← Blog Method

Language school: how to triple enrolments with a WhatsApp agent

Sector pattern: a language school loses 40% of leads on weekends because of a slow contact form. Here’s how to fix it with a WhatsApp agent — inside Meta’s new policy.

Recurring pattern in language schools and training centres: the site has decent traffic, but the contact form converts poorly at weekends — precisely when candidates are at home thinking about signing up.

The cycle is familiar: candidate fills the form on Sunday evening, nobody replies before Monday 10am, and by then Berlitz already answered on WhatsApp in 4 minutes. Lead lost.

How we’d approach it

Replace the form with a WhatsApp Business agent — inside Meta’s new policy, with a strictly educational scope (booking, course info, level tests). The agent does three things:

flowchart LR A([Visits the site]) --> B[Clicks
Enrol] B --> C[Opens WhatsApp
with context] C --> D[Agent asks
the essentials] D --> E{Level
known?} E -->|No| F[Sends short
online test] E -->|Yes| G[Suggests 3 compatible
class slots] F --> G G --> H([Reserves a spot
and sends proforma])

Three details that make the difference:

  • Short level test (10 min, gamified). Without it, the agent is guessing.
  • Reversible commitment. Spot held for 72h, no payment yet — reduces friction.
  • Clean handoff to a human if the candidate hesitates or asks about a special case.

Expected outcomes

In a school with reasonable lead volume and a well-calibrated system, the ranges sector benchmarks suggest, per quarter:

  • Completed enrolments: rise between +80% and +200% vs. the prior quarter (the base tends to be low in scenarios with slow forms).
  • Average first-response time: from several hours to seconds.
  • Weekend leads lost: from ~40% to single digits.
  • One reception person moves from working the funnel to handling special cases and retention — where the school couldn’t reach before.

What we wouldn’t automate

  • The in-person meeting for corporate courses (high ticket, long cycle).
  • Complaints or refund requests — always human.
  • Official certifications — involve complex paperwork a person reviews.

When this method makes sense

The pattern works in any short-course institution with > 50 leads/month and rolling enrolments (not only semester windows). Typical setup: one month to design the flow, two more weeks to calibrate. Payback usually shows up between week 8 and week 12 — right when the school starts asking "how did we ever do this before?".

Share

Related articles

Talk to QRA
Accepting new projects
© 2026 QRA · AI automation for SMBs Free diagnosis · no commitment