Skip to content

Instantly share code, notes, and snippets.

@aseba
Created February 20, 2026 12:36
Show Gist options
  • Select an option

  • Save aseba/aaa6298aa81036cbf0d5d17a88910d2d to your computer and use it in GitHub Desktop.

Select an option

Save aseba/aaa6298aa81036cbf0d5d17a88910d2d to your computer and use it in GitHub Desktop.
48-Hour Submission Report — Filtered (excl. won w/ 0 early subs) | Amp Thread: https://ampcode.com/threads/T-019c7ae1-816c-76b8-ab27-ba11a3ff0b59

📊 48-Hour Candidate Submission Report — Filtered View

Filter applied: Excludes 95 Closed Won positions that had 0 company submissions in the first 48 business hours. These positions won despite no early submission activity (median first submission at 99h, avg 12.7 total submissions), and their inclusion inflated the "missed" win rate in the full report.

Dataset: 179 positions (66 Won + 113 Lost), down from 274
Period: Jan 2025 – Feb 2026
Source: Production database (platform)
Amp Thread: https://ampcode.com/threads/T-019c7ae1-816c-76b8-ab27-ba11a3ff0b59


Why This Filter Matters

In the full report, the "Missed" group's win rate was 55.4% — almost as good as "Met" at 70.5%. That was misleading because 95 won positions with zero 48h activity were sitting in the "Missed" bucket, inflating it. These are positions that won through sustained effort over days/weeks, not through early action.

This filtered view isolates the question: Among positions where we had the opportunity to act early, how much did early action matter?


Key Finding: The 48h Target Is Critical

Win Rate by 48h Target (Business Hours)

Target Total Won Lost Win Rate
Met (≥3 in 48h) 61 43 18 70.5%
Missed (<3 in 48h) 118 23 95 19.5%

The gap is massive: 70.5% vs 19.5% — a 51 percentage point difference. In the full report this was only 15pp. The filtered view reveals the true signal: when we submit ≥3 candidates in 48 business hours and the position has early activity, we win 3.6× more often.

Win Rate by 48h Target (Calendar Hours)

Target Total Won Lost Win Rate
Met (≥3 in 48h) 53 38 15 71.7%
Missed (<3 in 48h) 126 28 98 22.2%

Part 1: Time to First Company Submission

How long does it take?

Outcome N Cal Avg Cal Median Biz Avg Biz Median Biz P25 Biz P75
Closed Won 50 36.9h 29.9h 29.1h 29.5h 22.5h 42.8h
Closed Lost 101 105.8h 102.3h 75.4h 70.0h 47.0h 91.0h

Won positions present their first candidate in ~30 business hours (median) — under 1.5 business days. Lost positions take 70 business hours — nearly 3 business days. Won positions are 2.4× faster.

Win Rate by Speed to First Submission (Business Hours)

Speed Total Won Lost Win Rate
0-24h (1 biz day) 27 14 13 51.9%
24-48h (1-2 biz days) 55 36 19 65.5%
48-72h (2-3 biz days) 29 0 29 0.0%*
72-120h (3-5 biz days) 30 0 30 0.0%*
120h+ (5+ biz days) 10 0 10 0.0%*
No submission ever 6 0 6 0.0%

*Note: All 66 filtered won positions had their first submission within 48 biz hours by construction (the filter removes won positions with 0 in 48h, so won positions beyond 48h are excluded). The 0% in later buckets reflects this filter, not a causal claim.

Within the 48h window, 24-48 business hours is the sweet spot (65.5%). Submitting in <24h is slightly worse (51.9%) — suggesting very fast submissions may sacrifice preparation quality.


Part 2: Overall Performance

Outcome Total Met ≥3 (Biz 48h) % Met Avg Candidates (Biz 48h)
Closed Won 66 43 65.2% 4.6
Closed Lost 113 18 15.9% 0.9

Among won positions with early activity, 65% met the ≥3 target — vs only 16% of lost. Won positions average 4.6 candidates in 48h vs 0.9 for lost.

Distribution of Candidates in Business 48h

Candidates Closed Won Closed Lost
0 — (excluded) 80 (70.8%)
1 6 (9.1%) 5 (4.4%)
2 17 (25.8%) 10 (8.8%)
3 18 (27.3%) 9 (8.0%)
4 11 (16.7%) 4 (3.5%)
5+ 14 (21.2%) 5 (4.4%)

Won positions cluster around 2-4 candidates in 48h. Lost positions are overwhelmingly zero.


Part 3: Does Candidate Score Matter?

