The problem in numbers
Filter your lost deals from last quarter for those that reached the proposal stage, and read the lost reasons. In most accounts, no decision and went quiet dominate that list, usually ahead of price and competitor combined. Then check the activity log of those deals: a typical pattern is one follow-up, sometimes two, then nothing. The quote that cost a discovery call, a scoping conversation and an hour of writing received less persistence than a newsletter.
This is the cheapest revenue problem you will ever fix. The prospect asked for the quote. The work is done. All the sequence does is make sure the question gets answered, and the answer, either way, beats the silence that poisons your forecast.
The design
The trigger is the stage change to Proposal sent. That presumes the stage means what it says: the quote left the building, which is exactly the kind of verifiable stage definition pipeline design argues for. If quotes go out from Smart Docs, the document's tracking tells you when the quote was opened, which is intelligence for the touches, and the send itself should coincide with the stage move.
The sequence is three touches with different jobs. Touch one, day two or three: a light check-in task for the rep, did it arrive, does anything need clarifying. Touch two, around day seven: a value nudge, one concrete addition to the case, a reference, an answer to an unasked question, never just checking in again. Touch three, around day fourteen: the honest closing question, asking directly whether this is going ahead, and making a no easy to say. In Pipedrive, build this as a workflow automation on the stage change: an immediate task, then delayed steps for the later touches, each creating the activity with the template linked. Where volume is high and values small, the second touch can be an automated templated email instead of a task.
Stop conditions keep the machine polite. The sequence must halt when the deal leaves the proposal stage, which delayed automation steps handle by re-checking their conditions before firing. Won, lost or moved back for rework all kill the remaining touches. And any real reply moves the conversation to human control: the rep completes or deletes the remaining tasks, and dates get re-anchored to what the prospect said. After touch three comes a decision, not a drift: close the deal lost with no decision, or park it deliberately to a named future date. The rotting playbook catches whatever tries to linger.
Implementation steps
First, tighten the stage: confirm Proposal sent means sent, and that reps move the deal the same day. Second, write the two templates with your best rep, short and specific, because template quality decides whether reps use the machine. Third, build the automation with its delays and condition checks, and test it with a dummy deal pushed through every exit: won, lost, stage back, reply received. Fourth, brief the team on one rule: the sequence is the floor, not the ceiling. Reps add human touches on top whenever they know something the machine does not. The wider craft of keeping such automations few, named and maintainable is covered in automations that matter.
How you measure that it works
Three numbers, monthly. Proposal-stage conversion: the share of deals entering the proposal stage that end won, which should climb within a quarter. Time in proposal stage: the median should fall, because answers arrive sooner in both directions. And the lost-reason mix: no decision should shrink as a share of proposal-stage losses, replaced by explicit reasons you can act on. All three come straight from the deal reports in Insights, and the third doubles as a health check on your lost reasons themselves.
Common failure modes
The robotic sequence: three template emails, no human judgement, and prospects feel processed. Keep the rep in touches one and three at minimum. The zombie sequence: automation fires on deals whose conversation moved on, because a stop condition is missing, and a prospect in contract review gets are you still interested. Test every exit path before go-live. The stage smuggle: reps park deals in the proposal stage before sending anything, and the sequence chases quotes that do not exist. The stage definition, enforced in the weekly review, is the fix. And the orphan template: the quote follows up while the quote itself is the problem. If touch responses keep raising the same objection, the fix belongs in your quoting setup, not in a fourth email.
Questions
Can Pipedrive send follow-up emails automatically?
Yes, on the plans that include email in workflow automation. An automation can send a templated email when its trigger and delay conditions are met. Whether follow-up two and three should be automatic emails or created call tasks is a design choice this playbook takes seriously.
How many follow-ups before we stop?
Three scheduled touches after the quote, then a decision, is a strong default: a check-in, a value nudge, and a clean closing question. Past three, response rates fall and irritation climbs. The end of the sequence is a decision point, not a fourth email.
What timing works between touches?
Day two or three for the first touch, roughly day seven for the second, day fourteen for the third. Adjust to your deal size: bigger decisions move slower. Anchor timing to real dates in the conversation whenever you have one, since an agreed decision date beats any fixed cadence.
Should follow-ups come from automation or from the rep?
The trigger and the timing from automation, the touch itself from the rep wherever deal values are meaningful. Automated task creation with a suggested template gets you consistency without robotic mail. Fully automated sends fit high-volume, low-value quoting.
How do we handle a prospect who says they need more time?
Move the conversation to a named date, then park the sequence until that date with a scheduled activity. The worst response to more time is three more template emails. The sequence exists to prevent silence, not to punish patience.