You’re setting up a pipeline for the first time — or cleaning up a pipeline that was set up poorly the first time. Either way, the decisions you make now will shape how useful your CRM data is for the next two or three years. Get the structure right early, and pipeline management becomes a competitive advantage. Get it wrong, and every report you pull will be contaminated by data that doesn’t reflect how your customers actually buy.

This guide walks through each step of building a sales pipeline from scratch: defining your stages before you touch the CRM, configuring the system properly, setting up the right fields, and establishing the first baselines that allow you to make decisions when you don’t yet have historical data.

Why Getting the Foundation Right Matters Before Scale

The most common complaint from sales teams that have been operating for two or three years is that their pipeline stages don’t accurately reflect how deals move. A stage called “Proposal Sent” is everywhere. The problem is that sending a proposal is a seller action, not a buyer decision. A prospect who receives a proposal and actively engages with it is in a completely different position than one who received it and went silent. But if your stage definition doesn’t capture that distinction, both deals look the same in your pipeline.

Poor early pipeline design creates technical debt. When you eventually need to fix it — and you will — you’re not just renaming stages. You’re migrating active deals, retraining a team, and rebuilding every saved report. At five reps, that’s an afternoon. At fifty reps with thousands of historical deals, it’s a project.

The two biggest setup mistakes are building too many stages and naming stages after seller actions instead of buyer decisions. Both of these problems are easier to prevent than to fix.

Step 1: Define Your Sales Process Before Touching the CRM

Before you open your CRM configuration, map how your best customers actually buy. Not how you sell — how they buy. The distinction matters because your pipeline should reflect the buyer’s journey, not your internal process.

Start by answering these questions:

• What does a prospect typically learn about your product before agreeing to a meeting?
• What question are they trying to answer before they agree to a deeper evaluation?
• What internal steps do they go through before they can approve a purchase?
• Who else gets involved at each stage of their decision process?

From those answers, identify the key buyer decisions that mark genuine progression. Each stage in your pipeline should correspond to a decision the buyer has made, not a task your rep completed.

A common framework for five to seven stages:

1. Identified — contact is in your target profile and has been added to the pipeline
2. Qualified — you’ve confirmed there’s a real problem, budget range, and timeline
3. Discovery Complete — you understand the specific situation well enough to propose a solution
4. Proposal Under Review — you’ve presented a solution and they are actively evaluating it
5. Decision Pending — they’ve verbally committed to move forward and are working through internal approvals
6. Closed Won / Closed Lost — final outcome recorded

Notice the language: “Proposal Under Review” instead of “Proposal Sent.” “Decision Pending” instead of “Contract Sent.” The stage name reflects the buyer’s state, not your last action.

Step 2: Configure CRM Pipeline Stages with Exit Criteria

Once you have your stage definitions, configure them in your CRM with explicit exit criteria — the specific condition that must be true before a deal advances to the next stage. Without exit criteria, reps advance deals based on effort and optimism, not actual progress.

For each stage, define:

Exit criteria — what specifically must be true? For “Qualified,” it might be: confirmed budget range, identified decision-maker, and agreed timeline documented in the deal record.

Required fields — what information must be filled in before a deal can advance? If you require company size at the Qualified stage, you’ll always have it in your data.

Stage probability — the historical likelihood that a deal at this stage will close. On day one, you’ll set assumptions. Plan to update these with actual data after six months.

Setting stage probabilities thoughtfully matters more than most teams realize. If you set “Proposal Under Review” at 50% probability but your actual close rate from that stage is 30%, every forecast you pull will be systematically optimistic.

Step 3: Set Up the Essential CRM Fields

The fields you configure determine the dimensions available for analysis later. Missing a field at setup means filling it in retroactively across hundreds of records — or going without that analysis entirely.

Core deal fields every pipeline needs:

• Close date (required at entry or by a specific stage)
• Deal value (required — estimated value, updated as you learn more)
• Deal source (how did this lead enter your pipeline?)
• Deal owner (the rep responsible)
• Stage and stage entry date (automatically tracked by most CRMs)

Account and contact fields:

• Company size (employee range or revenue range)
• Industry or vertical
• Contact role (champion, economic buyer, technical evaluator)
• Account location or territory

Qualification fields (based on your chosen framework, whether MEDDIC, BANT, or your own):

• Budget confirmed (yes/no or range)
• Decision-maker identified (yes/no + name)
• Timeline established (date or quarter)
• Pain or business case documented (text field or score)

Custom fields for your business context:

This is where teams often over-build. Before adding a custom field, ask: “Will we filter, group, or report on this field regularly?” If the answer is no, skip it for now. You can always add fields later. You cannot easily clean up years of inconsistently populated fields.

Start with no more than 10-15 custom fields across account and deal records. That constraint forces you to prioritize the data that actually matters.

