Singapore SaaS case: 12 to 47 qualified meetings in 90 days
A Series B HR-tech client doubled their pipeline without adding headcount. Here's the system change that actually drove the lift — and the two things we got wrong before it worked.
In late 2025 a Series B HR-tech company hired us to fix their outbound. They had two SDRs working hard, a Hubspot instance with mostly empty fields, and a Q3 that closed at 12 qualified meetings — well off the 30/quarter the founder had committed to the board.
We finished Q1 2026 at 47 qualified meetings with the same two SDRs. No new tooling beyond what they already had. Here's what changed, what didn't, and the two false starts we logged on the way.
The diagnosis (week 1)
We pulled their last 90 days of outbound activity from Hubspot and listened to ~20 connect calls. Three things stood out:
- Their ICP was three ICPs in a trench coat. Targeting "HR leaders at any company 100–1,000 employees in Singapore" meant the same email was being sent to a 120-person agency and a 900-person bank. Reply rates: 1.4%.
- The sequence had 9 steps. Eight of them were emails. One was a LinkedIn message that nobody actually sent.
- No one knew what "qualified" meant. A "qualified meeting" in their CRM included anyone who agreed to a call. About a third turned out to be students, partners, or people from companies they'd never sell into.
The two false starts
False start 1 — "Just write better emails"
Week two, we rewrote the email copy. Sharper subject lines, clearer ask, dropped the gimmicks. Reply rate moved from 1.4% to 2.1%. Meetings barely moved. Lesson: copy is downstream of targeting. Better emails to the wrong people don't unlock pipeline.
False start 2 — "Just narrow the ICP"
Week four, we cut the target list from ~3,000 accounts to ~600 based on a stricter ICP. Reply rate jumped to 4.8%. Meeting volume dropped in absolute terms because the SDRs ran out of accounts in week six and started repeating outreach. Lesson: a tighter ICP demands a faster trigger-detection loop, otherwise you starve.
What worked (week 5 onward)
Three changes, in order:
1. Trigger-based account scoring
We rebuilt the target list in tiers driven by buying-signal:
- Tier A — companies with a new VP People hired in the last 90 days, or that had announced a hiring freeze (yes, those are an opportunity for HR-tech, not a disqualifier).
- Tier B — companies on Tier A criteria from 6+ months ago (warm but stale).
- Tier C — companies matching firmographic ICP only.
SDRs spent 70% of their time on Tier A, 25% on B, 5% on C. The list refreshed weekly via a Clay → Hubspot pipe we built in two afternoons.
2. A 5-step sequence with one phone call
We collapsed the 9-step sequence to 5 steps over 14 days: email, LinkedIn connection (no pitch), email reply-bump, LinkedIn message only if connected, phone call if there was any signal. The phone step was new. It accounted for ~40% of booked meetings in Q1.
3. A real definition of "qualified"
"Qualified" became: works at a target-tier account, holds a buyer-or-influencer title (VP People, Head of HR Ops, CHRO, COO at <500-person companies), and has a current process or tool relevant to the conversation. SDRs got a 4-question screening script. Disqualifying calls before the calendar invite went out dropped no-shows from 22% to 9%.
The numbers
| Metric | Q3 2025 | Q1 2026 |
|---|---|---|
| SDRs | 2 | 2 |
| Accounts contacted | 3,100 | 1,400 |
| Reply rate | 1.4% | 7.6% |
| Connect calls | 38 | 81 |
| Qualified meetings | 12 | 47 |
| Meetings → opps | 33% | 51% |
Less surface area, more depth, faster trigger loop.
What we'd do differently
Two things, in retrospect:
- Start with the qualification rubric, not the copy. Half the early "improvement" was illusory because we hadn't agreed on what we were optimising.
- Get phone in the sequence on day one. It's culturally correct in Singapore mid-market and it's where the conversion happens.
If you want a similar diagnosis on your own outbound, drop us a note. We can usually tell you in 30 minutes whether your problem is targeting, sequence, or qualification — those three account for most stuck pipelines.