Case study · Zoho CRM + Zoho One
A sales forecast you can actually trust.
A services business ran its whole sales team on Zoho CRM, but the pipeline had ten stages nobody agreed on and every deal's probability was typed in by hand. Finance couldn't trust the number, so the team stopped trusting the CRM.
I rebuilt it: five clear stages with a rule for each, a four-question deal scorecard anyone can fill in after a call, and a probability Zoho works out by itself — plus the follow-ups that fire the moment a deal is won.
Ten stages, and no two people meant the same thing
The CRM had grown for years. Ten sales stages, several meaning roughly the same thing, and no written rule for when a deal was allowed to move from one to the next. Two salespeople could look at the same deal and put it in different columns — and both be right.
Probability was worse. It was a free-text number a rep set from a gut feel, so the weighted forecast was really a survey of how optimistic the team felt that week. Finance quietly rebuilt it in a spreadsheet. The sales team, sensing the CRM wasn't the source of truth, put less into it — which made it less true. A loop that only goes one way.
None of that needed new software. Everything required was already sitting inside the Zoho One subscription they were paying for.
Part one
Five stages, one rule each
A stage now means something specific: there is a written gate a deal must clear before it moves. Same deal, same stage, no matter who's looking at it.
CRM · Deals · Pipeline view
| Stage | What has to be true before a deal moves here | Base probability |
|---|---|---|
| Qualifying | There's a real need, we know who holds the budget, we know the timeline, and a next step is booked. | 10% |
| Proposal & Negotiation | A proposal is with them, scope and price are broadly agreed verbally, and a sponsor inside the business is confirmed. | 40% |
| Committed | A signed contract or purchase order is in hand. Not "they said yes on the phone" — in hand. | 75% |
| Closed Won | Won. Delivery is handed over automatically from here. | 100% |
| Closed Lost | Lost, with a reason recorded from a fixed list — so losses can be counted, not guessed at. | 0% |
The old stages didn't vanish — every existing deal was mapped onto the new five in a controlled pass, so the history still adds up.
Part two
Four questions after every call
The stage says how far along a deal is. The scorecard says how good it is. Four questions, each scored 0, 1 or 2 — about thirty seconds of work after a call, and nothing else to fill in.
CRM · Deals · Halcyon Systems
| Question | Score 0 | Score 1 | Score 2 |
|---|---|---|---|
| Fit | 0Poor fit | 1Partial fit | 2Strong fit |
| Sponsorship | 0No internal advocate | 1An advocate is emerging | 2A confirmed champion |
| Authority | 0No access to the decision-maker | 1Only via an intermediary | 2Direct access |
| Budget | 0Unknown or none | 1Indicative figure only | 2Confirmed |
The team got this as a one-page cheat sheet. That's the whole training requirement — no new tool to learn, no new tab to open.
Part three
Zoho works out the probability itself
Each stage has a base probability. A strong scorecard lifts a deal within its stage — but it can never leap it, so a well-liked deal in Qualifying can't masquerade as a signed one.
Every scorecard point above 4 adds 3 points. A 4-out-of-8 deal sits flat at the base.
Tops out well below the next stageEvery point above 4 adds 4 points, so a perfect scorecard is worth +16.
Capped at 60%Signed paperwork already carries most of the confidence, so the scorecard only nudges it.
Capped at 80%Worked through, using the deal above
Nothing is capped at 100% except a deal that's actually won. Nothing sits at 0% except a deal that's actually lost.
Why a cap matters
Without one, optimism compounds: a rep scores their own deal generously and the forecast inherits it. A hard ceiling per stage means the only way to a high number is to actually move the deal forward.
Calibrated against real history
The base percentages aren't invented. I read the business's own closed deals — years of them — and checked the win rate by deal type, because renewals and expansions close far more often than brand-new business. The model has to reflect that or it flatters new logos.
Part four
The follow-ups nobody has to remember
A tidy pipeline is only worth something if things actually happen off the back of it. These run in the background, inside Zoho, on the events that already occur.
A deal goes quiet — a task appears
If a deal sits untouched too long, Zoho raises a follow-up task on the owner, with the deal attached. No manager chasing a report to find out what's gone cold; the reminder finds the rep.
Trigger: no activity for X daysA deal is won — delivery starts itself
The moment a deal moves to Closed Won, a job is created on the delivery board in Zoho Projects, pre-filled from the deal — client, value, dates, what was sold — and set to New. Nobody re-types anything, and nothing is won-and-then-forgotten.
Trigger: stage = Closed WonSales changes something — ops sees it
Key fields flow from the CRM deal to the delivery job, so when sales flips "ready for delivery" from not yet to yes, the ops board updates on its own. One number, one status, two teams, no chasing.
Trigger: field updated on the dealA deal is lost — the reason is captured
Closing a deal as lost requires picking a reason from a fixed list. Small thing, big payoff: after a quarter you can see why you lose, in numbers, instead of retelling anecdotes in a meeting.
Trigger: stage = Closed LostWhat the hand-off looks like
CRM · Deals
automatically
Projects · Delivery board
Illustration — made-up clients and dates. Both sides live in the same Zoho One subscription, so the hand-off needs no third-party tool and no extra licence.
Part five
Now the forecast is just… there
Because every deal carries an honest probability, the weighted forecast is a live number rather than a monthly spreadsheet exercise. Same figure on the manager's screen, the founder's screen and the finance screen.
CRM · Analytics · Sales forecast
Changed without breaking anything
Restructuring a live CRM that a sales team works in every day is the risky part. Most of the effort went here — not into the build, into making the switch boring.
Built on a copy first
- The whole rebuild was constructed and tested in a sandbox copy of the CRM
- Nothing touched the live system until it was proven there
- Every existing automation was audited against the new stages before the switch
- One broken rule was found and fixed in the copy, not discovered on a Monday morning
Nothing moved without a net
- A full export of every deal taken before anything changed
- Old stages mapped to new ones on paper, and signed off, before a single deal moved
- Deals remapped in a controlled pass — preview, confirm, then verify
- Years of closed history preserved, so reporting still reconciles
Built with Zoho's own tools
- Native fields, formulas, pipelines and workflows — no bolt-on software
- No new licences and no extra monthly cost
- Nothing that stops working if I walk away
- Their team can change a rule themselves; I showed them how
Cleaned up on the way through
- Every custom field on the deal measured for how often it's actually filled in
- Dozens sat near-empty for years — listed for retirement, with the client deciding
- Duplicated and misspelt legacy fields flagged rather than silently deleted
- The form a rep opens got shorter, so it gets filled in
Does your forecast survive contact with finance?
If your pipeline stages mean different things to different people, or probability is a number someone types in, the forecast is a guess with a decimal point. That's fixable — usually inside the CRM you already pay for. Let's have a look at yours.
Book a callOr email harshan@flowlamina.com