Sales pipeline on memory and a stale spreadsheet: visibility before you buy a CRM

You are about two years in. Revenue is real, deals do close, and you can name your best customers from memory. Enterprise sales ops is rarely the bottleneck. Decisions now ride on a pipeline you cannot see: a spreadsheet someone meant to update, plus whatever is in your head after the last three calls.

That combination is common at this stage. It is also where missed follow-ups, deals going quiet without anyone noticing, and confident guesses about what will close this month start to cost money. Start with visibility and a few habits. A six-month CRM rollout can wait.

When informal tracking stops working

Informal sales works while one person holds the thread and deal volume is low enough that “I will remember Tuesday” is true most weeks. It stops working when the business crosses one or more of these lines:

• You cannot answer “what needs attention today?” in under five minutes without opening email, Slack, and three tabs.
• Forecasting happens in meetings without shared data: “I think three will close” vs “I only see one with a next step.”
• Handoffs fail: a founder goes on vacation, a contractor helps with outreach, or a part-time hire joins, and deals stall because context lives in private notes.
• The spreadsheet is always behind and everyone knows it, so people stop updating it and leadership plans from gut feel anyway.
• Quiet deals stay quiet until the buyer ghosts or a competitor wins, because nobody owns a regular “stale deal” review.

None of that means you failed at sales. It means the operating model outgrew the tooling. Add narrower visibility first. Skip the 40-field CRM on day one.

Minimum viable pipeline visibility (before a heavy CRM)

“Minimum viable” here means one shared place where every active opportunity has the same few fields, updated on a rhythm you will actually keep.

The shared place can be a CRM, Airtable, Notion database, or a spreadsheet with rules. The tool matters less than one source of truth everyone agrees is the pipeline for planning.

Fields worth having on day one (add more only when a decision needs them):

The rhythm matters as much as the columns:

• Daily (5–10 minutes for the owner): scan deals where next step date is today or overdue.
• Weekly (15–30 minutes for whoever sells): review anything with no activity in 7–14 days (pick one threshold and stick to it).
• Monthly: close out dead deals so your active count stays honest.

If you cannot maintain that rhythm in a spreadsheet, buying HubSpot or Pipedrive will not fix it. You will have a fancier graveyard of stale records.

What “proper” looks like at different team sizes

“Proper” means the least structure that matches who touches deals.

Two people (often founder + cofounder or founder + first hire)

One pipeline, one weekly sync, shared definition of stages. Email and calendar can stay the system of record for messages; the pipeline only needs to answer who owns the next step and when. Duplicates hurt less because you can still fix them in conversation.

Around five people with multiple people selling or delivering

You need stage definitions written down (three to seven stages; resist adding a dozen more), one owner per deal, and a rule for when a deal moves backward (it will happen). Reporting is still simple: count by stage, list overdue next steps, flag no touch in N days. This is where a real CRM often pays off because permissions and activity history start to matter.

First dedicated sales hire

“Proper” shifts from founder memory to repeatable inspection. The hire needs to log calls and next steps without asking you what happened in 2023. You need a Monday pipeline review that does not require you to narrate every deal. Forecasting can stay lightweight, but activity and stage hygiene belong in the job description, same as logging calls.

Simple stages and deal hygiene (without over-engineering)

Start with stages that mirror buyer actions: demo sent, contract out, and similar milestones.

1. Qualified conversation (problem and budget are plausible)
2. Proposal or trial (they are evaluating seriously)
3. Verbal yes / contract out (paperwork or procurement in flight)
4. Won / lost (close the record; do not leave zombies in “proposal” forever)

Rules that keep the pipe honest:

• No stage without a next step and date. If you do not know the next step, the stage is wrong or the deal is not really active.
• Lost is a valid outcome. Archiving lost deals keeps win-rate math and focus sane.
• One deal per real opportunity. Split only when there are separate contracts or separate buyers; otherwise merges and duplicates will poison any report you run later.

Automations can wait. A colored filter for “next step before Friday” is more useful than a workflow builder you will not maintain.

Missed follow-ups and deals going quiet

These are usually process problems wearing a software costume.

Missed follow-ups happen when next steps live in email drafts, sticky notes, or “I will ping them after this launch.” Fix: every open deal has one next step in the shared pipeline, and the owner gets a daily short list (calendar block, task tool, or CRM task view).

Deals going quiet happen when nobody asks “who have we not heard from?” on a schedule. Fix: a weekly stale deal list (no touch in 14 days, or whatever matches your cycle). The action might be qualify out, change channel, or escalate to a founder call instead of another template email.

If the playbook is right but nobody runs the review, the bottleneck is habit and ownership. Software helps when it makes the stale list automatic and visible. It rarely helps when only one person cares about hygiene.

First CRM vs spreadsheet plus rules (how to choose)

Stay on spreadsheet plus rules briefly if:

• One or two people sell, and you will run the weekly review yourself for the next 60 days.
• Deal count is small enough that manual hygiene takes a few minutes each week.
• You are still proving stage definitions and what “next step” means for your business.

Move to a first CRM when:

• More than one person needs to update and read the same pipeline daily.
• You are hiring sales help and need history they did not live through.
• Email search is your CRM and it is failing under volume.
• You want tasks, reminders, and mobile updates without building your own macros.

When you hit that point, try one pipeline-first tool before you run a month of demos. Pipedrive is a common shortlist pick for deal stages and overdue next steps; confirm trial length and seat pricing on their site. For a wider comparison, our CRM roundup for small teams prioritizes pipeline clarity and data you can keep clean ahead of long feature matrices.

Practical selection criteria at this stage:

• Can you import or recreate active deals in an afternoon?
• Will everyone actually log next steps on mobile?
• Does the default pipeline match how you sell, or will you fight the tool’s vocabulary?
• What is the bill at your seat count in 12 months, including the add-ons you will want next quarter?

You do not need marketing automation, lead scoring, or AI summaries to solve “I do not know what needs attention today.”

Ready for more structure vs over-engineering

You are ready to add structure when:

• The weekly stale-deal review is already happening (even painfully) and the pain is scale or handoffs more than discipline gaps.
• Leadership is making cash or hiring decisions from pipeline guesses and wants one shared number to stress-test.
• A new seller or ops person is starting in the next 90 days and needs a system that does not depend on founder recall.

You are over-engineering when:

• You are configuring automations before everyone updates stage and next step reliably for a month.
• You are buying tiers for reporting you will not trust because half the deals are outdated.
• You are modeling fifteen stages because it felt thorough without matching how buyers actually move.
• You treat the CRM as a substitute for saying no to bad-fit deals cluttering the pipe.

A reasonable sequence for a business at your stage: agree stages and fields, run the weekly review in whatever tool you already have, pick a CRM when handoffs or volume break the spreadsheet, then add automation after hygiene is boring.

Verdict (starter)

If deals close but you cannot see what needs attention until it is too late, treat that as a visibility problem: a small, shared pipeline and a weekly stale-deal habit. Shop for enterprise software only after that habit exists.

Get honest about signals (overdue next steps, quiet deals, planning from memory), run the minimum fields and rhythms above for a few weeks, then buy a CRM when more than one person must live in the same pipe or your first sales hire starts. Until then, a light process you will run beats a heavy one you will not.