insufficiency” is always cited as the main cause of M&A failure. But why is DD insufficient? Often it is because sensitivity to negative information dulls when rushing to close or from psychological anxiety to “not lose this deal.”
Platform-transacted deals are prone to bias toward reducing DD costs. Skipping expert engagement and relying on in-house document review creates a false sense of “having investigated.” But off-balance-sheet debt and contingent liabilities that become future litigation risk are never found in surface-level documents. Uncovering them requires experienced accountants and lawyers to persistently probe contradictions in management’s statements and scrutinize every detail.
Chapter 6: Behavioral Economics and AI Rejection: Dynamics of Trust Building
6.1 The “Weight” of Decision-Making and Distrust of Algorithms
M&A is for sellers the transfer of their life’s work; for buyers, a bet on company fate. In such high-stakes decisions, humans strongly demand “accountability” and “sense of conviction.”
Algorithmic matching tends to black-box the rationale for “why this counterpart is optimal.” Even if AI says “this counterpart is best,” management cannot take the final step without understanding the logic. In contrast, professional advisors provide explanations with narrative: “Combining Mr. XX’s technology with YY Corp’s sales network will produce this synergy. However, organizational culture differences require attention.” This narrative is what creates management’s “conviction.”
6.2 The Role of Human Mediation as “Emotional Cushion”
“Condition mismatch” almost inevitably arises in negotiation. A difference of millions or tens of millions in transfer price often develops into emotional conflict. At such times, AI offers only a “rational compromise point,” while skilled advisors function as a “cushion” to preserve both parties’ dignity and calm emotions.
Adjustments that appeal to both reason and sentiment—“The counterpart fully understands your hardship. However, given future risks, this price is their limit. In return, they promise maximum consideration for employee retention”—can only be made by humans. Platforms that skip this adjustment process and proceed mechanically naturally fail to reach closing.
6.3 Trust Propagation and Network Externalities
Successful intermediaries like Japan M&A Center do not directly seek sellers but partner with accounting firms and financial institutions that “already have established trust.” This is because trust in M&A is far more efficiently “transferred” from existing trust than built from zero.
Platforms have failed by disregarding this “trust infrastructure” and attempting to open management’s hearts through internet advertising and SEO alone. Management trusts the monthly accountant’s “try consulting there” more than a site appearing in search results.
Chapter 7: Optimal Solution for Platforms and Human Advisory
7.1 Evolution to Hybrid Model
Learning from existing platform failures, surviving services and emerging players are shifting to “hybrid (managed) type” that fuses technology and human capability. In the case of BATONZ, while boasting overwhelming user registration and deal volume on the platform, it clearly positions “accompaniment by experienced consultants” as the key to closing.
DX (digital transformation) handles “quantitative expansion” such as efficient reach and data management; consultants handle “qualitative deepening” such as emotional support and decision support. M&A is guided to success only when both wheels are in place.
7.2 Redefining AI’s Role: Not Interviewer but “Analysis Assistant”
AI’s capabilities should be maximized not in the front-end of interviewing but in back-end analysis of collected information. In “logical and structural tasks” such as risk extraction from vast financial data, comparison with similar deal closing prices, and contract clause checking, AI surpasses humans.
Human advisors conduct interviews with management; AI analyzes dialogue logs and provided materials to return feedback to advisors: “Additional confirmation needed here” or “This response contradicts earlier data.” This form of “AI that augments humans (Augmented Intelligence)” is the direction M&A platforms should pursue.
7.3 Improving Buyer-Side Deal Sourcing Skills
Leveraging platforms requires high literacy on the buyer side. Search techniques for finding quality deals and DD capability to assess the truth of presented information are essential skills for buyers.
Platform operators are expected to provide not merely deal listings but tools and expert networks for buyer education and proper DD implementation. Services that win market trust are those that provide “mechanisms” to find pearls amid low-quality information rather than lamenting quality.
Chapter 8: Conclusion—Market Redefinition and Future Outlook
Through this survey, the causes of M&A matching platform failures have been substantiated as deeply rooted in “information’s human nature” and “quality of interviewing,” as the client hypothesized.
Valuable information is disclosed only on a solid foundation of trust; replacing it with impersonal digital systems is currently impossible. AI interviewing failure stems from it not being a “process that respects the other’s life.” Unless this psychological gap is bridged, truly valuable information will not flow into the system.
This does not mean wholesale rejection of platforms. Technology still holds enormous potential for expanding information reach, reducing transaction costs, and increasing transparency. The winners in the future M&A market will be systems that achieve the following three points:
- Symbiosis with trust networks: Platforms that empower existing trusted professionals and financial institutions rather than monopolizing information.
- Design of emotional processes: Communication architecture that incorporates management’s psychological bias and emotional anxiety from the system design stage.
- Information quality assurance: Gatekeeper function that maximizes reliability of listed information through professional deal sourcing and multi-angle AI verification, not self-reports.
M&A is the transfer of “the company as a personality.” Redefining technology not as “a means to treat humans coldly” but as “a tool to create better encounters between humans”—without forgetting this essence—is the only path to break through the stagnant matching market.
References
- GENESIA Ventures: Structural competitive advantage learned from Japan M&A Center
- OnlyStory: M&A matching site guide 2025
- PR TIMES: Over 90% of M&A participants have regrets
- M&A Capital Partners: M&A failures
- Japan M&A Center: Causes of M&A failure
- SiteCatcher: Essence of M&A failure
- M&A Lead: 10 M&A failure cases
- M&A succeed: 15 M&A failure cases
- note: Overcoming psychological barriers in AI adoption
- Pro Seeds: Hearing in business
- Hammock: Sales hearing basics
- KOTORA: Consultant-style hearing examples
- M&A succeed: M&A benefits and drawbacks