Agency revenue forecasting: Models & metrics

Agency revenue forecasting means estimating what you will bill in the coming months from contracted retainers, projects in progress and weighted pipeline. Ask most owners for next quarter's number and you get either the best-case total or a shrug. Neither helps you decide when to hire, how much cash to hold or whether to push new business.

Key Takeaways:

• Bottom-up forecasting (building from individual contracts and deals) suits most agencies, because each client is a meaningful share of revenue
• Pipeline-weighted forecasting reduces optimism bias by discounting deals based on actual close probability
• The four metrics that matter most: MRR, pipeline velocity, win rate, and revenue churn
• Scenario planning with best, expected, and worst cases prepares you for volatility without creating paralysis
• Forecasts should be updated weekly and reviewed against actuals monthly to improve accuracy over time

This guide walks through the forecasting models that work best for agencies, the metrics you need to track, and how to build scenario plans that prepare you for whatever happens.

Why revenue forecasting matters for agencies

The cost of flying blind

Without a reliable revenue forecast, every major business decision becomes a gamble:

• Hiring: Do you have enough future revenue to support a new team member, or will you be scrambling to fill their time?
• Capacity planning: Will your team be overloaded next month or sitting idle?
• Cash flow: Can you afford that new tool, office space, or marketing campaign?
• Pricing: Are you leaving money on the table or pricing yourself out of deals?
• Growth investment: Is now the right time to invest in business development, or should you be preserving cash?

Agencies with accurate forecasts make these decisions proactively. Agencies without them react to problems after the damage is done.

The agency forecasting challenge

Forecasting is harder for agencies than for SaaS companies or product businesses because agency revenue is a mix of:

• Recurring retainer income (relatively predictable)
• Project-based income (lumpy and harder to predict)
• Ad hoc and overage work (nearly impossible to predict)

The key is using different forecasting approaches for each revenue type, then combining them into a unified view.

Forecasting models for agencies

Bottom-up forecasting

Bottom-up forecasting builds your revenue projection from individual contracts, deals and client relationships. It suits most agencies, especially smaller ones, because each client is a significant share of total revenue and you know each relationship well.

How to build a bottom-up forecast:

1. List all active contracts with their monthly or project values and end dates
2. Estimate renewal probability for each contract approaching renewal, based on that client's health and your own renewal history
3. Add pipeline opportunities weighted by their close probability
4. Include known upsell opportunities from existing client conversations
5. Sum everything by month for a 6-12 month forward view

Example bottom-up forecast for a single month:

The weighted total ($47,500) is your realistic forecast. The unweighted total ($71,000) is the optimistic ceiling.

Top-down forecasting

Top-down forecasting starts with historical trends and market data, then adjusts for known changes. It works better for larger agencies where individual deals have less impact on totals.

How to apply top-down forecasting:

1. Take your trailing 12-month revenue
2. Calculate your month-over-month growth rate
3. Project that growth rate forward, adjusting for known factors (seasonality, market conditions, capacity constraints)
4. Validate against your capacity ceiling (you cannot forecast revenue your team cannot deliver)

Top-down is most useful as a sanity check against your bottom-up forecast. If your bottom-up forecast projects 50% growth but your historical trend is 15%, something is off and you need to examine your assumptions.

Pipeline-weighted forecasting

Pipeline-weighted forecasting is the most practical method for agencies with an active sales pipeline in their CRM. It assigns a probability to every deal based on its current stage, which removes much of the optimism from a raw pipeline total.

Starting pipeline stage weights (replace with your own data):

Adjusting for your actual win rates: These generic weights are starting points. After 6-12 months of tracking, replace them with your actual conversion rates at each stage. If your data shows that 60% of proposals you send ultimately close (not 35%), adjust accordingly. Your agency sales process data is the best source for these adjustments.

The critical discipline: Update pipeline stages in real time. A forecast based on stale pipeline data is worse than no forecast at all, because it creates false confidence. Require your sales team to update deal stages within 24 hours of any status change.

