Use this when: You're tracking deals in a spreadsheet, your head, or sticky notes, and it's starting to break. You've lost track of where prospects are, missed follow-ups, or realized you have no idea what your actual close rate is.
You're done when: Every active deal has a stage, every stage has a clear entry/exit criteria, and you can answer "what's my pipeline worth?" in under 60 seconds.
The Sequence
1. Pick a tool and commit
For most startups under $5M ARR, the choice is simple: HubSpot (free tier), Pipedrive, or Close. All three work. Don't spend a week evaluating 12 options. Pick one, set it up, and switch later if you outgrow it. The tool matters far less than using it consistently.
If you have fewer than 20 active deals, a well-structured spreadsheet is genuinely fine. The signal to move to a CRM is when you miss a follow-up or lose track of a deal stage because your spreadsheet doesn't remind you.
- Deliverable: An account created and your team invited.
- Common mistake: Over-researching CRM options. The best CRM is the one your team actually uses.
2. Define your pipeline stages
Your stages should mirror how a deal actually moves, not how you wish it moved. A typical early-stage pipeline:
- Lead — Someone who fits your ICP but hasn't been contacted
- Contacted — You've reached out and they've responded
- Discovery scheduled — A call is on the calendar
- Discovery completed — You've had the conversation and confirmed fit
- Proposal sent — They've received pricing and scope
- Negotiation — They're evaluating, asking questions, or discussing internally
- Closed Won — Signed and paying
- Closed Lost — Dead, with a reason logged
Each stage needs clear exit criteria. A deal moves from "Discovery completed" to "Proposal sent" only when the proposal is actually sent, not when you plan to send it. Don't add stages you don't need. Five stages is better than twelve.
- Deliverable: Pipeline stages configured in your CRM with written entry/exit criteria.
- Common mistake: Creating stages that reflect internal process ("Reviewed by team") instead of buyer behavior. Stages should track where the buyer is, not where you are.
3. Set up the metrics that matter
Close rate is an outcome, not a leading indicator. It tells you what happened, not what will happen. Track the full funnel:
- Schedule rate: What percentage of outreach converts to a booked call?
- Show rate: What percentage of booked calls actually happen?
- Offer rate: What percentage of completed calls result in a proposal?
- Close rate: What percentage of proposals convert to signed deals?
- Cash collected upfront: What percentage of closed deals pay immediately vs. on terms?
- Average call-to-close: How many conversations before a deal closes (1-call, 2-call, 1.2)?
Improving your schedule rate by 20% has the same impact as improving your close rate by 20%. Most founders obsess over close rate and ignore everything upstream.
- Deliverable: A dashboard or report tracking these six metrics, updated weekly.
- Common mistake: Only tracking close rate. A 40% close rate with a 30% show rate means 70% of your booked calls don't happen. Fix the show rate first.
4. Implement lead scoring
Not all leads are equal. Track which characteristics predict a close. After 30-50 deals, you'll see patterns: leads from a specific channel close at 3x the rate, leads with a particular job title convert faster, leads from companies of a certain size have higher ACV. Score incoming leads on these attributes and route the best leads to your best closers.
This doesn't need to be sophisticated. A simple 1-3 scoring system (cold, warm, hot) based on 3-4 attributes is enough to start. Refine as your data grows.
- Deliverable: A lead scoring system with 3-4 attributes and a routing rule (best leads to best reps).
- Common mistake: Distributing leads equally regardless of quality. Your top closer on your worst leads is a waste. Your top closer on your best leads can 5x their output.
5. Build your data discipline
The CRM is only useful if the data is accurate. Set two rules and enforce them without exception:
- 5-minute rule: After every prospect interaction, update the CRM within 5 minutes. Not end of day. Not tomorrow. Five minutes.
- No deal without a next step: Every deal in the pipeline must have a next action and a date. If it doesn't, it's not a real deal, it's a wish.
If you have a team, review CRM hygiene in your weekly pipeline meeting. Deals without next steps get flagged. Deals that haven't moved in 14 days get a decision: advance or kill.
- Deliverable: Written CRM rules distributed to the team, with a weekly hygiene review cadence.
- Common mistake: Treating CRM entry as busywork. Bad data in means bad decisions out. Your forecasts, your pipeline reviews, and your hiring decisions all depend on what's in the CRM.
Template
Example
A 3-person sales team at an HR tech startup was tracking deals in a shared Google Sheet:
- Two deals fell through the cracks in the same week — one because nobody followed up after a demo, another because two reps were working the same account without knowing it
- They moved to HubSpot's free tier in a day
- Pipeline stages mirrored their actual process (lead, demo scheduled, demo done, proposal, negotiation, won/lost)
- They set up a simple dashboard tracking show rate, offer rate, and close rate
- Within a month, they discovered their show rate was 52%, meaning half their booked demos never happened
- Adding a pre-call confirmation sequence (a personalized video message the day before) pushed show rate to 71%
- That single fix, invisible without the CRM data, added 4 closed deals per month with zero additional pipeline
Written with ❤️ by a human (still)