Careers teams know students need more interview practice than one-to-one appointments can provide. What’s harder to see is where students actually go wrong, and how much practice it takes to make them ready.
We analyzed 4,391 AI mock interviews completed by students at 95+ universities and colleges on VirtualSpeech in 2026. The picture is clear: students stay calm and answer the question, but they don’t research the company, and they struggle to show the impact of what they’ve done or connect it to the role.
Key findings
- Company knowledge is the biggest gap. “Knowing the company and why you want the role” averaged 2.8 out of 10, and 85% of students scored below 5.
- Students tell stories without a result. Using a clear structure such as STAR averaged 4.6, and showing the impact of their actions averaged 5.0.
- Composure is a strength. Staying composed was the highest-scoring skill, at 6.4.
- Few students are interview-ready on their first try. Only 18% scored 7 or above on their first mock interview.
- Practice changes that, but most students stop after one. 63% of students who practiced five or more times reached 7 or above, compared with 18% of those who practiced once. Yet 58% practiced only once.
This report covers each finding, what it means for how careers teams prepare students, and how to measure interview readiness across a cohort.
Finding 1: Students don’t research the company
The biggest gap is not nerves or communication. It’s preparation. When an interviewer asked students about the organization and why they wanted the role, they averaged 2.8 out of 10, and 85% scored below 5.
The same pattern shows up in overall scores. Students averaged 6.0 in an open-ended general interview, but 4.4 in an interview for a specific company, where knowing the organization matters.
This isn’t only inexperience. Working professionals practicing the same interviews scored almost exactly the same on company knowledge (2.9). Researching an employer is a habit, and most candidates haven’t built it, whatever their age.
Finding 2: Students tell stories without a result
Behavioral questions such as “Tell me about a time you worked in a team” are where structure matters most, and where students lose points. Using a clear structure such as STAR (Situation, Task, Action, Result) averaged 4.6. Showing the impact of their actions averaged 5.0.
In practice, students describe what happened and what they did, then stop. They rarely say what changed as a result, or put a number on it.
Finding 3: Students don’t connect their experience to the job
Linking experience to the role averaged 5.1, with 38% of students scoring below 5. Students often have relevant experience from part-time jobs, group projects, societies, or volunteering. What they miss is the final sentence: why that experience makes them right for this role.
Finding 4: What students do well
Students are better at the parts careers teams often worry about most:
- Staying composed was the strongest skill, averaging 6.4, with only 14% scoring below 5.
- Listening and answering the question asked averaged 5.9.
- Engagement and enthusiasm averaged 5.8.
That matters for how careers teams spend their time. Confidence workshops help, but the bigger gains are in preparation and structure: researching the employer, structuring examples, and linking them to the role.
Finding 5: Practice works, but most students stop after one
Only 18% of students scored 7 or above on their first mock interview, and across all student interviews, just 1% scored 8 or above. Practice changes that. Among students who practiced five or more times, 63% reached 7 or above at least once.
The problem is that most students don’t come back. 58% of students practiced only once, and 85% did all their practice on a single day.
A second attempt helps most when the first goes badly. Students who scored below 4 on their first try improved by 2.7 points on their second. Those who already scored 6 or more gained little from repeating the same interview, and would benefit more from a harder one or a different type of interview.
Students who practice more may also be more motivated, so these figures show an association, not proof. But the pattern is consistent: one mock interview is not enough.
What this means for careers teams
- Teach employer research as a skill. Give students a simple framework: what the organization does, a recent development, its values, and why they want to work there specifically. Then make them practice using it out loud.
- Make STAR stick, with the R. Students know what STAR stands for but stop before the result. Coach them to finish every example with an outcome, ideally a number, and one sentence linking it to the role. Our STAR method guide is a ready-made resource to share.
- Practice the interview they’ll actually face. Interviews built on the real job description and the student’s own resume are more useful than generic questions, and expose the company and role gaps above.
- Require more than one mock interview. Build at least three into career modules, application workshops or placement preparation. A required second attempt after a weak first one is the single cheapest improvement.
- Spread practice out. Most students practice on one day. Scheduling practice across a few weeks, ahead of the recruitment season, gives time for feedback to sink in.
- Use one-to-one time where it counts. Let AI mock interviews handle volume and first-round feedback, and focus advisers on students who are stuck or about to face high-stakes interviews.
How to measure interview readiness across a cohort
Most careers teams measure activity: appointments booked, workshops attended. Mock interview data lets you measure readiness instead.
- Define “ready”. For example, a score of 7 or above out of 10 on a mock interview built on a real role. Only 18% of students reach this on their first attempt.
- Take a baseline early in the year, with one mock interview per student.
- Look at skills, not just scores. Use the skill breakdown to see where your cohort is weakest, and target workshops there.
- Track practice and progress, including how many students practiced more than once and how their scores changed.
- Re-assess before the recruitment season, and report the change to faculty and leadership.
These benchmarks from our 2026 data give you a starting point to compare against:
| Measure | Benchmark |
|---|---|
| Average mock interview score | 5.3 out of 10 |
| Students scoring 7+ on their first attempt | 18% |
| Students who practiced only once | 58% |
| Students reaching 7+ after five or more interviews | 63% |
| Knowing the company and why you want the role | 2.8 out of 10 |
| Using a clear structure (such as STAR) | 4.6 out of 10 |
| Staying composed | 6.4 out of 10 |
Frequently asked questions
What do students struggle with most in job interviews?
Knowing the employer. In our data from 4,391 student mock interviews, “knowing the company and why you want the role” averaged 2.8 out of 10, far below every other skill.
How many mock interviews should students do?
At least three. Among students who practiced five or more times, 63% reached a score of 7 or above, against 18% of those who practiced once.
How can career services measure interview readiness?
Set a readiness threshold on a role-specific mock interview, take a baseline early in the year, track skill scores and practice, then re-assess before the recruitment season.
Do AI mock interviews help students?
They let every student practice as often as they need with instant feedback. Students who scored below 4 on their first attempt improved by 2.7 points on their second.
How we did this analysis
We analyzed 4,391 completed AI mock interviews by students at 98 universities and colleges on VirtualSpeech, from January 1 to October 13, 2026. Interviews included general interviews, interviews based on a student’s resume and a job description, interviews for a specific company, behavioral (STAR) interviews and higher-education admissions interviews.
- Scoring: AI scored each interview from 0 to 10 overall and on several skills. We grouped scenario-specific skills into eight broader interview skills.
- Practice and progress: some institutions share one login between several students, so practice and improvement figures use only students with their own account (698 students).
- Comparison: we compared students with working professionals practicing the same interviews.
- Excluded: anonymous free-trial sessions, internal and test accounts, and sessions with no valid score.
- Limits: scores come from AI, not human interviewers. We are running a study comparing our AI scores with certified coaches. Students who practice more may be more motivated, so practice figures show an association, not cause and effect.
All data was anonymized and aggregated.
Related reading:
