Seasonal ramps fail in predictable ways. The team treats it as a sourcing challenge, floods the top of the funnel in week one, and then discovers in week three that scheduling capacity, not candidate supply, is the binding constraint. By week five they are hiring whoever is still answering the phone.
This is the structure we use for compressed ramps, anywhere from 250 to 2,000 hires over six weeks. It assumes you know the number and the start date, which is usually all you get.
Work backward from onboarding capacity
The first number to establish is not how many candidates you need. It is how many new hires the receiving organization can absorb per week. If onboarding runs classes of 40 and the operation can support two classes a week, your ceiling is 80 starts weekly regardless of how good the funnel is.
We have watched teams deliver 140 accepted offers into a system that could onboard 80, and the overflow does not politely wait. Start dates slip, candidates who accepted three weeks earlier take other jobs, and the reneged rate climbs past 20 percent. Overproducing offers is more expensive than underproducing them because it destroys candidate trust you will need next season.
The funnel math, stated honestly
For a 400-hire ramp with typical volume-role conversion, the arithmetic is unforgiving. These ratios are from our own high-volume desks in Austin, Bogota and Manila across three seasonal cycles, and they hold within about 15 percent for hourly and entry-level technical roles.
- Applicant to screened: 42 percent, so 400 hires needs roughly 11,000 applicants
- Screened to interview scheduled: 38 percent
- Interview scheduled to interview completed: 71 percent, and no-shows are the biggest leak
- Interview to offer: 44 percent
- Offer to accept: 82 percent, dropping to about 68 percent if the offer lands more than four days after the interview
The two lines to attack are scheduled-to-completed and the offer latency penalty. Both are operational, not market conditions. A same-day offer after a completed interview recovers roughly 14 points of acceptance, which at 400 hires is the difference between needing 11,000 applicants and needing 13,500.
Week by week
The ramp has a shape. Front-loading everything into week one wastes the pipeline, because candidates sourced in week one who cannot interview until week three go cold. We stage it.
- Weeks 1 to 2: rebuild capacity first. Lock interview slots, confirm onboarding class sizes, publish the scorecard, activate TalentGraph rediscovery against last season's silver medalists.
- Weeks 2 to 4: peak sourcing. Beacon Sourcing runs outbound, Signal Screening handles pre-qualification over WhatsApp and SMS, Voice Outreach covers inbound call volume from job board traffic.
- Weeks 3 to 5: peak interview and offer throughput. This is where scheduling capacity binds. Hold it steady rather than chasing new sourcing.
- Week 6: backfill and reneges only. Assume 8 to 12 percent of accepted offers will not start and plan the buffer explicitly rather than discovering it.
Silver medalist rediscovery in week one is the single highest-yield move on the list. Candidates who reached final stages last season and were not selected convert at roughly three times the rate of cold applicants, and they need less screening because the prior evaluation still exists. On a well-maintained database that alone can cover 15 to 25 percent of a seasonal ramp.
Every ramp is won or lost on the constraint you did not measure. It is almost never candidate supply.
The three things that break first
Across the ramps we have run, the same three failure points recur. None of them are visible in week one, and all three are cheaper to prevent than to fix mid-ramp.
Interview no-shows come first. At volume, a 29 percent no-show rate is normal and a 12 percent rate is achievable with a confirmation sequence sent 24 hours and 2 hours before the slot over the channel the candidate actually uses. In LATAM and Southeast Asia that is WhatsApp, and the fix is close to free.
Hiring manager availability breaks second. In week four the operations leaders who agreed to interview panels are also preparing for the same seasonal peak. Get the slots into calendars in week one as recurring blocks, and staff a backup interviewer for every panel.
Quality drift breaks third and is the one nobody catches until January. Under throughput pressure, the effective bar slides. We hold it by keeping the rubric fixed and monitoring pass-rate by interviewer weekly. If one interviewer's pass rate moves 20 points above the panel median, that is a calibration conversation, not a productivity win.
Measure the ramp, not the recruiters
During a compressed ramp, report three numbers daily: offers accepted against the weekly onboarding ceiling, interview slots filled against slots available, and median hours from completed interview to offer sent. Everything else can wait for the retro.
Then run the retro while it is fresh, in the week after the last start date. The most valuable output is not the hire count. It is the updated conversion table, because next season's plan is only as good as this season's honest ratios.
Cal NakashimaVP of Delivery at Novexhire