B2B Sales Process Guide: Seven Stages, Real Exit Criteria, and a Forecast You Can Trust

Three different things hide behind the phrase "sales process," and conflating them produces most of the confusion in small-team selling. The process is the agreed sequence of stages and the facts required to advance through them. The pipeline is the artefact — the board or report — that displays deals against that process. A methodology is a philosophy of selling (consultative, challenger, solution) that shapes how conversations are conducted within the stages. A small team needs a clear process, a visible pipeline, and only as much methodology as it can practise honestly. This guide is about the first two; get them right and methodology becomes a coaching choice rather than a rescue attempt.

What makes this table a process rather than a list is the exit criterion column: an observable event, not a feeling. A deal advances when the fact exists.

The seventh stage is the one small teams most often omit, and its absence is expensive: a deal that closes and then onboards badly becomes a churned customer and a burned reference. Keeping handover inside the sales process — with a named owner and a first-value date — is the cheapest retention programme that exists.

Qualification is not a gate to keep prospects out; it is a decision about where your scarcest resource — selling time — goes this week. The four-question frame from the FAQ above (problem, money, decision, timing) works because each question can be answered from things the prospect actually said, and each missing answer converts directly into a discovery objective. The discipline that makes it work: answers must be quotes or events, not inferences. "They said the current vendor's downtime cost them a client last month" qualifies; "they seemed frustrated" does not. Write the answers on the deal record, because they are the raw material for the proposal and the ammunition for the negotiation.

Discovery earns the right to propose

The discovery conversation has three jobs: surface the real problem in the buyer's language, quantify the cost of doing nothing, and map who else is involved in the decision. The most reliable close of a discovery call is a summary played back — "here is what I heard, here is what I think it costs you, have I got that right?" — because a confirmed summary becomes the first page of a proposal the buyer recognises as their own thinking.

Single-threaded deals are forecast risk wearing a smile

Above a threshold value you set — a lakh, five thousand dollars, whatever makes the loss hurt — every deal should name at least two stakeholders: the champion you talk to and the person whose budget it is. Deals known through one friendly contact feel warm and die suddenly: the champion changes role, loses an internal argument, or simply stops replying. In pipeline review, "who else have we spoken to?" should be as routine a question as "what is the next step?"

The proposal section of this process is mostly cadence, covered in the FAQ: no proposal without a booked review, an expiry date on every quote, and the two-five-ten day rhythm afterwards. Negotiation needs guardrails agreed before the pressure arrives: a discount floor the rep can offer without escalation, and a rule that concessions are traded rather than given — a better price for a longer commitment, a faster start for a reference call. And every deal must end cleanly: won with a handover, or lost with a reason from the fixed list. The quarterly read of loss reasons — are we losing to price, to a named competitor, or to no decision at all? — is the strategy meeting most small businesses never hold, sitting unread in their own data.

A process only stays real if something inspects it, and the inspection should be light: stage-conversion rates (where do deals leak?), cycle time per stage (where do they stall?), ageing (which deals have outsat your norm and need help or honesty?), weighted pipeline against target (is there enough in play for next quarter?), and activity per rep (is effort where we agreed?). Run it from the live pipeline, not a reconstructed sheet. This is where tooling either carries the process or quietly kills it: in HelloGrowthCRM the stages, exit-criteria checklists, follow-up sequences, and these five reports are the out-of-the-box shape of the product, with AI scoring ranking each morning's queue — the process runs on rails rather than on memory. Whatever system you use, the test is the same: if the weekly review takes more than ten minutes to prepare, the tooling is failing the process.

Build the machinery behind this process: lead management software for the capture end, sales automation for cadences, the CRM dialer for logged calling, CRM with WhatsApp for conversation-heavy markets, AI CRM features for scoring and drafting, CRM vs Excel for the spreadsheet question, and use cases for team-by-team examples.