Step 4: Define Your First Baselines

You won’t have historical data when you launch. That means you need to set assumptions explicitly — and document them so everyone understands these are starting points, not validated numbers.

Average deal size — Set this based on your pricing, not on historical wins. If your product ranges from $10,000 to $50,000 annually, a reasonable starting assumption might be $20,000. Update with actual data after 20-30 closed deals.

Sales cycle length — Estimate based on how long a typical purchase process takes in your segment. A two-week evaluation is very different from a six-month enterprise cycle. Be conservative: new teams almost always underestimate cycle length.

Win rate — Without data, start with a conservative assumption. A 20-25% win rate is reasonable for most B2B sales motions with proper qualification. You may perform better or worse — the point is to have an explicit assumption so your forecasts are based on something documented.

Pipeline coverage ratio — If your win rate assumption is 25%, you need four times your revenue target in pipeline to hit it. Set this ratio explicitly so you know when you have enough pipeline and when you don’t.

Document all four assumptions somewhere visible — in the CRM, in a team wiki, in a shared doc — with a plan to review and update them after the first six months of real data.

Pipeline Setup Step	What to Decide	Common Mistake	Best Practice	How to Validate Later
Stage definition	Buyer-decision names and sequence	Naming stages after seller actions (e.g., “Proposal Sent”)	Use buyer-state language (“Proposal Under Review”)	Review stage conversion rates after 90 days
Exit criteria	Condition required to advance each deal	No exit criteria — deals advance on gut feel	Write exit criteria as a checklist for each stage	Audit a sample of “stalled” deals per stage quarterly
Required fields	Fields that must be populated to advance	Too many required fields at entry slows adoption	Require 3-5 fields at entry; more at key stages	Track field completion rate monthly
Stage probability	Win likelihood from each stage	Setting stage probabilities too high from the start	Start conservative; adjust after 20-30 closed deals	Compare forecast to actual wins quarterly
Core field setup	Which fields to include at launch	Over-building: 40+ custom fields that are never populated	Maximum 10-15 custom fields; add more as needed	Review field usage rate at 3 months
Reporting setup	Which views and reports to build first	Building 20 reports before knowing what you need	Start with 3 core views: pipeline by stage, by rep, by close date	Add reports in response to specific questions
Team training	What to train on first	Training on software features instead of process	Train on stage definitions, exit criteria, and data entry standards	Run first pipeline review within 2 weeks of launch

Step 5: Launch with Discipline, Not Complexity

The temptation when setting up a new pipeline is to build for the future. You imagine all the analysis you’ll eventually want to do and configure everything in advance. The result is a CRM with dozens of underused fields, stages that require data nobody enters, and a team that avoids logging deals because it’s too much work.

Start simple. Five or six stages. Ten to fifteen fields. Three core reports.

Train your team on the process first and the software second. Most CRM training sessions get this backwards: they spend an hour on how to create a deal record and five minutes on what a “Qualified” deal actually means. The field names don’t matter if the team isn’t aligned on the underlying definitions.

Run your first pipeline review within two weeks of launch. Not to analyze data — you won’t have meaningful data yet — but to establish the habit and surface the process questions that arise when the theory meets real deals. Those first reviews will reveal gaps in your stage definitions that you didn’t anticipate during setup.

After three months, run a brief audit: Are required fields being populated? Are stage conversion rates making sense? Are there deals sitting in a stage too long with no movement? Use those findings to make small adjustments before the pipeline has grown too large to reshape easily.

---

Frequently Asked Questions

How many pipeline stages is too many when starting out?

More than seven is usually too many for a new pipeline. Each additional stage adds cognitive overhead for reps and requires more data entry to maintain. Most sales processes can be accurately captured in five or six stages. If you’re finding that five stages don’t capture enough nuance, that’s usually a sign that one of your stages needs to be split — but start there before adding stages across the board.

Should we buy CRM software before we have a defined sales process?

You should define your sales process — at least a draft of your stages and exit criteria — before you configure the CRM. The CRM is a tool for executing and tracking your process. If you configure it before defining the process, you’ll configure it around the software’s defaults rather than around how you actually sell. That said, you don’t need a perfect process definition to start. A working draft is enough to begin, and you’ll refine it.

How do we handle deals already in progress when we set up the CRM?

Import active deals into the most accurate stage you can assign based on where they actually are. Don’t try to reconstruct detailed history. Set a “pipeline migration date” field so you know which deals were in progress before the new system went live, and exclude them from baseline calculations that require entry-to-close data. Let new deals entered after launch be your clean dataset.

When do we revisit and revise our pipeline stages?

Do a formal review of your stage definitions every six months in the first two years. Look for deals consistently stalling in one stage (may indicate the exit criteria are unclear), stage skipping (may indicate a stage doesn’t reflect real buyer progression), and team confusion about when to advance a deal. After two years with stable data, annual reviews are usually sufficient unless your sales process changes significantly.