Bottom-up vs top-down forecasting. Why the best number uses both.

The short answer. Two different starting points, one number.

Bottom-up vs top-down forecasting isn't really a choice between a right method and a wrong one. Bottom-up forecasting starts with the pipeline: every open deal, its stage, its probability and its expected close date, rolled up by rep and then by team. Top-down forecasting starts somewhere else entirely: a board target, a market-sizing exercise, or simply last year's number with a growth rate applied, divided down into a quarterly or monthly figure. Both produce a number that looks the same on a slide. They are built from opposite directions, and that's exactly why comparing them is useful.

Xactly's 2024 Sales Forecasting Benchmark Report found that 80% of sales and finance leaders had missed a quarterly forecast in the past year, with more than half missing more than once. That's not a methodology failure specific to one approach. It's what happens when a team commits to a single number, from either direction, without a second, independently built number to check it against.

It's worth being specific about what "missing a forecast" usually means in practice, because it rarely means the whole number was wildly wrong. More often, a handful of deals the bottom-up model counted on slipped to the next period, or a top-down target assumed a growth rate the market simply didn't deliver that quarter. Either failure is survivable on its own. What makes it costly is finding out in the final week of the quarter rather than the first month, which is exactly the gap a second, independently built forecast is designed to close.

How they differ. What each one is actually measuring.

Neither failure mode announces itself. A bottom-up forecast built on stale pipeline stages will look confident right up until deals start slipping en masse in the final week. A top-down forecast built on an optimistic market assumption will look fine on a spreadsheet for months before the gap to reality becomes impossible to ignore. Running both doesn't eliminate either weakness, but it gives you two independent chances to catch it before the quarter closes rather than after.

There's also a practical difference in who trusts each number. Reps and frontline managers generally trust a bottom-up forecast more, because it's built from deals they can point to by name. Boards and investors generally lean on a top-down number, because it's tied to a growth story they've already approved. That split in who trusts which number is itself worth knowing: when a founder presents a top-down target the sales team privately doesn't believe, the gap tends to surface as quiet disengagement from the forecast process long before it shows up as a missed quarter.

Which to use and when. Match the method to the decision.

Use bottom-up forecasting for anything short-term and rep-specific: this quarter's commit number, which deals need a manager's attention this week, and coaching conversations about why a particular rep's pipeline doesn't support their number. It's the more granular view and the one closest to the day-to-day pipeline, our guide to pipeline forecasting covers how that number actually gets built stage by stage.

Use top-down forecasting for anything annual, strategic or board-facing: setting next year's revenue target before the pipeline to support it exists yet, sizing a new market entry, or making a headcount decision that has to be made before enough bottom-up data accumulates. It's the right tool when you need a defensible number and don't yet have, or can't wait for, granular pipeline data to build one from the ground up.

Most founder-led and SaaS teams need both running at the same time rather than choosing one permanently. Our guide to improving sales forecast accuracy covers the data-quality habits that make the bottom-up side trustworthy enough to actually reconcile against a top-down target.

Company stage matters here too. A very early-stage team with only a handful of deals in the pipeline at any time has little choice but to lean on a top-down target, because the bottom-up number is too thin a sample to mean much on its own. Once there's enough deal volume for the bottom-up roll-up to be statistically meaningful, usually once a team has more than a quarter or two of consistent pipeline history, it becomes the more trustworthy of the two for anything short-term, while the top-down number stays useful for the longer planning horizon it was always better suited to.

How Lauren would decide. A worked disagreement, and the rule that came out of it.

A founder-led team I worked with set an ambitious top-down target at the start of the year, derived from a board-level growth rate applied to the prior year's number. Three months in, the bottom-up pipeline roll-up showed the team tracking meaningfully behind that target, but when we looked rep by rep, the deals were there, several reps simply had stages sitting stale well past their expected close dates, a classic sign of mild sandbagging rather than a genuine shortfall. The instinct in the room was to lower the top-down target to match the gloomier bottom-up number. We didn't. Instead we audited the stale-stage deals directly with each rep, found that most of them were further along than the CRM showed, and the corrected bottom-up number moved most of the way back towards the original target on its own.

This is an illustrative pattern, not a single named client result, but the decision rule is the useful part: when the two forecasts disagree, don't average them and don't automatically trust the more cautious one. Investigate the specific gap first. A bottom-up number that's lower than a reasonable top-down target is usually a pipeline-hygiene problem to fix, not a target to quietly lower.

Getting this right. Build the habit, not just the model.

Set your top-down target at the start of the period and your bottom-up roll-up running continuously, then compare them monthly rather than waiting for quarter-end to notice they've drifted apart. The value sits almost entirely in the comparison, not in either model's sophistication; a simple spreadsheet target checked monthly against a basic CRM pipeline report catches most of the same problems a more elaborate forecasting tool would. If your sales dashboard doesn't currently surface both numbers side by side, that's usually the first thing worth fixing before investing in a more complex forecasting model on top.

One more habit worth building in early: write down why the two numbers disagree each month they do, not just the size of the gap. A one-line note, "bottom-up below target because two large deals slipped to next quarter" or "top-down target revised down after Q2 renewal data came in," turns the monthly comparison into a record you can look back on at year-end to see whether one side of the forecast is consistently the one that needs correcting. Teams that skip this step tend to keep re-litigating the same disagreement every quarter without ever learning from it.