What Mismatches Between a Pitch Deck, Model and Data Room Cause Investors to Lose Confidence?

TL;DR: The mismatches that cost the most trust are the simple ones: a revenue figure, a raise amount, a growth rate or a return that reads one way in the deck and another way in the model or data room.

• Why do the deck, model and data room drift apart? They drift apart because founders build each piece at a different time, often with different people.
• Where do mismatches hide? Mismatches hide in five places.
• How do institutional reviewers find mismatches? Reviewers find mismatches by rebuilding your numbers from the source files.

The mismatches that cost the most trust are the simple ones: a revenue figure, a raise amount, a growth rate or a return that reads one way in the deck and another way in the model or data room. Investors can live with a weak number. They struggle with a number that changes. One number that moves between documents costs more trust than a weak number that holds.

The stakes are highest in due diligence. In a survey of 885 venture capitalists, Gompers, Gornall, Kaplan and Strebulaev (How Do Venture Capitalists Make Decisions?, NBER Working Paper 22587, 2016) found that only about a third of startups that reach due diligence are offered a term sheet. Due diligence is where reviewers read your documents side by side, and it is where mismatches show up.

Why do the deck, model and data room drift apart?

They drift apart because founders build each piece at a different time, often with different people. The deck gets built first, for early meetings. The model keeps changing as assumptions move. The data room fills up last, often in a rush. Each piece looks fine alone. The problem shows up when a reviewer reads all three side by side, which is exactly what institutional reviewers do.

Where do mismatches hide?

Mismatches hide in five places.

1. Headline numbers. The cover slide says a $20M raise. The model's financing tab says $21.5M. The sources and uses in the data room says $22M. Same line item, three answers.

2. Growth and return figures. The deck shows next year's revenue growth or a projected return. The model output tab shows a different figure because assumptions changed after the deck was built. This is the most common one, and reviewers check it first.

3. Use of funds. The deck's use-of-funds chart says 40% to sales hiring. The model's hiring plan only spends a quarter of the raise on sales. The reviewer now has to guess which plan you will actually run.

4. Cap table and terms. The deck shows founder ownership after the round. The cap table file shows a different number once the option pool and notes convert. The legal documents win, so the deck is wrong.

5. History and old files. The traction chart in the deck does not tie to the financial statements in the data room. Or the data room still holds last quarter's model and an old investor deck with different numbers. A reviewer who opens the wrong file sees a different company.

Here is how reviewers catch each one, and the fix.

How do institutional reviewers find mismatches?

Reviewers find mismatches by rebuilding your numbers from the source files. A typical diligence team will:

• Trace every deck number back to a model cell or a data room file.
• Tie the model's starting point to your last reviewed or audited financial statements.
• Rebuild the cap table from the legal documents, with every note, SAFE and option converted.
• Open every file in the data room, including the old ones you forgot were there.
• Compare your answers on calls with what the documents say.

None of this is a trick. It is the job. Assume every number will be checked against every other place it appears.

Why do investors treat mismatches as a character test?

A mismatch changes what the reviewer is evaluating. Before the mismatch, they are judging the deal. After it, they are judging you.

They start asking: Does this team know its own numbers? Who checks the work? If the numbers drift on documents the team fully controls, what will investor reports look like after the money is wired?

The reviewer also cannot tell which number is true. So they stop, write a follow-up list and wait. Every follow-up cycle adds time and pulls attention away from your deal. With only about a third of diligence-stage deals reaching a term sheet in the Gompers study, a deal that needs a long question list starts at a disadvantage.

A clean package sends the opposite signal. It tells the reviewer your team runs a controlled process. That same discipline is what they want to see in reporting, budgets and board updates after they invest.

What changes for a real estate sponsor?

A real estate sponsor faces the same five risks, with different documents in play.

