Fill Your Cohort in 90 Days With a Two-Person Team

A two-person admissions office can still fill a cohort in 90 days — if it stops doing everything by hand

You do not need a bigger team. You need to stop the two people you have from spending their days on work a chatbot can do, and put every human hour into applicants actually close to enrolling. This plan splits 90 days into three 30-day phases — audit, activation, conversion — with a weekly task list, a table of what humans should touch versus what gets automated, and tactics specific to Canadian colleges, private career schools and universities working a shrinking calendar.

Assumption: it is late July or August, the cohort is short, and the fall or January intake is fixed with no budget for a third hire. Sector-wide enrolment pressure tracked by Universities Canada means most small admissions offices run this exact math every cycle.

Days 90–61: audit and triage — find out what is actually salvageable before you touch anything

Do not launch a new campaign in week one. Spend the first 30 days finding out which prospects in your CRM are worth working and which are dead weight — a two-person team that works every prospect equally will run out of hours before it runs out of applicants.

Week 1 — pull the funnel apart by stage

List every prospect in your CRM by stage: inquiry, started application, submitted application, invited to open house, attended, offer sent, deposit paid. Count how many are stuck at each stage and for how long. The national drop-off is steep at every step: 91% of website visitors never make first contact, 64% of first contacts never start an application, 42% of applications never turn into an open house signup, 35% of open house registrants never show up, 28% of attendees never submit a complete file, and 18% of complete files never convert to a final enrolment — for an overall visit-to-enrolment conversion of just 0.8% (source: content/zpd-bank.json#prospect-dropout-funnel). If your numbers land near this range, the problem is the funnel, not the program. Whichever stage is worst versus these benchmarks is where the two people spend week two.

Week 2 — triage the CRM into three piles

Split every open prospect record into three groups:

1. Work now — inquired or applied in the last 45 days, responded to at least one outreach attempt, fits the program's admission profile.
2. Dormant, worth one attempt — inquired 45–180 days ago, no response yet, but downloaded a viewbook, attended a fair, or came through a high-intent channel (program-name search, a referral).
3. Cut — inquired more than 180 days ago with zero engagement, wrong fit, or already enrolled elsewhere.

Do not work pile three. This is the single highest-leverage call in the first 30 days: a team that tries to "not waste a prospect" by calling everyone gives the same six minutes to a five-month ghost as to yesterday's inquiry. Cut the dead weight and reinvest the hours in piles one and two.

Week 3 — find where your own process is losing prospects

Walk your own inquiry-to-application path as a prospective student would: fill out the contact form and time what happens next. Contact form inquiries take a median of 72 hours to get a response, email 47 hours, while a chatbot answers in 3 seconds around the clock; phone gets picked up only 34% of the time even at a fast 3 minutes 20 seconds average (source: content/zpd-bank.json#response-time-by-channel). Two people cannot staff a phone line or inbox 24/7 — that is the case for automating first response before week four.

Week 4 — decide what you will NOT do this cycle

Write down the tactics you are skipping: a website redesign, a CRM migration, a paid social campaign built from scratch, every fair in the province. A team that tries all of it finishes week 12 having done none of it well. Pick the two or three channels with the best signup-to-attendance economics and commit the remaining hours there.

Days 60–31: activation — automate first response, reactivate dormant prospects, run one tightly scoped push

Once the CRM is triaged, the goal of the middle 30 days is a chatbot or automated sequence handling first response and basic qualification, so the two admissions staff only pick up conversations that are genuinely close to a decision.

Automate first response before you reactivate anyone

Turn on a website chatbot that answers program, tuition, deadline and admission-requirement questions immediately, qualifies the prospect (program, intended intake, current stage), and routes judgment calls — a transcript review, a scholarship exception, a program transfer — to your two staff. This addresses the biggest leak in the funnel: schools running an AI chatbot cut the visit-to-first-contact drop-off from 91% to 76%, a 167% increase in effective first contacts for the same traffic (source: content/zpd-bank.json#prospect-dropout-funnel) — prospects already paid for, previously lost to a slow reply, now captured without adding headcount.

Reactivate the dormant pile with a specific reason to respond

For pile two, send a short message tied to something concrete: an approaching deadline, a supplementary offer window, a new bursary, an invitation to a smaller info session. Route replies through the same chatbot-first workflow. Treat anything above 8–12% reactivation as a good result for a cold list, and do not chase prospects who skip a second attempt.

Run one open house or info session push, not three

Pick a single acquisition push for this phase and route registration through the highest-converting channel available. Chatbot-driven signups convert three to four times better than most alternatives: chatbot on the school site converts registrants at 18.4%, against 6.2% for a contact form, 4.8% for an email campaign, 3.7% for paid social and 2.1% for organic social (source: content/zpd-bank.json#jpo-registration-by-channel). By week eight, staff time should be almost entirely on prospects the chatbot has already qualified; still manually answering "what are your tuition fees" at that point means the automation step was skipped.

