# Revenue forecasting for a sales manager: what your CRM knows about next quarter, and what it doesn’t https://salesbond.app/blog/revenue-forecasting-sales-manager-crm/ Published: 2026-10-23 · Updated: 2026-10-23 Revenue forecasting is estimating how much revenue the company will recognize in a coming period. From your CRM you can build the sales side of that estimate out of three things: closed-won deals that have already been signed, open pipeline weighted by how often each stage has really closed, and the renewals you expect in the period. Present it to finance as a range, not a point. The CRM does not know when the money arrives or which customers will leave, so those two inputs come from outside it, and the forecast should say so on the page. ## Sales forecast vs revenue forecast: what changes A sales forecast says how much we will close by the end of the quarter. A revenue forecast says how much the company will recognize as revenue in it. Same deals, different clock. The first counts bookings: the value of contracts signed. The second counts recognized revenue: the part of those contracts the company has earned by delivering the service inside the period. A contract signed on the last day of March counts in full as a booking for the quarter. As recognized revenue it counts only for whatever part of the service falls inside the quarter, which may be nothing. Revenue forecasting models work from sales history, customer behavior and market trends, and finance keeps two other documents next to them: a budget, which sets a spending limit for a period, and a financial plan, which runs over several years ( [Orb: revenue forecasting models](https://www.withorb.com/blog/revenue-forecasting-models)). None of those is the sales manager’s job. Yours is the sales-side input. Call your part exactly that, the sales-side input to the revenue forecast: three numbers from the CRM, each labeled, so finance can convert bookings into recognized revenue with the assumptions only finance has. For finance you build that input from deals up, every time; the target-down version is a sanity check, and when it is worth running is in [bottom-up vs top-down forecasting](https://salesbond.app/blog/bottom-up-vs-top-down-forecasting/). ## Three numbers your CRM already holds The first is closed-won deals with a start date inside the period. This is the only number in the forecast that is a fact rather than an estimate. It needs two fields on the deal: the won date, which the CRM records when the deal is marked won, and the date the service starts, which you will usually have to add yourself as a custom date field. Make it required on the won stage. Without it, a deal signed in September for a January start lands in the wrong quarter. Even with it, the start date tells you which quarter revenue begins in, not how much of the contract falls inside that quarter. That split is finance’s to make. The second is open pipeline weighted by the historical win rate of each stage. Not the default stage probabilities the CRM ships with, and not the rep’s gut. Computing the rates from your own won and lost deals is covered in [weighted pipeline forecast](https://salesbond.app/blog/weighted-pipeline-forecast/), with the HubSpot setup in [deal stage probability in HubSpot](https://salesbond.app/blog/hubspot-deal-stage-probability-forecast/). For the revenue version you add one filter: only deals whose expected close date and start date both fall inside the quarter. Keep in mind what the weight can’t do. It describes the average deal at that stage, and the one in front of you may not be average. The third is renewals and repeat purchases. If your team creates a renewal deal for every expiring contract, this number is already in the pipeline and you only need to tag it, with a renewal flag or a link to the existing customer, so it is not counted twice. If renewals are not deals in your CRM the number is zero in the report, and you write “renewals: not tracked in CRM” rather than leaving the line blank. A blank looks like zero risk. One more field matters for all three: deal value. It’s the amount the rep entered, before any discount, credit or return that happens later in billing. ## Two numbers it can’t hold A CRM can’t hold two inputs: when customers pay and which ones churn. Payment timing first. A CRM records when a deal was won, not when the customer pays, and the gap between the two is the whole story of the quarter’s cash. The fields that bridge it, from invoice trigger to deposit, are in [cash flow forecast from the CRM pipeline](https://salesbond.app/blog/cash-flow-forecast-crm-pipeline/). The short version here: payment terms live in the contract and the billing system, and finance owns them. Churn is the second, and its the one that quietly ruins a renewals line. The CRM knows what was sold. It does not know who has stopped using the product or whose contract was renewed by habit for the last time. A lost renewal shows up only if someone created a renewal deal and then marked it lost, and that is not a deal anyone enjoys opening. Churn data lives in the billing system, in support tickets and in