Forecast Revenue from Project Leads Using Your Pipeline

The most useful number in an A&E firm's monthly partner meeting is the answer to one question. Based on what we have in the pipeline today, how much revenue are we likely to book in the next 90 days. Most firms cannot answer it.

The fix is a probability weighted revenue forecast that reads from the pipeline. This guide walks the model, the math, and the accuracy improvement loop that makes it sharper every quarter.

The forecasting model in one paragraph

For each lead in the pipeline, multiply the estimated fee by the historical win probability of its current stage. Sum across all leads. Distribute the resulting weighted value across the months when those projects would actually invoice. That is your revenue forecast.

The data inputs

The model needs five fields per lead.

• Estimated fee. Total contract value if won.
• Stage. Current pipeline stage (Inquiry, Qualified, Proposal Sent, Shortlist, Won, Lost).
• Stage probability. Historical win rate at that stage.
• Expected close date. When the contract would be signed.
• Project duration. Number of months to deliver, which determines the invoicing distribution.

Stage probabilities: get them right

The single biggest determinant of forecast accuracy is the probability you assign to each stage. Default values are a starting point. Real numbers come from your firm's history.

Defaults to start with.

• Inquiry: 5 to 10 percent.
• Qualified: 20 to 30 percent.
• Proposal Sent: 35 to 50 percent.
• Shortlist or Interview: 60 to 75 percent.
• Verbal Won (not yet signed): 90 percent.
• Won and signed: 100 percent.

After two quarters of data, replace these with your actual numbers. Most firms find their probabilities differ from defaults by 5 to 15 percentage points per stage.

A worked example

A 25 person firm forecasting the next 90 days.

Pipeline contents.

• 3 leads at Inquiry, total estimated fee 1.2M dollars.
• 4 leads at Qualified, total 1.8M.
• 5 leads at Proposal Sent, total 2.4M.
• 2 leads at Shortlist, total 0.9M.
• 1 verbal Won, 0.4M.

Apply firm specific probabilities (Inquiry 8 percent, Qualified 25 percent, Proposal 42 percent, Shortlist 68 percent, Verbal Won 90 percent).

• Inquiry weighted: 1.2M × 0.08 = 96K.
• Qualified weighted: 1.8M × 0.25 = 450K.
• Proposal weighted: 2.4M × 0.42 = 1,008K.
• Shortlist weighted: 0.9M × 0.68 = 612K.
• Verbal Won weighted: 0.4M × 0.90 = 360K.
• Total weighted pipeline: 2,526K dollars.

That is the expected total fee value of currently active pursuits. The next step is distributing it across months.

From total weighted to monthly revenue

Total weighted pipeline is interesting. Monthly revenue forecast is actionable. The mapping requires three more pieces of data per lead.

1. Expected contract sign date.
2. Project duration in months.
3. Invoicing pattern (monthly even, milestone weighted, or front loaded).

For a typical A&E firm, monthly even invoicing is the default. A 600K project signing in May with a 12 month duration produces about 50K of revenue per month from June through May of next year.

Spread each weighted lead across its expected delivery months. Sum by month. The result is a monthly revenue forecast that explicitly reflects pipeline probability.

The accuracy tracking loop

The first version of this forecast will be wrong. The point is not to be right month one. The point is to be sharper month six.

Track three numbers quarterly.

• Forecasted revenue 90 days ago vs actual revenue this quarter. Variance under 10 percent is excellent. 10 to 20 percent is good. Over 25 percent means the model needs work.
• Stage probability accuracy. For each stage, what percent of leads at that stage 90 days ago actually closed. Compare to your model probability.
• Sign date accuracy. How many leads signed when forecast vs how many slipped. If average slip is over 30 days, build slip buffer into future forecasts.

After two quarters of tuning, your forecast accuracy lands within 10 percent of actual revenue. That is good enough to make hiring, BD investment, and cash decisions on.

What this unlocks

A reliable revenue forecast changes how a firm operates.

• Hiring decisions get made on data, not vibes.
• Cash flow planning gets a real input.
• BD investment timing gets sharper.
• The principal stops being the only person who knows where the firm is going.

Combine with capacity forecasting

Revenue forecasting answers "can we book the work." Capacity forecasting answers "can we deliver it." Together, the two forecasts are the operating dashboard for an A&E firm.

Most firms run one or the other. The strongest firms run both, in the same weekly cadence, on the same pipeline data.

The five mistakes that wreck the forecast

1. Stale pipeline data. A forecast on stale data is fiction. Update weekly.
2. Optimistic stage probabilities. Use real history, not aspirational numbers.
3. No sign date discipline. If owners do not record realistic sign dates, the monthly distribution is meaningless.
4. Including dormant leads. Anything more than 60 days in stage with no activity should drop out of the active forecast.
5. Treating the forecast as a target. The forecast is a prediction. Targets and predictions are different things.

The 90 day starting plan

Set up the model with default probabilities. Run it weekly for 13 weeks. Track the variance against actual closes. Adjust the probabilities. By quarter two, the forecast is meaningfully accurate. By quarter four, it drives real decisions.

That is what a real revenue forecast buys you. Not just a number. A firm that can plan with eyes open.