20 January 13:24
At the Eduson Academy, Krasnoyarsk residents were told how to calculate a sales forecast not through Excel, but through real signals.
For example, the head of the sales department at a SaaS company built a forecast every quarter using the formula: “last quarter and 1.1.” As a result, there is a constant gap between plan and fact: ±27%. I decided to change my approach. I didn’t take revenue, but a metric further down the funnel: the number of leads who completed the demo and asked at least 2 clarifying questions. It turned out that the correlation between them and closing in 30-45 days is 0.84. Now he builds a forecast not from the past, but from the current supply of such leads. The error dropped to ±8%.
The Academy noted that the forecast is not an extrapolation, but an interpolation from the present. Another example is a B2B startup in logistics. They only looked at the total number of leads and didn't notice that 74% of hot leads come from webinars, not from advertising. When they started tracking: “how many leads after the webinar – how many demos – how many were closed,” the forecast became 33% more accurate. This happened because experts took into account the quality of the source, and not just the quantity.
When making a forecast, it is not at all necessary to calculate everything with an accuracy of the ruble; the range is important. One of the most common failures here is ignoring the timing of decision making. An EdTech company noticed that if a lead is out of action for 7 days, the likelihood of closing drops from 41% to 9%. They introduced the rule: “60-day forecast = (leads in work and 0.38) + (leads after 7 days without activity and 0.07).” Previously, the forecast was based on the total number of leads, now adjusted for customer behavior. The result is fewer redevelopments and a more stable cash flow.
Another life hack is analyzing questions on a demo. The owner of a CRM implementation agency began to record: what three questions are asked by leads who ultimately buy. It turned out: “How quickly can we launch?”, “Who will supervise?”, “Are there any cases in our niche?” If these questions are asked, the probability of a deal is 82%. If not, 14%. Now managers at the end of the demo ask: “Do you still have questions about deadlines, responsibilities or cases?” This is not imposition, but filtering.
There should not be only one forecast. There should be a range: pessimistic (only leads with 3+ questions), basic (all active leads), optimistic (all leads + return ones). The consulting team uses such a system and over the past six months has never experienced a loss in cash flow due to an error in planning.
“One of our colleagues, a former analyst, now makes a forecast like this: he takes 3 weeks now, counts how many leads have moved to the next stage. Divide by 3 to get average speed. Multiplies by the number of leads in the current stage – gets a score. This is not a model. This is common sense, strengthened by data,” the Academy said.
If you are interested in what 5 metrics replace complex models, how to build a forecast without IT and Data Science, and how to use it to manage a team, read more in the materials of the Eduson Academy. There are also table templates and scripts for negotiations that increase conversion.
Advertising. Eduson LLC. erid: 2Vfnxx2oASu
Newslab now in MAX. News is also published in OK | VK | Zen | Telegram
