Most forecast calls start the same way. A leader opens the CRM (customer relationship management) system, hears a stage and a close date for each deal, and then has to explain the total to a CFO (chief financial officer) or a board without a shared record of how the forecast was built. When the answer lives in the leader's memory, every miss becomes a debate about judgment, with no process to review. This article covers how to set a consistent operating cadence, how to use data and tracked commitments to steady the forecast, and how an AI layer can support that cadence so you can build a forecast your CFO will trust.
The usual failure is inconsistent inputs: each manager runs the call a little differently and each rep updates deals on a different schedule. The history of a deal, such as who said what and why a close date moved, sits in the heads of the people closest to it. When the leader walks into the room, that history has to be rebuilt by hand.
Lesley Young, a go-to-market leader who has worked at Oracle, Meta, Twilio, and Box, tells a story on Revenue Mavericks that shows how this plays out at the deal level. Early in her career she was an account executive on a large deal with a large legal software company. The buying committee praised the demo, the capability questions were answered, and she believed she had a champion guiding her through the process. She flew home feeling confident, and then the deal was lost. She believed an incumbent with an executive-level relationship had overridden the committee. Looking back, she did not have a champion, because that person lacked the influence to advocate for her. It had a material impact on the quarter.
A deal that looks strong in a review but has no verified path to the person who decides is exactly what a forecast call has to catch, and memory is a weak tool for catching it.
Young says the operating cadence matters at any company size, because the people who work for you need it. She compares it to a college syllabus. Before a student walks into class, they have no context for what needs to get done, and a sales team without a defined rhythm over the next 13 weeks of a quarter is in the same position.
She offered an example of what that rhythm can look like.
Week one, we look at slip deals and what we're going to do to close them. The middle of month two, we look at where the forecast is, we lock the forecast down.
From there, the rep's job is to work toward the commit they gave. This is her illustration, not a prescribed schedule. Her point is that people should know in advance what will happen and when, and that it needs to be consistent.
Includes the forecast, that includes the QBRs (quarterly business reviews), that includes how we talk about a forecast.
For an operator, the practical step is to write your forecasting process down as a calendar. Decide which week is for slipped deals, when the forecast locks, what a rep must bring to each review, and who signs off on changes after the lock. In many companies RevOps owns those definitions (revenue operations) while sales leaders own the forecast itself. A locked forecast gives you a fixed baseline to compare against, so you can improve forecasting instead of arguing about it each week.
Once the calendar exists, the content of each review needs a standard too. She can get emotional about a deal, and she asked how you take the emotion out while still giving people feedback they can act on. Her answer is to lead with data. When she goes in to coach, whether in a deal review or a coaching conversation, the question she asks: What is the data we're using?
She gave an example from Box. A rep described herself as a horrible salesperson. Young had tracked the rep's commitments and walked her through them: where she started the quarter, what she said she would do, where she stood at the end of month one, and what she said she would land at in month three. In Young's recollection, the rep landed within 10K of her month-three commit, even though she felt stressed about the process. Young told her she now had a formula and should keep following it. This was one rep, and Young was describing her own experience, not a rule for every team. It was also done in a spreadsheet at the time.
A commitment recorded early becomes evidence. When you capture the commit at the start of the quarter, at month one, and at expected landing, you can see whose forecasts hold, whose drift, and where coaching will help. Landing close to the commit is a sign of strong forecast accuracy at the rep level, and when you measure forecast accuracy this way, a manager can defend a forecast without relying on recollection.
A forecast has an audience beyond the sales floor, and Young described how she learned to write for it. While in a leadership program at Informatica, she met quarterly over lunch with a board member who gave her a simple summary of what the board wants.
The board cares about predictability. They care about growth and they care about efficiency.
His advice was to learn what the CFO cares about, so she asked the company's CFO to meet monthly and teach her how the finance side of the business worked. That changed how she thought about the business and what a go-to-market team could do to affect it.
The impact of what happens when you don't know and when you miss.
Package the forecast with its locked commitments and the reasons for each change. A forecast alone invites questions about how it was reached. A forecast that arrives with the locked commitments, the changes since the lock, and the reasons behind each change answers many of those questions before they are asked. Inspectable commitments are easier to defend than a narrative.