The four metrics that drive revenue predictability

1. Monthly recurring revenue (MRR)

MRR is the total value of your recurring contracts (retainers, subscriptions, ongoing engagements) normalized to a monthly figure.

MRR = Sum of all recurring monthly contract values

Why it matters: MRR is your revenue floor, the minimum you can expect each month assuming no cancellations. The higher the share of revenue that is recurring, the easier forecasting gets, because most of each month's revenue is already contracted.

Track MRR movement monthly:

• New MRR: Revenue from new retainer clients
• Expansion MRR: Increased revenue from existing clients (upsells, scope increases)
• Contraction MRR: Decreased revenue from existing clients (scope reductions, downgrades)
• Churned MRR: Revenue lost from cancelled clients

Net MRR change = New + Expansion - Contraction - Churned. If this number is consistently positive, your base revenue is growing. If negative, you are on a treadmill.

2. Pipeline velocity

Pipeline velocity measures how quickly revenue moves through your sales pipeline.

Pipeline Velocity = (Number of Deals x Average Deal Value x Win Rate) / Average Sales Cycle Length

Example: If you have 20 deals in your pipeline with an average value of $15,000, a 30% win rate, and a 45-day average sales cycle:

Pipeline Velocity = (20 x $15,000 x 0.30) / 45 = $2,000 per day

This tells you that your pipeline is generating roughly $2,000 in expected revenue per day, or about $60,000 per month. If your revenue target is $100,000 per month, you know your pipeline needs to be roughly 67% larger or your win rate needs to improve.

3. Win rate

Win rate is the percentage of proposals or qualified opportunities that convert to signed contracts.

Win Rate = Deals Won / Total Proposals Sent x 100

Win rates differ sharply by lead source. Referrals usually close at the highest rate, inbound leads in the middle and cold outbound at the lowest, so a single blended rate hides a lot. Track win rate by lead source, deal size and service type, and use the segment rate when you weight each deal.

For more on optimizing your conversion metrics, see our guide to agency KPIs and metrics.

4. Revenue churn rate

Revenue churn measures the percentage of recurring revenue lost each month due to client cancellations or scope reductions.

Monthly Revenue Churn = Lost MRR / Starting MRR x 100

Rules of thumb:

• Under 3% a month (roughly 30% a year) is a stable retainer base
• 3-5% a month needs attention
• Above 5% a month means you are replacing a large part of your client base every year

High churn undermines every other forecasting effort. If you are losing 5% of your MRR every month, you need to replace that revenue just to stay flat. On $100,000 of MRR, cutting monthly churn from 5% to 3% keeps an extra $2,000 of MRR every month, which compounds over a year.

Scenario planning: Preparing for uncertainty

Even the best forecast is a single point estimate. Scenario planning acknowledges uncertainty by modeling multiple outcomes.

The three-scenario framework

Build three versions of your revenue forecast:

Best Case (80th percentile):

• All confirmed revenue closes as expected
• Pipeline deals close at the high end of probability ranges
• Win rate runs above your 12-month average
• No unexpected client churn

Expected Case (50th percentile):

• Confirmed revenue closes as expected
• Pipeline deals close at average probability weights
• Win rate matches your 12-month average
• Normal churn rate applies

Worst Case (20th percentile):

• One or two confirmed renewals do not materialize
• Pipeline deals close at below-average rates
• Win rate drops 20-30% below average
• One major client reduces scope or churns unexpectedly

Using scenarios for decision-making

The power of scenario planning is in how you use it for business decisions:

• Hiring decisions: Only hire if the worst-case scenario still supports the new salary. If you need the best case to justify the hire, wait.
• Investment decisions: Fund investments (marketing, tools, office space) based on the expected case, not the best case.
• Cash reserves: Size your cash reserve to survive the worst case for at least 2-3 months.
• Sales targets: Set stretch goals based on the best case, but set quotas based on the expected case.

Updating scenarios monthly

Review all three scenarios monthly. As pipeline deals close or fall through, and as new information emerges, adjust the scenarios. The gap between your best and worst case should narrow as you get closer to the forecast period, because more information is available.

