Deal analysis and review

Understand the score and choose the next action

DealPack starts with a historical win rate for comparable deals and adjusts it using engagement, meetings, deal size, and close-date changes. It pairs that estimated chance of winning with a suggested explanation and next step. Its approach to buyer indecision is partly inspired by The JOLT Effect. The score has not been validated as an exact probability for an individual deal or a current-quarter close.

Review the inputs that moved the score. Few substantive replies might call for checking whether the project is still active; repeated date pushes might call for clarifying the approval process. Compare later wins and losses by saved score band to check whether higher-scoring deals actually win more often.

• Check what the buyer said or agreed to do. Adding meetings just to raise the score does not resolve a budget, product, or approval problem.
• Keep source thresholds, multipliers, and silence clocks adaptable to the sales motion.
• A score does not automatically close a deal, change ownership, or release an account.

Read the factors behind the win proxy

DealPack uses a historical base win rate multiplied by observed factors for reply ratio, meetings, people on activity, deal size, and close-date pushes. The model then applies its limits and overrides. Review those inputs alongside the underlying buyer evidence.

The factors can overlap: more meetings and more people may describe the same evaluation. Treat the result as a proxy estimate until later outcomes establish how its score bands perform. An eventual win estimate also needs a separate timing judgment before it supports a period forecast.

Keep account coverage separate from deal risk

When a high-fit account has a low-scoring opportunity, investigate the buying process before changing coverage. A low proxy score does not establish that the account is a poor fit or that another rep should own it.

Use new engagement to ask whether the account is actively shopping, what buying cycle it is working toward, and what motivates the decision. On an existing opportunity, pass that context to the owner and update the next buyer action.

Check the buyer facts behind the score

Compare deals from the same sales motion, such as new business or expansion. Use the deal value and buyer evidence recorded at the review date. Trial users and product activity can help identify people to ask about the evaluation; confirm who is involved in the purchase.

• Check the problem under evaluation, substantive replies relative to outreach, meetings that clarified a buying requirement or produced a next step, buyer roles, latest commitment, and date changes.
• Distinguish missing activity links from confirmed inactivity, and respect buyer-agreed return dates.
• Name alternative explanations. Few replies may reflect priority, relevance, the wrong person, or incomplete logging; many replies may be administrative.
• Keep competitive choice, no active project, timing, economics, product requirements, and unknown loss reasons distinct.

Choose the next seller action

Use the next buyer conversation or evaluation milestone to answer the unresolved question. The seller updates the action, stage, close date, or closure reason. RevOps maintains the scoring rules and data refresh; the manager helps with pricing, resources, or deal strategy.

• Engaged but unable to choose: recommend one option and explain why.
• Repeated slips with ongoing replies: ask what must be true for the buyer to proceed confidently.
• Silent after an active evaluation: ask whether the project is still a priority and whether a specific concern is preventing the next step.
• If activity is missing from the CRM, fix the links before scoring it as silence. If nobody has confirmed a problem and evaluation, return to qualification.
• Save the buyer fact, your explanation of the blocker, the next action and date, and any stage or forecast change in the existing CRM record.

The buyer keeps asking for more information

The buyer agrees there is a problem but keeps requesting another demo, comparison, or reference. First establish what the next piece of information would help them decide. A missing requirement needs an answer; uncertainty about choosing or implementing the product needs a different response.

Dixon and McKenna’s JOLT approach addresses this kind of indecision through a clear recommendation, limits on unnecessary exploration, and reduced purchase risk.

• Ask which specific question still prevents the decision.
• Recommend the option that fits the agreed requirements and explain the tradeoffs.
• If implementation is the concern, agree on a realistic first phase, responsibilities, and success criteria.
• Record what the buyer agreed to decide next. Another meeting alone is not a reason to raise the forecast.

Source: Matt Dixon and Ted McKenna: customer indecision and the JOLT approach

Check whether the score predicts outcomes

Use only point-in-time information and test predictive rules on an untouched later cohort. Correlated engagement factors can overstate confidence. Historical association does not establish that adding meetings or contacts causes wins.

• Review later outcomes by saved score band with counts, a defined horizon, and unresolved cases visible.
• Track whether the seller resolved the question raised in the review and whether the buyer agreed to a next step.
• Check timing separately before a win proxy feeds a period forecast.
• Check the report's last calculation date. Editing source rows does not refresh a precomputed report.