finance’s own history of non-renewals. We think the right move is to refuse to estimate either number from the sales side. Ask finance for their churn assumption and their average days-to-payment, write both as a line called “assumptions from finance” under the forecast, and leave your name off those two lines. A forecast that states where each input came from survives a bad quarter. ## From weighted pipeline to a range (low / likely / high) One number invites one argument about whether it is right. Three numbers invite a better conversation, because each has a definition anyone in the room can hold up against the CRM. Low is what has already happened: closed-won deals with a start date in the quarter, plus renewals that are already signed. Nothing estimated, nothing weighted. Likely is low plus the weighted open pipeline, filtered to deals that close and start inside the quarter. High is likely plus the open deals in your final stage taken at full value instead of their weighted value, again only those with a close date inside the quarter. That is the most you can defend, and it still leaves out the deal a rep swears will come in from stage two. ![Low, likely and high bars built from CRM data, with payment timing and churn assumption sitting outside the CRM.](https://salesbond.app/wp-content/uploads/2026/10/revenue-forecasting-sales-manager-crm-2.png) If your team already splits [commit vs best case](https://salesbond.app/blog/commit-vs-best-case-forecast/), keep that vocabulary and just add the start-date filter. Send all three to finance, with likely marked as the working number. If they ask for a single figure, give them likely and the width of the range: the width says how much of the quarter is still open deals, and nobody else can tell them that. ## Data checks before the number goes to finance The problems under a forecast are the same handful every quarter, so run through them the morning the number leaves your hands. Start with won deals in the period that have no start date, because each one is sitting in the wrong quarter or in no quarter at all. Then look at open deals whose expected close date has already passed; the fix is a new date or a lost deal, not silence. Final-stage deals with no next activity scheduled come next, since those are the ones padding your high number. A renewal deal and a repeat purchase for the same customer in the same period should be counted once. And watch deals in a second currency with no converted value, because the CRM will happily sum them anyway. How to measure which of these costs you the most accuracy over a few quarters is in [sales forecast accuracy for small teams](https://salesbond.app/blog/sales-forecast-accuracy-small-teams/). A disclosure, since this is the point where a manager usually starts building filters: we build SalesBond, a sales management layer on top of the CRM a team already uses; we started with Pipedrive. Its Sales Plan keeps fact, forecast and recommended action separate, with the assumptions, period and currency stated explicitly, which is the same discipline this article asks you to keep in a spreadsheet. [Check your CRM with SalesBond](https://salesbond.app/) if you would rather not build the checks by hand. ## How to review the forecast after the quarter Compare each of the three numbers with what actually happened, separately, and resist the urge to look at the total first. The total can be right by accident. A quarter where the weighted pipeline came in low and renewals came in high looks fine on one line and tells you nothing. Each miss has its own owner. If closed-won with start date was wrong, the start-date field was missing on some deals, and that is a hygiene problem you fix in the the won stage. If the weighted pipeline was wrong, the stage win rates have drifted from the history you computed them on, and they need recomputing before next quarter. If renewals were wrong, the churn assumption was wrong, and that conversation goes back to finance with your numbers attached rather than your apology. Keep the three lines, the two outside assumptions and the actuals in one sheet, one row per quarter. After a few quarters you will know which input has the widest error, and that is the one worth a field, a rule or a Monday review. Our bet is on the start date. It is the least glamorous field in the CRM, and it decides which quarter the money belongs to. ## FAQ ### How do you forecast revenue for next quarter from a CRM? Start from what is already signed with a start date in the period, add open pipeline weighted by your historical stage win rates, then add expected renewals. Finance then applies payment timing and churn, and the result goes out as a low, likely and high range. ### Is a sales forecast the same as a revenue forecast? No. A sales forecast counts deals closed in the period, while a revenue forecast counts the revenue recognized in it. A deal signed late in the quarter may add nothing to the second. ### What can’t a CRM tell you about revenue? When customers pay and which ones won’t renew. Both come from billing and finance, and the forecast should label them as outside assumptions.