Building your forecasting process

Weekly forecast updates (15-20 minutes)

• Update pipeline deal stages and values in your CRM
• Note any new opportunities or lost deals
• Flag any at-risk retainer clients
• Adjust near-term (next 4 weeks) projections based on current information

Monthly forecast reviews (60-90 minutes)

• Compare last month's forecast to actual results
• Calculate forecast accuracy (actual / forecast x 100)
• Identify which assumptions were wrong and why
• Update all three scenarios for the next 6 months
• Review MRR movement, win rate, and churn trends
• Adjust pipeline stage probabilities if actuals differ from assumed weights

Quarterly strategic reviews (2-3 hours)

• Analyze year-to-date forecast accuracy trends
• Review whether revenue mix (retainer vs. project) is shifting
• Assess whether current pipeline is sufficient to hit annual targets
• Make capacity and hiring decisions based on 6-month outlook
• Set or adjust annual revenue targets

Common forecasting mistakes

Confusing pipeline value with forecast. Your pipeline total is not your forecast. A $500,000 pipeline with a 30% win rate is a $150,000 forecast, not a $500,000 one. Always weight by probability.

Ignoring churn in projections. If you forecast $50,000 in new MRR but ignore that you typically lose $15,000 in churned MRR per month, your net growth forecast is $35,000, not $50,000.

Using the same probability for all deals. A $5,000 retainer proposal to a warm referral and a $50,000 project proposal to a cold lead should not have the same close probability. Differentiate by deal size, lead source, and relationship strength.

Forecasting beyond your capacity. If your team can deliver $200,000 per month at maximum utilization, forecasting $300,000 is not optimistic, it is impossible. Always cap your forecast at your delivery capacity.

Updating too infrequently. A forecast updated monthly is a history report, not a planning tool. Weekly updates keep forecasts actionable. The agencies with the best forecast accuracy are the ones that treat it as a living document, not a quarterly exercise.

Getting started

If you do not currently forecast revenue, start simple. Build a bottom-up forecast of your next 3 months using the model described above. Track your four key metrics (MRR, pipeline velocity, win rate, churn) for one quarter. Compare your forecast to actuals each month and refine your assumptions. Within two quarters, you will have a forecasting system that gives you real confidence in your numbers.

Frequently asked questions

How accurate should an agency revenue forecast be?

Accurate enough to make hiring and cash decisions. Measure it each month (actual divided by forecast) and aim to narrow the gap over time. Expect wide misses in the first few quarters, when pipeline data is patchy and you have little history. Even established sales teams miss: in Xactly's 2024 Sales Forecasting Benchmark Report, 4 in 5 sales and finance leaders had missed a quarterly forecast in the past year, and 52% of sales leaders said their forecasts were off by 10% or more. Accuracy improves as you calibrate stage weights to your own results.

What pipeline weighting should agencies use?

Start with stage weights like those in the table above (for example 30-40% at proposal sent, 70-80% at verbal agreement), then replace them with your own historical close rates by stage after 6 to 12 months.

How far out should an agency forecast revenue?

Forecast a rolling 90 days at high confidence and 6 to 12 months at lower confidence. Beyond 12 months, the variables compound and predictions become guesses. Update the rolling 90-day forecast weekly and the longer view monthly.

Should retainer and project revenue be forecast separately?

Yes. Retainers are MRR with churn risk; projects are pipeline conversion. Forecast each on its own model (retainer churn rate plus expansion vs. project pipeline weighted by stage) and combine for the total. Mixing them hides whichever side is breaking.

What metrics drive forecast accuracy improvement over time?

Track three: forecast variance (predicted vs actual by month), pipeline velocity (days from create to close by stage), and stage-to-close rate. Reviewing these monthly catches the assumptions that need recalibrating before they distort the next quarter.

AgencyPro keeps recurring billing and invoices on every plan and a leads pipeline on the Pro plan, so contracted and pipeline revenue live in one place. Book a demo to see it.