Teams often expect a sales methodology to produce a forecast, and that expectation causes trouble. MEDDIC stands for Metrics, Economic buyer, Decision criteria, Decision process, Identify pain, and Champion, and it is commonly used to qualify complex deals. Young says MEDDIC is how a team tests a deal, not a forecast tool.
A methodology scorecard tells you whether a deal has the right ingredients, while a forecast tells you when and how much revenue will land. Read this way (our reading, not hers), her lost deal fails the test: a champion without influence fails even if every other box is filled in. If you want to improve forecast accuracy, use the methodology inside the review cadence to challenge each deal, and keep the forecast tied to the commitments you track against it.
She also ties this to AI, and in her opinion the fundamentals matter right now.
If those insights are built off of an inconsistent operating model, they're not going to enable you to take advantage of what AI has to offer to go to market.
She was speaking about the AI opportunity broadly, and applying it to forecasting is our own step. Use this test before you buy anything. If two managers would score the same deal differently, or if no one can say when the forecast locks, any AI sales forecasting tool is likely to inherit that inconsistency and make it faster.
Terret's own view is that a defined cadence is the foundation and AI should build on it. That is our position as a company, separate from Young's remarks. Revenue data lives in fragments across the CRM, email, calls, and the data warehouse. No single AI sees the whole picture, so answers stay incomplete when you ask a question like "where will we land this quarter and what are the headwinds?" Leaders end up reconstructing that picture by hand before every forecast call, and even a good answer then has to be turned into action by someone else.
Our platform, Terret Nexus, takes a revenue question to action. It starts with our Revenue Graph, which gives you a complete revenue picture across structured and unstructured data from every system, behind an enterprise-grade governance layer, so the answer reflects the whole picture. You ask a question, our AI Architects analyze your complete revenue picture and design the go-to-market system, and our AI Agents execute what the Architects design, including scoring deals and generating forecasts, so the answer is acted on at the right moment. That shows up on the forecast call, where you can run forecasting on Terret Nexus.
Your sales team remains the drivers. The Architects and Agents do not replace the cadence, the locked commitments, or the judgment a leader brings to the room. They support the cadence you have defined, so the forecast conversation covers what to do next.
It means the leader explains the forecast from what they remember about each deal instead of from a shared record. The history of who said what and why a close date moved sits with the people closest to the deal, so it gets rebuilt by hand before every call. A locked forecast with tracked changes moves that history into the record.
Consistency matters more than any one schedule. As an example, Lesley Young described on Revenue Mavericks looking at slipped deals in week one and locking the forecast in the middle of month two. Your calendar may differ, but it helps to write it down as part of your forecasting process and then run it the same way each quarter.
Record each rep's commit at the start of the quarter, at the end of month one, and at expected landing, then compare those commits with the final result. Over several quarters the pattern shows who holds their commit and who drifts. This lets you measure forecast accuracy by rep as well as in total, which points coaching at the right people.
MEDDIC is not a forecast tool.
MEDDIC (Metrics, Economic buyer, Decision criteria, Decision process, Identify pain, Champion) is a way to test whether a deal has the right ingredients. To improve forecast accuracy, pair that deal testing with tracked commitments, because the forecast is a separate record of what will land and when.
Young relayed advice from a board member that the board cares about predictability, growth, and efficiency. In practice that suggests a forecast that arrives with its locked commitments and the reasons behind each change. A forecast your CFO will trust is one they can inspect.
It is better to settle the basics first. In Young's opinion, insights built on an inconsistent operating model will not let you take advantage of what AI has to offer. Applied to forecasting, that is our reading: if managers score deals differently or no one can say when the forecast locks, AI sales forecasting is likely to inherit that inconsistency and repeat it faster.
In many companies, RevOps owns the definitions and the calendar, while sales leaders own the forecast itself. What matters most is that one person or team is accountable for the lock date, the stage definitions, and the rules for changing a commit after the lock.
Our AI Agents can generate forecasts for your forecast call. Our platform, Terret Nexus, takes revenue questions from answer to action, and our Revenue Graph gives a complete revenue picture across structured and unstructured data from every system, so the forecast draws on more than a single system. It runs on your cadence, and your sales team remains the drivers.
With Terret Nexus, you ask the forecast question, our AI Architects answer it from the complete revenue picture, and our AI Agents act on it at the right moment, while your team stays in charge.
Request a demo to see it on your cadence and systems.