Case study
A private career college running six campuses in Atlantic Canada had steady programme traffic and an application form that finished last in every report. We rebuilt both, starting with eleven interviews rather than a redesign.
Named only with client approval Figures marked are illustrative Reply within one business day
Attribution · this quarter
GA4 · server-side312
Booked calls
3.1x
ROAS
−31%
CPL
OlDevs rebuilt the enrolment journey for a private career college running six campuses in Atlantic Canada, starting with session recordings, a first click study on the programme template and eleven interviews with recent applicants, two of them people who had abandoned the form part way through.The twenty-three question application became a three step flow that asks for a name, a programme and a phone number first, with the qualifying questions moved to a follow up call with admissions. The programme page template was rebuilt so accreditation, instructor bios, graduate outcomes and a plain answer on funding sit above the fold. Everything shipped as tested variants in a staged rollout, and application starts rose 3.2x while form abandonment fell 41%.
Key facts
The challenge
Programme pages pulled steady search traffic and paid campaigns added thousands of visits a month, yet the application form finished last in every report the college ran. It asked twenty-three questions before a prospective student could speak to a human, and the pages around it said almost nothing about funding, class sizes or what graduates went on to do. Admissions staff also felt the few enquiries that did arrive were poorly matched to the programmes on offer.
The application asked for education history, employment details and funding intentions in one long form. A prospective student with a straightforward question about a programme had no route to admissions that did not begin with filling all of it in.
Pages described curriculum and duration but said almost nothing about how study is funded, what class sizes look like or what graduates go on to do. Prospective students left to search elsewhere for answers the college was best placed to give.
Admissions reported that the few enquiries arriving were often for programmes the applicant was not eligible for or not really interested in. Staff time went into sorting mismatches rather than into conversations with people ready to enrol.
Every report showed the form finishing last, and none explained why. Without session recordings or interviews, the college was choosing between redesign opinions, and any change would have shipped everywhere at once with nothing to compare it against.
What we built
We started with research rather than a redesign: session recordings, a first click study on the programme template, and eleven interviews with recent applicants, two of them people who had dropped out mid form. The application became a three step flow asking for a name, a programme and a phone number first, with qualifying questions moved to a follow up call, while the programme template was rebuilt so accreditation, instructor bios, graduate outcomes and a plain answer on funding sit above the fold. Everything shipped as tested variants in a staged rollout so campuses could be compared honestly before the change went everywhere.
Step one asks for a name, a programme and a phone number. The remaining qualifying questions moved to a follow up call with admissions, so a prospective student reaches a person early and the college still collects what it needs to assess eligibility.
Accreditation, instructor bios, graduate outcomes and a plain answer on funding now sit above the fold, in the order the interviews said people look for them. The template is built to WCAG 2.2 AA and works on a phone as readily as a desktop.
Eleven interviews with recent applicants, two of them people who abandoned the form part way through, plus session recordings and a first click study on the programme template, gave the college evidence for each change before a line of code was written.
Changes shipped as tested variants across the six campuses rather than everywhere at once, so results could be compared before the college committed. Measurement was defined ahead of launch, including what counted as an application start.
Process
Watch and listen first
Session recordings and funnel analysis showed where the twenty-three question form lost people, and the drop off points gave us the questions to put to applicants rather than assumptions to defend.
Interview applicants and abandoners
We ran eleven interviews with recent applicants, two of them people who had dropped out mid form, which is where the funding, class size and graduate outcome gaps surfaced most clearly.
Test the template before rebuilding it
A first click study on the programme template showed where prospective students expected to find accreditation, funding and outcome information, and how far that sat from where the page actually put it.
Rebuild the form and the template
The application became a three step flow with qualifying questions moved to a follow up call, and the programme template was rebuilt to WCAG 2.2 AA with a working demo the college reviewed every week.
Roll out in stages and compare
Variants went live campus by campus with measurement agreed in advance, so the college could compare like with like and admissions could feed back on enquiry quality before the change went everywhere.
Stack
Research
Measurement
Front end
Content
Ways of working
3.2x
Application starts
−41%
Form abandonment
11
Applicant interviews run
Outcomes
Application starts rose 3.2x once the first step asked for a name, a programme and a phone number instead of twenty-three questions.
Form abandonment fell 41%, because the questions that used to stop people now happen in a call with admissions instead.
Staff spend their day on prospective students who want to enrol rather than chasing half finished forms, and qualifying happens in conversation where it belongs.
11 applicant interviews run, plus session recordings and a first click study, so the college changed its pages on what applicants said rather than on internal preference.
In their words
Admissions spends its day on people who are ready to enrol rather than chasing half finished forms.
Head of admissions, private career college
Post-secondary education
FAQ
It did not here. The qualifying questions were not removed, they moved to a follow up call with admissions, where staff can ask them in context and correct a mismatch on the spot. Admissions reported spending their time on people ready to enrol rather than sorting half finished forms and poorly matched applications.
People who complete a form tell you what worked. People who left tell you what stopped them, which is the thing you are trying to fix. Two of the eleven interviews were with applicants who dropped out mid form, and their accounts of what felt intrusive or unanswerable shaped the three step design directly.
Six campuses draw different applicants, so a change that lifts one may do nothing at another. Shipping tested variants in stages let the college compare like with like, gather admissions feedback on enquiry quality, and commit to the new template with evidence rather than on the strength of an early week.
Ten weeks from first session recordings to a staged rollout across the campuses, including the interviews, the first click study, the rebuilt three step application and the new programme page template. The college saw a working demo every week, so approval happened alongside the build rather than after it.
Yes. The programme template and the application were built to WCAG 2.2 AA and tested on phones and desktops, including keyboard and screen reader paths through each of the three form steps. Accessibility was part of the build, not a remediation pass afterwards.
Let’s connect
Tell us what is not working. We’ll reply within one business day with how we would approach it and a tailored quote.