Days 30–0: conversion sprint — cut no-shows, apply yield management, protect the team

The final 30 days convert the prospects already in the pipeline — reducing open house no-shows, moving offers to deposits faster, and running the sprint without the two-person team burning out.

Cut open house no-shows with automated, personalized follow-up

No-show rates are driven almost entirely by follow-up quality, not prospect intent. No follow-up: 52% no-show. Email-only: 38%. SMS-only: 31%. Personalized chatbot follow-up: 19%. Chatbot-plus-SMS combo: 14%. A reminder naming the prospect's specific program: as low as 11% (source: content/zpd-bank.json#jpo-no-show-rate). Set up an automated sequence — confirmation, a reminder 48 hours out, a same-day nudge — and reserve personal staff time for prospects flagged high-intent.

Apply yield management once offers go out

Treat every stage between offer and deposit as manageable, not fixed. Set a deposit deadline with real urgency, follow up personally with admitted students who have not responded, and use a supplementary or late-offer round — the mechanism most Ontario colleges publish through ontariocolleges.ca — to fill remaining seats from your qualified-but-not-yet-offered pool instead of a fresh campaign. For deeper tactics, see our guide to yield management for student enrolment.

Protect the two-person team in the final stretch

A sprint that runs two people at full intensity for 30 straight days produces mistakes in the final week, when stakes are highest. Split the load: protected time blocks per person, a rotation for evening and weekend chatbot escalations, and treat the chatbot's qualification and reminder work as infrastructure, not something skipped when busy. A tool that complements the two staff by absorbing repetitive volume makes the sprint survivable; a tool they have to babysit adds work instead.

How to split 90 days between two people and automation

Judgment calls stay with the two humans; everything repetitive moves to the chatbot. Here is the split by phase:

That last row reflects a consistent pattern in chatbot deployment data: most prospect questions are repetitive, and only a small share genuinely need a person.

What a two-person team should NOT attempt in a 90-day window

Be explicit about trade-offs, because the instinct under pressure is to try everything at once.

• Do not rebuild the website or migrate CRM platforms mid-sprint — the disruption costs more than it recovers.
• Do not run more than one or two acquisition channels at once; a scattered push across paid social, organic social, email and print spreads two people across four half-executed campaigns instead of one well-run one.
• Do not personally call every dormant prospect; the reactivation message and chatbot qualification exist so staff time goes only to prospects who respond.
• Do not promise capabilities you have not deployed — a chatbot configured in an afternoon and billed as "handles everything" will escalate more, not less.

The result of doing this well

None of this requires new headcount — the ceiling on what a lean team can recover is set by the funnel, not by ambition. Institutions combining faster first response with fewer open house no-shows and disciplined triage see qualified prospect volume move from roughly 120 to 195 per month (+62%), cost per qualified prospect down about 38%, and a median 280% return over 12 months (source: content/zpd-bank.json#chatbot-roi-metrics). These come from institutions running chatbot deployments alongside funnel optimization — not a guarantee, but a realistic order of magnitude for a two-person team in one admissions cycle.

For the full-year version of this plan, see our 12-month admission campaign timeline for private schools; to audit your funnel first, start with our student recruitment funnel audit guide. For the broader strategy, see how to recruit more students in higher education.

FAQ

Can two people really fill a cohort in 90 days without hiring?

Yes, if the two people stop doing manual first-response and repetitive Q&A and reallocate that time to high-intent prospects. The constraint is rarely headcount — it is how many hours go to prospects who were never going to enrol.

What is the single highest-leverage change to make first?

Automating first response. 91% of website visitors never reach first contact, and a contact form reply can take 72 hours while a chatbot responds in 3 seconds, 24/7 (source: content/zpd-bank.json#prospect-dropout-funnel, #response-time-by-channel). No other change touches as many prospects for as little staff time.

Should we prioritize domestic or international prospects in a 90-day push?

Domestic prospects usually convert faster because they do not depend on a study permit timeline. International admissions involve provincial attestation letter allocations and processing times through Immigration, Refugees and Citizenship Canada that can extend well past 90 days — treat it as a parallel, longer-cycle track.

Does OSAP or financial aid affect the 90-day timeline?

It can, particularly close to a fall intake. Ontario students often wait on OSAP or provincial student aid confirmation before committing a deposit; other provinces have similar waits tied to their own aid programs and Canada Student Loans. Build a short buffer for prospects citing a pending aid decision rather than treating silence as a lost prospect.

Does data privacy law affect how we contact dormant prospects?

Yes — under PIPEDA federally, and Quebec's Loi 25 for institutions operating there, reactivation messages should respect the consent basis under which the contact was collected. Most CRM-sourced inquiries already carry consent for follow-up about the program they inquired on.

Two people, 90 days, one under-enrolled cohort — the plan works by removing repetitive work from the two humans, not by adding pressure to their calendar.