Sales Forecasting
Sales Forecasting
What is sales forecasting?
- Estimating revenue from the pipeline and history
- Basis for business planning and resource allocation
- If forecasts keep missing, audit your data
Once you adopt sales forecasting, the language of your meetings changes. “This quarter looks fine” becomes “Commit KRW 1.2 (≈ US$1) billion, best case KRW 400 (≈ US$1) million, KRW 180 million (≈ US$134,700) short of target,” and the team discusses on the spot how to close the gap. Instead of discovering a shortfall only in the last month of the quarter, you spot it 6–8 weeks ahead, which leaves time to act, whether that means running a promotion or reactivating idle leads. The real value of a forecast lies not in hitting the number but in buying you time to respond.
In practice, teams combine three methods. A weighted pipeline, which multiplies deal values by stage probabilities, suits organizations with many opportunities of fairly even size; a commit approach, where reps judge deal by deal, suits enterprise sales with large deals and few of them. A statistical forecast built from historical trend and seasonality serves as a control that checks the other two. If the three values differ by more than 20%, one data source is wrong, so find the cause before choosing a number. A forecast is different from a quota or a plan. A quota is the number you need to hit, while a forecast is where you will land on your current trajectory, so if the two are reported as the same number, it is already hope, not a forecast.
The most common failure happens when “forecast accuracy” is made a performance metric. Reps start lowballing and then beating their numbers, so forecasts turn chronically conservative and management under-hires and under-stocks. When orders then surge, delivery deadlines slip and customers are lost. Conversely, if stage definitions are loose and a deal moves to the proposal stage simply because “we sent a proposal,” a deal where the customer has no budget gets counted at 40% probability, the pipeline fills with phantom deals, and the forecast always falls short. To fix forecast accuracy, start by writing down stage entry criteria, not tweaking the numbers—a meeting with the decision maker, confirmed budget, an agreed timeline.
John KimAuthor of B2B Marketing Design and CEO of Nanum Consulting
Seungtae ParkAI transformation (AX) practitioner who teaches AI business development and marketing automation by building it hands-onA practice that took hold as sales pipeline management and CRM matured. Using AI and data analytics to improve forecast accuracy is now a major trend.
Consider a hypothetical case. Company A sells industrial measuring equipment, with 140 employees and annual revenue of about KRW 32 billion (≈ US$23,941,300). Its 9 sales reps logged opportunities in the CRM, but quarterly forecasts missed actual results by 30–40% for four quarters running. In Q3 the forecast was KRW 8.5 billion (≈ US$6,359,400) against actual results of KRW 5.8 (≈ US$1) billion, and the materials ordered in advance ended up as excess inventory.
- When the head of sales pulled the CRM records for the past four quarters and calculated actual conversion rates by stage, the win rate at the “Proposal” stage was 17%, not the company’s assumed 40%. The “Negotiation” stage, by contrast, came in at 78%, higher than the assumption.
- Digging into the cause, they found reps were moving a deal to the proposal stage as soon as they sent a quote. The head of sales changed the entry criteria so that the proposal stage requires “confirmed customer budget + at least one in-person meeting with the decision maker,” and moved all 47 deals that did not meet the criteria back to an earlier stage.
- At the weekly Monday pipeline review, reps had to classify each opportunity as commit, best case, or pipeline. When one rep put a KRW 1.2 billion (≈ US$897,800) deal in commit, the division head asked, "Have you exchanged a draft contract?" With no answer, the deal was moved down to best case.
- A manager on the planning team separately ran a statistical forecast that reflected the seasonality of the past three years of sales, then compared it with the sales team’s forecast. At the first Q4 comparison the two figures had diverged to KRW 6.1 billion (≈ US$4,563,800) and KRW 7.4 billion (≈ US$5,536,400). The review found that three Middle East export deals were about to slip to the next period because of shipping delays.
- The final Q4 forecast was KRW 6.3 billion (≈ US$4,713,500) against an actual close of KRW 6.6 (≈ US$1) billion, an error of just 4.5%. With forecast error cut from the 30% range to single digits, the team could size advance material orders to real demand, and next quarter surplus inventory fell by KRW 620 million (≈ US$463,900) from the prior quarter.