Scratchpad vs Dooly for AI Agency Pilot-to-Contract Sales

The most common failure mode for an automation agency's pipeline is a stack of opportunities stuck at 'pilot in progress' for two months past when the pilot should have wrapped. Nobody moved the stage because nobody was sure whether it had actually stalled or was just waiting on the client's engineering team, and updating the record felt like a distraction from the technical work itself.

That ambiguity is exactly what a stale pipeline looks like from the outside, and it is why a weekly review of open pilots matters as much as a weekly review of net-new opportunities. A founder-led sales motion makes this worse rather than better, since the same person running the pilot is also the one who is supposed to remember to update Salesforce about it.

The details that decide whether a pilot converts, which specific workflow steps the automation replaced, how much manual review time it actually saved, and who on the client side needs to sign off, are usually discussed in a technical debrief call, not a sales call. If that debrief never gets typed up anywhere structured, the person writing the statement of work is reconstructing it from a Slack thread days later.

Dooly fits that debrief call directly: a solutions engineer can run a note template built around pilot outcomes, link the measured time savings to the opportunity, and push it straight into the Salesforce record the account executive is working from.

An agency running six or eight pilots at once needs a fast way to see which ones are approaching their review date and which have gone quiet. Scratchpad's grid view lets a founder or BD lead scan every open pilot's stage and next step in one sitting, rather than opening each opportunity separately to check on it.

Its Deal Spotlight view is useful here specifically because it flags opportunities with no recent activity, which for a pilot-based sales motion is often the earliest warning that a champion has gone cold or moved to another project.

Scan open pilots each week for these signals:

• Pilots approaching their review date, so nobody is surprised when the client expects a result.
• Pilots whose stage has not moved even though the work continues, which usually means nobody confirmed whether they stalled or are waiting on client engineering.
• Pilots with a blank or outdated next step in the opportunity record.
• Pilots where the technical debrief was never typed up, leaving the statement of work to be written from memory.

Running unpaid or discounted pilots costs real engineering time before any revenue lands, which is exactly the kind of spend a burn multiple is meant to catch. For an automation agency under $10 million in ARR, a burn multiple under 1.1x is the good end of the range, and anything approaching 3.8x signals that too much is being spent for too little net new revenue¹.

A pipeline full of stalled, unpriced pilots is one of the fastest ways to push that number in the wrong direction, since engineering time keeps getting spent while nothing converts.

Say a firm's pilots convert at a healthy rate once they are approved, but half of them never get approved to start in the first place. That is not a CRM tooling problem; it is a qualification problem, and neither Scratchpad nor Dooly will fix a sales process that agrees to run pilots for prospects who were never going to buy.

If pilots that do start convert reliably, the actual bottleneck is visibility into which ones are stalling and why, which is the problem both tools are built to solve, starting with whichever one, capture or bulk editing, is currently costing the most account executive time.

A useful gut check before buying either tool: list every pilot that has gone quiet for more than two weeks and ask whether the agency actually knows why. If the answer is usually a specific, remembered reason, the team is tracking pilots fine and just needs a faster place to log it, which points to Scratchpad. If the answer is usually a shrug, the information was never captured in the first place, which points to Dooly.