A private buyer — a searcher, an independent sponsor, a family office, or an operator writing a check — does not buy your forecast. They rebuild it. If the monthly data behind the model is missing, they either cut the price or walk.
Historical tax returns still matter. They are not a model. A predictive financial model is a set of drivers, tied to source documents, that shows what the company is likely to earn after you are gone. This is the package serious private buyers ask for before they believe the number on the teaser.
What they are trying to prove
Three questions sit under every request:
- Is the earnings number real? Not the tax return. The profit a new owner can keep, after a market wage for your job and after one-time items are removed.
- Will it repeat? Customers, contracts, and a pipeline that is not your personal phone.
- What cash does the deal actually need? Working capital, maintenance capex, and debt that does not show up in a one-page P&L.
If your model cannot answer those three, it is a wish. Buyers price wishes at a discount.
The data they open first
Have this ready before the first management call. “We’ll pull it together” is how a 30-day exclusivity turns into a retrade.
1. Monthly financials, not year-end only
Trailing 36 months of monthly profit and loss, plus a current balance sheet and trailing-twelve-month view. Private buyers look for seasonality, a sudden margin change, and whether the last quarter was dressed up for a sale. Annual statements hide all three.
2. A written EBITDA bridge
Start at net income or tax-return profit. Add back, line by line: owner compensation above a replacement wage, personal expenses, one-time legal or moving costs, PPP or other non-recurring items, and related-party rent if it is off market. Every add-back needs a document. Unsupported add-backs are the fastest way to lose credibility.
3. Revenue the buyer can re-underwrite
- Revenue by customer for three years, with the top 10 named and their share of total sales
- Revenue by product, service line, or location
- New versus repeat, and contracted versus one-off
- Backlog or recurring revenue schedule, if you have one
- Churn, if customers cancel — logo count and dollars, not a vibe
A single customer above roughly 15% of revenue will be in the first email. Concentration is not fatal. Hiding it is.
4. Margins that tie to the work
Gross margin by line of business, not one blended number. Cost of goods or direct labor that moves with volume. If price increases carried last year’s growth, say so. Buyers would rather see a plain driver than a hockey stick with no cause.
5. Working capital and cash
Accounts receivable aging, inventory if you hold it, accounts payable, and a sense of normal working capital — not the low point you hit the month you stopped buying inventory. Bank statements or reconciliations for the same 12–36 months. Cash that does not tie to the books stops a deal.
6. Capex, debt, and what you actually take home
- Maintenance versus growth capital spending for three years
- A debt schedule: lender, balance, rate, maturity, personal guarantees
- Owner salary, distributions, and any related-party payments
- Leases a buyer must assume
The model they will rebuild anyway
Give them a model they can edit. Three cases — base, upside, downside — for three years. Each case should be driver-based:
- Volume, price, and mix, not “revenue grows 12%”
- Gross margin that moves when those drivers move
- Operating costs split into what stays and what flexes
- A replacement wage for the owner’s role
- Maintenance capex and a normal level of working capital
- An assumptions tab a stranger can read without calling you
Upside is allowed. It has to name the trigger: a contract already in late stage, a second crew you have priced, a price increase you have already tested. “The new owner will professionalize sales” is not a driver. It is a hope, and private buyers do not pay for hope on day one.
CO. BUY SELL PERSPECTIVE
The sellers who hold price are not the ones with the prettiest chart. They are the ones whose monthly file, add-back memo, and customer list all tell the same story. When those three disagree, the buyer believes the worst one.
What triggers a retrade
- Add-backs with no invoice, no payroll record, and no explanation.
- A forecast that assumes you keep selling after close, with no one else in the pipeline.
- Customer concentration discovered in diligence instead of in the model.
- Working capital handed over below the level the business actually runs on.
- Last year’s growth that was one job, one price hike, or one customer.
How to package it
One folder. Clear names. No scavenger hunt.
- Monthly P&Ls, 36 months, plus year-to-date and a TTM column.
- Balance sheet, AR aging, and debt schedule.
- EBITDA bridge with a note and a backup file for each add-back.
- Customer and revenue-by-line export.
- A three-case driver model and a one-page assumptions memo.
- Bank reconciliations for the trailing twelve months.
You do not need a 40-tab model or a full quality-of-earnings report to start. You need the source data a buyer’s analyst would pull on a Sunday night. If you cannot produce it in a week, that is the work to do before you ask anyone to pay a premium.
The bottom line
Private buyers demand proof that earnings are real, repeatable, and convertible to cash without you in every deal. Give them monthly numbers, a documented bridge, customer-level revenue, working capital, and a forecast tied to drivers. The model is the argument. The data is the evidence. Price follows the evidence.
Put a real number under the model.
A forecast is not a valuation. If you want the range a buyer can test — and the gaps that move it — start with a professional valuation or benchmark sale readiness first.
