← Insights

Study · Enquiry experience · 2026

We sent the same enquiry to 30 institutions. Seven answered the question.

One candidate, one three-part question, up to four channels per institution, across 10 markets. We measured who replied, how fast, whether the reply answered what was asked, and whether anyone remembered the candidate.

Institutions out of 30 that…

{{ b.label }} {{ b.v }}

Edvion secret-shopper study, 2026, n = 30. Each institution counted once if any channel replied. "Replied at all" includes automated acknowledgements and chatbots.

23 of 30

gave no direct answer to the question asked.

14 of 30

did not reply within 48 hours.

16 of 30

publish two or more enquiry channels, each with its own queue.

The enquiry

Three questions, each a different kind of test.

One needs an eligibility judgement, one is factual, one needs a policy lookup. A generic reply can manage the second and skip the other two, which is exactly the gap we wanted to see. The line about deciding within two weeks tests whether anyone triages by urgency.

Subject: Enquiry about [programme] – entry requirements and next intake

Hello,

I am interested in the [programme] and I am applying from [country]. Three questions:

  1. Am I eligible with a [qualification] from [institution]?
  2. When does the next intake close, and what is the total tuition for an international student?
  3. Is there a scholarship I could be considered for?

I am comparing a few universities and would like to decide within the next two weeks.

Thank you

What we saw

Seven patterns that cost institutions their strongest candidates.

1 Redirected, not answered

The first reply was often a pointer to another inbox, team or form. The candidate had to start again.

2 Asked to repeat themselves

Replies asked for nationality, qualification or programme, all of which were already in the enquiry.

3 No memory across channels

Every institution we contacted on more than one channel treated each contact as a new person. One gave two different eligibility answers on two channels.

4 Chatbots were fast, then stopped short

Chatbots replied instantly, but none answered all three questions. Some handoffs to a person failed outside office hours.

5 Wrong programme, wrong campus

Several replies answered for a different programme or a different campus from the one asked about.

6 Nobody followed up

No institution followed up on its own. The only further emails were marketing.

7 Weekends are dark

An enquiry sent on a Saturday waited until the following week, including at an institution whose email signature promised weekend counsellors.

What good looks like

Every channel answered within four hours, the actual question addressed, nothing asked twice, and the candidate recognised on their next contact.

Institutions are not named. [Average reply times by channel: add from the final 14-day dataset.]

The sample

30 institutions, 10 markets.

{{ m.label }} {{ m.v }}

By type

Five kinds of provider.

{{ t.label }} {{ t.v }}

How we ran it

  1. One real candidate identity, one email address and one phone number throughout.
  2. The same enquiry, word for word, on each channel the institution publishes: web form, email, WhatsApp, chatbot.
  3. Roughly half sent inside local business hours and half outside, plus a separate weekend test.
  4. Second channels never mentioned the first, so we could see whether anyone connected the two.
  5. Counsellors' time respected: loops closed politely, and enquiries stopped if they triggered an application fee or document request.

Caveats

  • A snapshot of one intake period, with a 14-day response window.
  • The candidate was Singapore-based, so some Singapore institutions may have treated him as a local applicant.
  • A few enquiries used slightly different wording; these are noted in the full study.

Get the full study.

Method, every metric, reply times by channel and the patterns in detail. Or ask us to run the same audit on your own channels.

Audit my channels →

Thank you. The study is on its way to your inbox.