AI Prompt Library for Dealmakers

The Evolving Role of AI in Due Diligence

The integration of artificial intelligence into M&A due diligence processes represents a significant evolution in strategic analysis and deal execution. While the technology's capability to process vast datasets rapidly is undeniable, its true value is realised through precise and contextually aware prompting. Effective AI prompting transforms a mere data processing engine into a sophisticated analytical co-pilot.

Traditional due diligence often involves extensive manual review of documents, a process susceptible to human error and time constraints. AI, particularly large language models, offers a demonstrable advantage in automating preliminary scans, identifying anomalies, and summarising critical information. However, the quality of AI output remains directly proportional to the clarity and specificity of the input prompts. Generic queries yield generic responses, whereas finely tuned prompts unlock deeper insights.

This article outlines a foundational approach to prompt engineering for dealmakers, focusing on principles that enhance analytical rigour. It moves beyond superficial interactions to explore how structured prompting can drive more effective engagement with AI tools, such as those integrated into platforms like Lens, our AI-powered data room. The objective is to equip professionals with a practical prompt library, facilitating a more precise and efficient due diligence cycle.

Foundational Principles of Prompt Engineering for Dealmakers

Effective prompt engineering for due diligence is predicated on several core principles. Foremost is specificity. A prompt must clearly define the task, the scope, the desired output format, and any relevant constraints. Ambiguity in a prompt will inevitably lead to ambiguous or unhelpful outputs.

Contextualisation is equally critical. AI models perform best when provided with sufficient background information about the deal, the industry, and the specific documents under review. This enables the AI to interpret information through a relevant lens, mimicking the contextual understanding a human analyst would apply. For instance, when reviewing a SaaS contract, specifying the industry and common SaaS metrics allows the AI to highlight pertinent clauses more effectively.

Iteration is another fundamental principle. Initial prompts rarely yield perfect results. Dealmakers should anticipate refining their prompts based on the AI's initial responses. This iterative process allows for a progressive narrowing of focus and an improvement in the quality and relevance of the output. It is a dialogue, not a monologue.

Finally, the concept of 'persona' can be instrumental. Instructing the AI to adopt a specific role, such as 'You are a corporate finance analyst reviewing a target company's financials' or 'You are a legal counsel identifying key risks in a commercial contract', can steer the AI's analytical framework and response style to better align with professional expectations. This enhances the utility and applicability of the AI-generated insights.

Prompt Library: Battle-Tested Prompts for Key Due Diligence Areas

CIM Analysis

Analysing the Confidential Information Memorandum (CIM) is an initial, critical step. AI can accelerate the identification of key investment highlights, potential red flags, and areas requiring deeper investigation.

Prompts:

1. “Summarise the key investment highlights and potential risks outlined in this CIM for a technology company in the enterprise software sector. Focus on revenue growth drivers, customer churn rates, and competitive differentiation. Output in bullet points.”
2. “Extract all financial projections for the next three years, detailing revenue, EBITDA, and free cash flow. Identify any underlying assumptions that appear optimistic or conservative. Present this information in a table format.”
3. “Identify any mentions of contingent liabilities, ongoing litigations, or regulatory compliance challenges. For each, summarise the nature of the issue and its potential impact on valuation. List these sequentially.”
4. “Based on the provided CIM, what are the top three strategic synergy opportunities for a potential acquirer in the [specific industry]? Elaborate on the rationale for each.”
5. “Perform a SWOT analysis of the target company as presented in the CIM. Prioritize factors relevant to a private equity investment thesis.”

Contract Review

Reviewing voluminous contracts is a time-consuming but essential aspect of due diligence. AI significantly reduces manual effort by highlighting critical clauses, obligations, and potential areas of concern.

Prompts:

6. “From this customer contract, identify all clauses related to change of control, assignment, and termination rights. For each, summarise the key provisions and any notice periods required.”
7. “Extract all material liabilities, indemnification clauses, and warranties from this vendor agreement. Quantify any monetary caps or limitations of liability mentioned.”
8. “Analyse this intellectual property licence agreement. Identify the scope of the licence, any geographical or time limitations, and specific provisions regarding sublicensing or transferability.”
9. “List all non-compete and non-solicitation clauses present in this employment contract. State their duration, geographical scope, and any specified remedies for breach.”
10. “Review this lease agreement and identify all critical dates, renewal options, and tenant improvement allowances. Note any extraordinary landlord obligations.”

