Quick answer: Match the forecasting method to how the business sells. Forecast project and industrial work from backlog conversion, services from a stage-weighted pipeline, subscriptions and managed services from recurring revenue and churn, and consumer products from cohorts and repeat orders. Buyers accept signed orders, CRM exports, contract schedules and order-level history as proof. A forecast built on the wrong method, or on summaries instead of those records, is replaced by the buyer’s own.
Key points
- The method follows the way the business wins revenue. Backlog fits work booked before delivery, pipeline fits work still to be sold, and order history fits products that customers buy again.
- Every method splits the forecast into revenue already secured and revenue still to be won, and buyers give credit mainly for the first part.
- Conversion, win, churn and repeat rates must come from the company’s own history, measured the same way each period. A rate chosen to fit the target is the first assumption a buyer replaces.
- A buyer re-runs the method from source-system exports, so a forecast that cannot be rebuilt from the order book, CRM, billing or sales data earns little credit.
Choose the method that matches how the business wins revenue
Forecast with the method that follows how customers commit. Work booked before delivery points to backlog, revenue that starts with a proposal points to the pipeline, contracts that renew point to the contract base and its churn, and products that customers choose again point to cohorts. A buyer rebuilds the forecast from the company’s records, so the method must also match the records the business already keeps.
The table sets out each method, the evidence behind it and where buyers push back. The guide to driver-based forecasting shows how any of them turns into units, price and cost.
| Business model | Method | Evidence buyers accept | Where buyers push back |
|---|---|---|---|
| Project and industrial work | Backlog conversion plus book-and-ship revenue | Order book export with dates, contracts, cancellation terms | Orders a customer can cancel at no cost, ship dates that keep slipping |
| Professional and technical services | Stage-weighted pipeline converted by start date | CRM export with stage history, record of proposals won and lost, staffing plan | Stale opportunities, win rates from the best year, revenue that needs more people than are on staff |
| Subscriptions and managed services | Recurring revenue roll-forward with churn | Contract schedule, billing by customer by month, churn history | Churn measured on a flattering base, price increases counted as expansion |
| Consumer and branded products | Cohorts and repeat orders | Order data by customer or account, retailer sell-through, promotion calendar | Repeat rates from the best cohort, new-customer volume with no spending behind it |
Each method divides next year’s revenue into the part already secured and the part still to be won. Buyers give credit for the first and test the second, as the guide to how buyers test projections describes. The guide to recurring and project revenue explains why the multiple depends on which kind is secured.
Backlog conversion for project and industrial businesses
Backlog is the value of signed orders not yet delivered. A backlog forecast converts the backlog on hand into revenue by delivery date, then adds revenue from orders that will be booked and shipped inside the year, which can only come from history. Backlog earns credit by its terms. An order the customer can cancel for free is closer to pipeline, a date set by the customer’s release schedule is firmer than one set by the plant’s capacity, and an order priced before a cost increase may carry less margin than history suggests. The valuation guide for metal fabricators covers backlog and price pass-through.
Illustrative example: Backlog plus book-and-ship revenue for a fabricator
A fabricator starts the year with $6,000,000 of signed orders, of which $5,400,000 is scheduled to ship in the year. The rest is long-lead work for the following year. Last year, $6,000,000 of revenue came from orders booked and shipped within the same year, and the forecast holds that figure.
| Component | Amount |
|---|---|
| Backlog scheduled to ship in the year | $5,400,000 |
| Book-and-ship revenue, held at last year’s level | $6,000,000 |
| Forecast revenue | $11,400,000 |
Signed orders cover 47 percent of the forecast, and the other 53 percent depends on orders not yet received. If $800,000 of the scheduled backlog could be cancelled at no cost, the secured amount falls to $4,600,000, about 40 percent. The numbers are invented to show the mechanics and are not benchmarks.