• Waterfall: A waterfall is the order in which cash flows to investors and to the sponsor. The preferred return, catch-up and promote must read the same in the deck, the model and the LPA or operating agreement.
• Sources and uses: The equity raise, the loan amount and the closing costs must add up the same way in every file.
• Debt terms: The rate, term and leverage in the model must match the lender's term sheet.
• Track record: Past deals should show projected and realized returns side by side, and they should match the track record file in the data room.

For a sponsor, the legal documents always control. If the deck and the LPA disagree, the LPA wins and the deck is wrong.

How do you build a one-page tie-out sheet?

The fix is a one-page tie-out sheet. List every number that appears in more than one document. Name the one document that owns it. Then check every other place it shows up.

The ownership rule runs one way. Legal documents set the terms. The model turns those terms and your history into projections. The deck summarizes the model. The deck is always the last thing you update. If a number in the deck cannot be traced to a model cell or a data room file, take it out.

Two more checks belong on the sheet. First, confirm the model's starting point ties to your last reviewed or audited period. IRC Partners covers that step in How to Bridge Historical Financials to the Forecast. Second, search the data room for old versions. Remove them or label them superseded with a date.

Who should sign off before anything goes out?

One person owns the tie-out sheet. Usually that is the CFO. In a smaller company it is the founder. If everyone owns it, no one checks it.

Use three rules:

1. Run the sheet before every investor touch. That means before first outreach, before the data room opens and before any updated file goes out mid-raise.
2. Counsel signs the terms rows. Your lawyer confirms the deck and model match the legal documents.
3. Date the sheet. Keep each dated version. If a reviewer asks why a number changed, you can show when and why.

What does a tie-out catch? A made-up example

This is a made-up example. It is not an IRC Partners client and the numbers are for illustration only.

A software company is about to open its data room for a $20M round. The CFO runs the tie-out sheet the week before launch and finds three mismatches:

1. Growth rate: The deck says 70% growth next year. The model says 58%, after a pricing change. The deck gets updated to match the model.
2. Ownership: The deck shows the founders at 61% after the round. The cap table, with two convertible notes converted, shows 56%. The deck gets updated to match the cap table.
3. Old files: The data room holds three versions of the model. Two get moved to a dated archive.

Each fix takes less than a day. Found during diligence, each one would have turned into a follow-up list and a question about who checks the work.

What should you do if an investor finds a mismatch first?

Fix it fast and say so plainly. Send the corrected file the same day with a short note on what changed and which document now controls. A quick, plain correction keeps more trust than letting the investor find a second one.

Catch it before they do

Investors read your deck, model and data room as one system. One loose number makes them doubt the rest.

The Capital Raise Pre-Flight is IRC Partners' fixed-fee readiness audit. It reads your deck, model and data room side by side across 12 gates, including Pitch Deck, Financial Model, Data Room and Decision Friction, and flags every mismatch before an investor finds it.

Frequently Asked Questions

Which document should control a number when the deck and model disagree?

The model controls projected numbers. The financial statements control history. The term sheet or LPA controls terms. Update the deck last, to match those sources.

How often should I tie out my documents during a raise?

Before every investor touch. A small model change can create a deck mismatch the same day.

What is the most common mismatch investors find?

A growth rate or return in the deck that no longer matches the model, because assumptions changed after the deck was built.

Should old versions stay in the data room?

No. Keep one current version of each core file. Move old files to a dated archive or remove them so a reviewer cannot open the wrong one.

What should I do if an investor finds a mismatch first?

Fix it the same day and send the corrected file with a short note naming the controlling document. Speed and plain answers protect trust.

Who should own the tie-out process?

One named person, usually the CFO or founder, with counsel signing off on terms.

Continue reading this series:

• How to Build a Data Room That Holds Up to Institutional Diligence
• How to Bridge Historical Financials to the Forecast
• IC Deck vs Investor Presentation

IRC Partners advises founders and sponsors raising $5M to $250M of institutional capital. We take 7 strategic partners per quarter. To talk with our team about your raise, start HERE.