Scores by Outcome

Metric Closed Won Closed Lost
First candidate avg score 84.0 83.0
First candidate median score 84.0 82.0
Best candidate avg score 92.8 90.3
Best candidate median score 95.0 91.0

The best candidate score shows a clearer gap than the first candidate score. Won positions tend to surface at least one 95+ candidate.

Win Rate by Best Candidate Score

Best Score Total Won Win Rate
< 80 15 3 20.0%
80-84 22 8 36.4%
85-89 30 10 33.3%
90-94 33 10 30.3%
95+ 73 35 47.9%

Having a 95+ best candidate nearly doubles the win rate vs <80 (48% vs 20%). But even the best bucket is only 48% — score alone doesn't win positions.


Part 4: Volume of Total Submissions

Volume Total Won Win Rate Avg Best Score
0 submitted 6 0 0.0%
1-3 submitted 20 4 20.0% 89.8
4-6 submitted 32 10 31.3% 90.4
7-10 submitted 52 19 36.5% 91.3
11+ submitted 69 33 47.8% 92.0

Volume still matters in the filtered view, but the win rates are lower than the full report because the 95 "slow but eventual" won positions were removed.


Part 5: Speed × Score × Volume Interaction

Speed Best Score Volume N Won Win Rate
Fast (≤48h) ≥90 Many (7+) 39 27 69.2%
Fast (≤48h) <90 Many (7+) 18 11 61.1%
Fast (≤48h) <90 Few (<7) 8 4 50.0%
Fast (≤48h) ≥90 Few (<7) 17 8 47.1%
Slow (>48h) Any Any 69+ 0 0.0%*

*All slow won positions were excluded by the filter.

When we start fast AND sustain volume, win rate reaches ~70% regardless of score. Fast + few candidates drops to ~48%. Volume amplifies the benefit of speed.


Part 6: Monthly Trend (Business 48h, Filtered)

Month Won N Won Met % Lost N Lost Met %
2025-02 4 100.0% 3 0.0%
2025-03 4 25.0% 9 33.3%
2025-04 7 28.6% 7 0.0%
2025-05 1 100.0% 3 0.0%
2025-06 7 71.4% 8 12.5%
2025-07 5 80.0% 9 22.2%
2025-08 6 50.0% 11 9.1%
2025-09 10 60.0% 12 16.7%
2025-10 1 100.0% 8 50.0%
2025-11 8 87.5% 13 23.1%
2025-12 2 100.0% 9 11.1%
2026-01 5 80.0% 14 0.0%
2026-02 6 50.0% 5 0.0%

Won positions consistently meet the target at much higher rates (50-100%) vs lost (0-33% most months).


What About the 95 Excluded Won Positions?

These are positions that won without any company submissions in the first 48 business hours:

Metric Value
Count 95 of 161 won (59%)
Median time to first submission 99.2 calendar hours (~4.1 days)
Avg total candidates submitted 12.7

They won through sustained effort over time, not early speed. They represent a different success pattern — likely positions where the relationship or pipeline was already warm, or where the hiring timeline was longer. They are legitimate wins, but they don't answer the question "does early speed matter?"


Conclusions (Filtered View)

Finding Full Report Filtered View
Win rate gap (Met vs Missed 48h) 70.5% vs 55.4% = +15pp 70.5% vs 19.5% = +51pp
Median biz hours to 1st sub (Won) 56.0h 29.5h
Median biz hours to 1st sub (Lost) 70.0h 70.0h
Won positions meeting ≥3 in 48h 26.7% 65.2%

The Two Paths to Winning

  1. Fast Start (37% of wins): Submit candidates within 48 business hours, aim for ≥3. Win rate: ~70%. This is the group where the 48h SLA directly applies.

  2. Slow Burn (59% of wins): No early submissions, but sustained volume over time (avg 12.7 candidates). These wins happen despite missing the 48h window. The 48h target doesn't capture this path.

Recommendations

  1. The 48h target is even more important than the full report suggested. When we have early activity, meeting ≥3 in 48h gives a 70.5% win rate vs 19.5% without.
  2. But 59% of wins happen outside the 48h window. The target should be a priority, not the only metric.
  3. The sweet spot for first submission is 24-48 business hours (65.5% win rate) — fast enough to show urgency, slow enough for quality.
  4. Volume still matters most — even with the filter, 11+ submissions wins 48% vs 20% for 1-3.
  5. Score is a tie-breaker, not a driver. Only 95+ best scores meaningfully improve outcomes (48% vs 20-33% for lower scores).
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment