I have spent the last year reading IMs at a volume no human practitioner ever could. Building NaS_OS has meant running fifty-plus real and reference IMs through extraction pipelines, comparing them against the underlying data, and watching where the documents succeed and where they fail. The pattern is consistent enough that I want to write it down for the analysts and senior associates drafting these documents in 2026.
This is not a stylistic guide. It is an operational guide. The IM that closes a deal is structurally different from the IM that doesn’t, and the differences are mostly fixable with a few hours of deliberate revision.
The five structural mistakes I see most often
Mistake 1: Industry overview that overwhelms the company story. The most common opening in a weak IM is a 10-page industry primer that could appear in any IM in the sector. By the time the buyer reaches the company itself on page 12, the document has communicated nothing specific. The fix is to compress the industry overview to the minimum needed to frame the company’s position, usually 2-4 pages, and to make every paragraph in that section explicitly relevant to why this company wins in this industry.
Mistake 2: Executive summary that does not actually summarize. The executive summary of a weak IM lists facts. The executive summary of a strong IM makes an argument. If a buyer read only the first two pages of the document, they should understand the equity story, the financial trajectory, the management team’s credibility, and the strategic rationale for the sale. Most executive summaries I read give the buyer four of these and miss one, usually the strategic rationale, which is the section that signals what kind of process this will be.
Mistake 3: Financial section that doesn’t speak to the equity story. I covered this in the equity story post and I will repeat it here because it is the single highest-leverage fix: the financial section should be designed to prove the equity story, not to dump three years of P&L. Every chart should be there for a reason that connects back to the central claim of the document. If a chart does not serve the equity story, it does not belong in the IM.
Mistake 4: Misordered sections. The default IM template orders sections roughly as: executive summary, company overview, market, business model, financials, growth strategy, management, risks, projections. This order works for some equity stories and is wrong for others. A platform story should put the platform architecture early, before the historical financials. A consolidator story should put the M&A roadmap front and center, not buried in growth strategy. The structure should serve the story; the template should be edited to fit, not followed by default.
Mistake 5: Management section that reads as a LinkedIn page. The weakest management sections list resumes. The strongest ones make the case for execution. If the CEO previously scaled another business through a similar transition, that is the story. If the CFO came from a public company and brings institutional finance discipline, that is the story. If the COO has been with the company for fifteen years and represents continuity, that is the story. The buyer does not care where the team went to college. The buyer cares whether the team can execute the plan being sold.
Three patterns that distinguish high-converting IMs
Pattern 1: A defensible thesis paragraph. The strongest IMs have a single paragraph somewhere in the executive summary that articulates the investment thesis in a way that could be lifted directly into a buyer’s investment committee memo. Test: if a buyer copy-pasted that paragraph into their IC memo, would it stand up on its own? If yes, the IM has done its job at the headline level. If no, the IM is making the buyer write the thesis themselves, and most buyers won’t.
Pattern 2: Numbers that tell a consistent story. Every figure in the IM should be reachable from every other figure through visible logic. Revenue grows from X to Y because of A, B, and C. EBITDA margin expands from M to N because of D and E. Customer count moves from P to Q at acquisition cost R. The buyer should be able to trace any output back to its inputs without leaving the document. When numbers appear without this connective tissue, they read as data rather than evidence.
Pattern 3: Honest treatment of risks. The risk factors section in a weak IM is generic boilerplate. The risk factors section in a strong IM is specific to the deal and acknowledges the issues a sophisticated buyer would raise anyway. The advantage of doing this is twofold. First, it builds credibility. Second, it lets the IM control the framing of each risk rather than ceding that framing to the buyer’s diligence. A risk that the advisor names and contextualizes is a risk on the advisor’s terms. A risk the buyer discovers feels different.
Mistakes that AI tools cannot fix
There are categories of IM weakness that no amount of AI tooling will repair. They are worth naming because they explain why “use AI to write your IMs” is not a complete answer.
The first is poor data. If the underlying financials are messy, the management accounts are unreliable, and the data room is incomplete, the IM will be limited no matter how it is produced. Every hour spent cleaning the source data has higher leverage than an hour spent revising prose.
The second is weak positioning. If the equity story has not been thought through before the IM is drafted, the IM will read as confused regardless of how well it is written. The discipline of forcing the team to articulate the dominant narrative archetype before drafting is one of the highest-leverage interventions a managing partner can make.
The third is misaligned incentives. If the analyst drafting the IM has not been in the room when the equity story was decided, the IM will reflect the analyst’s interpretation rather than the partner’s intent. This is a workflow problem, not a writing problem, and it is one of the things AI can actually help with, the AI does not lose context between meetings the way a junior analyst does.
What I would do if I were drafting an IM today
If I were an analyst or associate drafting an IM in 2026, my workflow would look something like this.
Day one: read every document in the data room. Note the figures that don’t reconcile across documents. Have the conversation with the senior associate about which version of the truth the IM is going to present. Write the one-paragraph equity story.
Day two: use an AI tool to extract a clean first draft of every section based on the data room, with full source citations. Run it against the data room a second time to flag inconsistencies.
Day three: revise the draft section by section, with the equity story paragraph open in front of you the entire time. Every paragraph that does not serve the story gets cut or rewritten. Every paragraph that serves the story but doesn’t make the case clearly gets sharpened.
Day four: senior review. Most of the review time should go to the executive summary, the equity story articulation, and the financial section’s connection to the thesis. The mechanical sections should already be clean.
Day five: founder review and final revisions.
That is a five-day process. Most boutiques today take three to six weeks for the same deliverable. The difference is not that the AI is doing the analyst’s job. The difference is that the analyst’s job stops being mechanical extraction and starts being editorial judgment. Which, not coincidentally, is the work that actually develops the analyst into a senior associate.