Pipeline weighting for services firms
A services firm has little backlog, because engagements start soon after signing. Revenue begins as a proposal, so the forecast weights each open opportunity by its chance of closing and spreads the value over the months in which the work is delivered. Weights come from the firm’s own win rates by stage, since a rate from its best year fails the first buyer question. Stages need written definitions, stale opportunities come out before weighting, and start dates must agree with the hiring and utilization assumptions in the model.
Illustrative example: A stage-weighted pipeline
A consulting firm’s CRM export shows the open proposals below. The win rates are placeholders for the firm’s own history by stage.
| Stage | Open value | Placeholder win rate | Weighted value |
|---|---|---|---|
| Proposal sent | $2,000,000 | 20% | $400,000 |
| Verbal award, contract not signed | $1,000,000 | 70% | $700,000 |
| Contract in legal review | $500,000 | 90% | $450,000 |
| Total | $3,500,000 | $1,550,000 |
If $600,000 of the proposals sent has had no activity for months, removing it cuts the weighted total to $1,430,000. The weighted value is contract value and not revenue, because a project starting in the last quarter adds only a fraction of its value to the year. The values and rates are invented to show the mechanics and are not benchmarks.
Recurring revenue and churn for subscriptions and managed services
Recurring revenue comes from contracts that continue without a new sale each period, such as managed service agreements and subscriptions. The forecast starts from the recurring revenue in place at the beginning of the year and rolls it forward for losses, downgrades, expansion and new customers. Gross revenue retention is the share of starting revenue still billed after losses and downgrades, and net revenue retention adds back expansion and price increases. The guide to KPIs and operating metrics covers how to define and report both.
Definitions decide whether the numbers are believed. A buyer asks what counts as recurring, since monthly billing that a customer can stop on short notice behaves differently from a committed term, and it recalculates churn from billing records by customer by month. A churn rate measured on a base that includes recent customers flatters the result, so the base is the customers in place at the start of the period. Revenue also lags the base, because revenue recognized in a year is closer to the average of the opening and closing base than to the closing figure. The valuation guide for IT services and managed service providers covers contract terms and the mix of recurring and project work.
Illustrative example: A recurring revenue roll-forward
A managed services provider starts the year with $4,000,000 of annual recurring revenue.
| Roll-forward | Annual recurring revenue |
|---|---|
| Start of year | $4,000,000 |
| Customers lost | ($400,000) |
| Downgrades | ($100,000) |
| Expansion and price increases | $300,000 |
| New customers | $600,000 |
| End of year | $4,400,000 |
Gross revenue retention is 87.5 percent ($3,500,000 divided by $4,000,000) and net revenue retention is 95 percent ($3,800,000 divided by $4,000,000). The existing base shrinks by $200,000, so all of the growth rests on the $600,000 from new customers, which is the line a buyer tests. If changes arrive evenly through the year, revenue is close to the $4,200,000 average of the opening and closing figures, not the $4,400,000 year-end base. The numbers are invented to show the mechanics and are not benchmarks.
Cohorts and repeat orders for consumer products
A consumer product company sells to people or accounts that decide again with each order, so the forecast models repeat behavior. A cohort groups customers by the period of their first order and follows each group’s revenue afterward, and stacking the cohorts shows how much of a year’s revenue comes from customers already acquired. For products sold through retailers the cohort is usually the retail account, and the evidence is reorder history plus retailer sell-through reports, not the first stocking order. The repeat curve must come from several cohorts and not the best one. The new-customer forecast must also agree with the marketing, promotion and trade spending in the model, since volume bought with discounts repeats differently.
Illustrative example: Stacking cohorts into next year’s revenue
A direct-to-consumer brand tracks customers by the year of their first order. A new customer spends $60 in the first year. A cohort’s revenue in its second year is 45 percent of its first-year revenue, and in each later year it is 80 percent of the year before. These rates are placeholders for the company’s own history.
| Cohort (first order in) | New customers | Year 3 actual | Year 4 forecast |
|---|---|---|---|
| Year 1 | 10,000 | $216,000 | $172,800 |
| Year 2 | 12,000 | $324,000 | $259,200 |
| Year 3 | 12,000 | $720,000 | $324,000 |
| Year 4 (new customers held at 12,000) | 12,000 | $720,000 | |
| Total | $1,260,000 | $1,476,000 |
Earlier cohorts supply $756,000 of the Year 4 forecast, about 51 percent. The other 49 percent depends on winning 12,000 new customers at the same cost as before. The figures are invented to show the mechanics and are not benchmarks.
