Renewal Forecasting vs New Business Forecasting: Why One Model Cannot Do Both

What Is the Difference Between Renewal Forecasting and New Business Forecasting?

New business forecasting predicts whether revenue that does not exist yet will be won, and renewal forecasting predicts whether revenue you already have will be kept. Those are opposite problems, and running them through one model produces a number that is wrong about both.

A new business forecast starts from a pipeline of opportunities that may or may not close. The inputs are behavioral: stage movement, close date changes, deal value, engagement from the buying group. The base case is zero, and every dollar has to be earned.

A renewal forecast starts from a contract with a known date and a known value. The base case is the full amount, and the forecasting question is what fraction leaks away through churn or contraction, plus what gets added through expansion. The inputs are product usage, support engagement and relationship health, none of which appear in a sales pipeline.

Why Does Blending Them Hide Risk?

Because the two halves fail at different times and for different reasons, and a combined number lets a strong half mask a collapsing one. A quarter where new business runs hot and renewals quietly slip looks fine on the roll-up and is a serious problem on the balance sheet.

The failure modes do not overlap. New business misses when a competitor creates pricing pressure and average deal size falls, or when capital tightens and win rates decline, or when uncertainty stretches the cycle from qualified to closed. Renewals miss when a champion leaves, when usage drops, or when a customer never adopted the product in the first place.

Reporting matters here too. Accuracy expectations differ. A carefully built manual forecast on new and expansion business usually lands around 90 percent, and ORM targets 95 percent on that same new and expansion business without manual adjustments, holding from day 1 to day 90. Renewals need their own scorecard, because a blended accuracy figure tells you nothing about which motion is working.

What Signals Predict a Renewal?

Support case volume, read at both extremes rather than as a simple more-is-worse metric. This is one of the more counterintuitive patterns in retention data and it is worth building into a model.

Customers with no support cases at all are at risk of churn. Silence is usually not satisfaction. It generally means nobody is using the product deeply enough to run into anything. Customers with seven or more cases in the last year are also at risk, because that volume points to real friction.

The healthy band sits in the middle. Three to five tickets, typically tier two or three rather than severe, correlates with customers who stay. Those customers are engaged, they are getting support, and they are usually happy. A renewal model that treats ticket volume as a linear risk score will flag exactly the wrong accounts.

What Signals Predict a New Business Deal?

Change, and specifically change in stage, close date or amount. Notes and logged calls feel like progress and rarely predict revenue.

The strongest slippage signal is a rep moving a close date. When a deal slips from one quarter into the next, it becomes less likely to close, even when it sits in commit. That single field change carries more information than most qualification frameworks.

The earliest signal is the absence of any signal. No stage movement, no data changing, no notes. From the seller's side, the equivalent is a buyer who stops returning email, stops taking calls, and stops responding to texts. Most ORM customers apply a 12-month rule to opportunity aging on exactly this basis, and 10 percent or more of pipeline across those customers has not been touched in that window.

Should Expansion Sit With Renewals or New Business?

With new business, because expansion behaves like a sales motion even though it happens inside an existing account. Expansion requires a buying decision, a budget and usually a business case, which is why forecast accuracy is measured on new and expansion business together.

The practical structure splits the revenue base into three forecasts rather than two. New logo acquisition, expansion within the installed base, and renewal of existing contracts. Each gets its own model, its own owner and its own accuracy measurement.

That split also maps cleanly onto a retention waterfall. Running the month-by-month sequence of beginning ARR, churned customer ARR, churned product ARR, product decrease ARR, new customer ARR, new product ARR, increased product ARR and ending ARR gives you a reconciling view where beginning ARR always equals the prior month's ending ARR. Both net revenue retention and gross revenue retention fall out of that same chart, which is why the waterfall is worth the setup effort.

How Do You Build a Renewal Forecast That Beats Default Optimism?

Start from the contract list, apply usage and support signals to score risk, and force a called outcome earlier than ninety days out. The visibility advantage of renewals is that you know the dates a year ahead. Most teams throw that advantage away by starting the conversation at the last minute.

Four steps produce a usable model. Pull every contract with a renewal date in the next four quarters, with value and term. Score each account on usage trend and support case count, applying the pattern where zero cases and seven or more both signal risk. Layer relationship signals, particularly whether the original champion is still in the role. Then require a called outcome at two quarters out and again at one, with any account moving from renew to at-risk triggering an intervention rather than a note.

Two rules keep it honest. Never carry an account at 100 percent by default, since the default assumption is where renewal forecasts go to die. And measure renewal forecast accuracy separately from the sales number, because the two motions share nothing except the total they roll into.

For the acquisition half, the mechanics in how to create a sales forecast still apply, and the sales forecasting definition covers the shared vocabulary. Keep the models separate and the reporting joined, rather than the reverse.