How to forecast upside without sandbagging or fantasy

Every quarter, the same theatre plays out. Reps submit a commit number they're 95% sure they can hit, a best case that includes a deal they secretly think is dead, and an "upside" column that's either padded with dreams or stripped so bare it's useless to the VP trying to plan hiring.

The problem isn't the rep. It's that most forecast categories are designed to protect people, not to predict revenue. Upside, in particular, has become the dumping ground for anything a manager can't categorise — which means it carries almost no information by the time it rolls up.

A useful upside number is one a CRO can actually plan against. That requires forcing rigour into the column most teams treat as guesswork.

What "upside" should actually mean

Strip the word back to its job. Upside is the set of deals that are not in your commit, but have a credible path to closing in the period. Not "could close if the planets align." Not "we have a champion and a budget conversation next week." Credible path means: you can name the next three steps, the buying committee is identified, and the customer has acknowledged a timeline that fits the quarter.

If a deal fails any of those three tests, it doesn't belong in upside. It belongs in pipeline for next quarter, or in best case with a flag.

This sounds obvious. The reason teams get it wrong is that upside gets used as a political instrument. Reps inflate it when they're behind on pipeline coverage and need to look healthier in the 1:1. They deflate it when they're already on track and don't want to raise expectations. Both behaviours destroy the signal.

The fix is to make upside cost something to put a deal in, and cost something to leave one out.

The two-question test that kills both sandbagging and fantasy

Before any deal goes into upside, the rep answers two questions in writing inside the opportunity record:

1. What specifically has to be true for this to close in the period? Not "they sign the order form." The actual sequence: legal review starts by week 8, security questionnaire returned by week 9, procurement kicks off the week after, signature in the final week. If the rep can't write the sequence, the deal isn't upside.

2. What's the single most likely reason it slips? Forces honesty. Common answers: "the economic buyer hasn't confirmed budget," "we're waiting on a competitive evaluation we don't control," "their CFO is on parental leave until July." A rep who can't name a slip risk is either not close enough to the deal or is fantasising.

These two answers become the inspection points in the forecast call. The manager doesn't argue about the deal; the manager argues about whether the answers are still true.

Teams that run this exercise tend to find their upside column shrinks substantially in the first month — and then becomes a far more accurate predictor of what actually lands.

Build an upside conversion benchmark you trust

You cannot forecast upside without knowing what fraction of it has historically converted. Most teams either don't measure this or measure it at the company level, which is useless for the AE trying to call her own number.

Pull the last six to eight quarters of forecast snapshots. For each AE, calculate: of the deals they marked upside at the start of the final month, what percentage closed in the period? Then segment that further by deal size band and by sales stage at the time of the call.

A hypothetical: say an AE has called $1.2M of upside heading into the last month of Q2 2026. Looking back, her upside has historically converted at about a third when deals are in late-stage proposal, and almost not at all when they're still in technical evaluation. If $800K of her current upside is in technical evaluation, the realistic contribution to commit is closer to $130K than $400K.

That's the number the manager should be planning against. Not the rep's gut. Not a flat company average. The rep's own demonstrated conversion rate, segmented by stage.

This also gives you a clean way to coach. A rep whose upside consistently converts at a low rate isn't being conservative — they're miscategorising deals. A rep whose upside converts at a high rate is sandbagging and should be pushing more deals into commit.

The weekly inspection ritual that holds it together

Forecast accuracy decays without a weekly cadence. The ritual that works is short and structured. For every deal in upside:

• Read back the two-question answers from when the deal was added.
• Ask what has changed in the last seven days that confirms or contradicts those answers.
• If nothing has changed in two consecutive weeks on a deal you expect to close in the next month, that's a signal — usually that the customer's urgency isn't real.

Movement is the signal. A deal where the champion went quiet, the next meeting got rescheduled, and the procurement intro hasn't happened is not upside, regardless of what the rep wants to believe. A deal where the customer sent over their standard MSA, looped in their security team unprompted, and asked about provisioning timelines is upside even if the rep is nervous about calling it.

The discipline is to weight observable buyer behaviour over rep sentiment. Sentiment is where both sandbagging and fantasy live.

What changes when you do this

The first quarter is uncomfortable. Upside numbers drop, sometimes sharply, because deals that were padding the column get reclassified. Senior leadership may panic at the apparent shrinkage. This is the moment to hold the line: a smaller, more accurate upside number is more useful for planning than a larger, fictional one.

By the second or third quarter, the forecast call gets faster. Most of the argument has already happened in the opportunity record. Managers stop relitigating the same deals every week. And the gap between commit and actual closes — which is the only metric that actually matters — tightens.

The reps who resist hardest are usually the ones whose upside was doing the most political work. That's fine. The system isn't there to make them comfortable. It's there to make the number real.

The takeaway

• Require every upside deal to have a written sequence of remaining steps and a named slip risk before it counts; deals that fail this test belong in next-quarter pipeline.
• Calculate each AE's historical upside conversion rate segmented by deal stage, and use that — not gut feel — to translate upside into expected revenue.
• Run a weekly inspection where the only question is whether observable buyer behaviour has changed since the deal was added; two stagnant weeks is a reclassification trigger.