Q&A Drafting

Diligence Q&A lists are central to uncovering detailed information. AI can assist in anticipating questions and structuring comprehensive inquiries based on initial document review.

Prompts:

11. “Given the financial section of this CIM, draft five incisive questions regarding the company's working capital management and cash conversion cycle. Assume the role of a seasoned corporate finance analyst.”
12. “Based on the identified risks in the technology due diligence report, formulate three follow-up questions for management concerning cybersecurity protocols and data privacy compliance.”
13. “Review the provided cap table. Draft two clarifying questions regarding preferred stock liquidation preferences and vesting schedules for key employees.”
14. “Using the sales contracts reviewed, propose three questions for the sales leadership team regarding average contract value (ACV) trends, customer retention strategies, and sales force productivity metrics.”
15. “From the environmental report, formulate two questions for the operations team regarding their waste management practices and compliance with recent environmental regulations.”

IC Memo Critique

Internal Committee (IC) memoranda synthesise due diligence findings for investment decisions. AI can provide a quick, objective critique, identifying gaps or biases.

Prompts:

16. “Critique this investment committee memo, identifying any unsupported assertions regarding market size or competitive landscape. Highlight references that lack specific data points.”
17. “Review the identified risks in this IC memo. Are there any significant risks from the due diligence findings that appear to be understated or omitted? List them with rationale.”
18. “Assess whether the valuation methodology presented in this IC memo is adequately supported by the financial projections. Point out any discrepancies between the two.”
19. “Analyse the strategic rationale section of the memo. Does it clearly articulate the value creation thesis, and are there any aspects that seem vague or overly optimistic given the due diligence findings?”
20. “From a risk management perspective, identify any areas in this IC memo where mitigating actions for identified risks are not sufficiently detailed or appear unrealistic.”

Anti-Patterns and Calibration Strategies

Common Anti-Patterns

While AI offers considerable advantages, certain anti-patterns in prompting can impede its utility. Vagueness is perhaps the most prevalent. Prompts like 'Summarise this document' are too broad, leading to generic and often unhelpful outputs. Similarly, over-reliance on default settings without customisation neglects the specific nuances of M&A.

Another anti-pattern is lack of iterative refinement. Treating the AI as a 'one-shot' solution rather than a tool for ongoing dialogue limits the depth of insights. Ignoring AI limitations, particularly regarding highly subjective or complex legal interpretations requiring human judgment, can lead to misinformation or flawed analysis. AI can flag, but human expertise must interpret and decide.

Calibration Strategies

Effective calibration involves a continuous feedback loop. Refining prompts based on initial outputs is paramount. If an output is too broad, narrow the scope. If too brief, ask for more detail. Specify examples of desired output formats (e.g., 'table format with columns for X, Y, Z').

Providing specific examples within the prompt can also guide the AI's understanding. For instance, when asking for a specific type of clause, provide an example of such a clause from a different document. This 'few-shot learning' can greatly improve relevance.

Adjusting temperature and randomness in more advanced AI interfaces, where available, can control the creativity versus factual adherence of the output. For due diligence, a lower 'temperature' is generally preferred to ensure outputs remain grounded in the provided text rather than generating speculative content. Utilising tools like Lens with its integrated AI capabilities allows for bespoke prompt refinement within a secure data room environment, ensuring that the AI is not just processing data, but generating precisely actionable intelligence for the deal. This approach aligns with Beyond M&A's philosophy of leveraging technology to enhance, not replace, strategic human judgment in complex transactions.

Conclusion

The strategic application of AI in due diligence is no longer experimental; it is an increasingly integral component of efficient dealmaking. By mastering prompt engineering, dealmakers can transcend basic AI interactions, transforming the technology into a potent analytical partner. The curated prompts and insights into effective calibration presented here serve as a robust foundation. As AI capabilities continue to advance, so too will the methodologies for leveraging them, making continuous learning and adaptation essential for competitive advantage in M&A. The future of due diligence is demonstrably more intelligent, precise, and agile.