Blending methods and tying the total to the statements
Few businesses fit one method. A fabricator with a maintenance base, a distributor with project customers or a food company selling at retail and online forecasts each revenue stream with its own method and adds the streams. Four steps keep the total credible.
- Split revenue into streams by how each is won, using the same categories in the history and the forecast.
- Back-test each method. Run it from the opening position of the last full year and compare the result with the revenue the stream produced. A method that would have missed last year needs better inputs or a different method.
- Label the secured part of each stream and keep the exports behind it, such as the order book, CRM, billing and sales data with dates, so a buyer’s analysts can re-run them. The guide to base, upside and downside cases shows where the unsecured part belongs.
- Tie the total to revenue in the financial statements for each historical period, and to the CIM and management presentation for the forecast.
A buyer’s case usually changes the rates these methods rely on, as the guide to management case versus buyer case explains. The complete guide to building financial projections shows how revenue connects to margins and cash.
Frequently asked questions
Which revenue forecasting method should a seller use?
The one that matches how customers commit. Backlog conversion suits work booked before delivery, a stage-weighted pipeline suits services sold by proposal, a recurring revenue roll-forward suits contracts that renew, and cohorts suit products customers buy again. Most businesses blend methods, one for each revenue stream.
How do buyers treat backlog in a forecast?
They credit signed orders with firm delivery dates and test the rest. A buyer reconciles backlog to contracts, checks whether the customer can cancel or defer without cost, and compares scheduled dates with past slippage. Backlog that is easy to cancel is treated more like pipeline.
What is the difference between gross and net revenue retention?
Gross revenue retention is the share of starting recurring revenue still billed after customer losses and downgrades, so it cannot exceed the whole base. Net revenue retention adds back expansion and price increases from the same customers and can exceed it. Buyers ask for both, calculated from billing records.
What is a cohort analysis in a revenue forecast?
A cohort analysis groups customers or accounts by the period of their first order and tracks each group’s revenue in later periods. Adding the cohorts shows how much of a year’s revenue comes from customers already acquired and how much depends on winning new ones.
Preparing a company for sale or a buyer's diligence?
The Impetus Group offers fractional investment banking support, including earnings normalization, financial projections, CIM preparation and help navigating buyer diligence, with an engagement scoped to what the business needs.
Related guides
- Driver-Based Forecasting for a Business Sale: Build a Model Buyers Can Test: Driver-based forecasting builds a sale projection from units, price, utilization and retention. See driver trees for a manufacturer and a services firm.
- KPIs and Operating Metrics Buyers Ask For, by Business Model: The KPIs buyers request in diligence for manufacturers, distributors, services, recurring revenue and consumer brands, with definitions and reporting rules.
- How Buyers Test Your Projections in Diligence, Test by Test: How buyers evaluate financial projections in a sale: the bridge from LTM, forecast accuracy, backlog and pipeline checks, customer calls, capacity and sensitivities.
- Customer Due Diligence: Contract Review, Retention Analysis and Customer Calls: How buyers run customer due diligence: contract terms, retention by cohort, pricing history, credits and disputes, and the customer calls that test revenue.
- Base, Upside and Downside Cases: Scenario Analysis for Sale Projections: Scenario analysis for a sale: what moves between base, upside and downside cases, how to label assumptions and how lenders and buyers use the downside.
- Management Case vs Buyer Case: How Buyers Adjust Your Projections: How a buyer turns a management case into its own case: the adjustments it makes, a reconciliation of the gap, and how sellers defend assumptions with evidence.
This guide is general education. It is not a valuation opinion, appraisal, investment advice, or legal or tax advice. How these guides are made.