Koji vs Jisc Online Surveys (BOS) for Course Evaluation 2026

Short answer: Jisc Online Surveys — the tool many UK and European universities still call BOS (Bristol Online Surveys) — is a dependable, UK-hosted, GDPR-conscious general-purpose survey platform. For a short closed-question unit evaluation that you export to Excel or SPSS, it does the job at modest cost, and a large share of UK academics already know how to use it. But it is a survey builder, not a purpose-built course-evaluation system. It collects static Likert grids and free-text boxes that a human still has to read, code, and turn into action by hand. Koji is built specifically for course and student evaluation: AI-moderated conversational interviews that ask intelligent follow-ups, automatic thematic analysis of every open-text answer, bias-aware standardized moderation, and built-in closing-the-loop action tracking — with EU data residency. If your evaluation is ten Likert items and a comment box, Jisc Online Surveys is perfectly adequate. If you want deeper evidence with far less manual analysis — and audit-ready quality-cycle documentation — Koji is the stronger choice.

What Jisc Online Surveys is, and what it does well

Jisc Online Surveys began life as Bristol Online Surveys (BOS), developed at the University of Bristol, and is now operated by Jisc, the UK not-for-profit digital agency for tertiary education and research. It is an entirely web-based survey tool licensed by a large number of UK and international institutions, typically through CHEST/Jisc agreements. There are three licence tiers, and Jisc states that all tiers have access to the same feature set; as of publication, full per-seat list pricing was not published as a single public figure and is negotiated through those agreements.

Its strengths are real and worth acknowledging honestly:

Breadth of question types and survey logic. It supports a wide range of question formats, plus piping, routing/branching, survey access control, and email distribution (with a standard sending capacity of up to 5,000 on project and organisation licences). There are no hard limits on the number of surveys you create or respondents you collect.
Clean export for analysis. Results export to Excel and to statistics packages such as SPSS, which makes it a familiar choice for institutional-research teams and quantitative social scientists.
Strong compliance posture. Jisc acts as data processor while the institution remains data controller, operating under UK GDPR and the Data Protection Act 2018. The service is ISO 27001 certified and UK-hosted, and Jisc states that new features must meet WCAG 2.2 AA accessibility.
Versatility. The same tool handles research surveys, staff pulse checks, organisational-change evaluation, and student feedback — one licence, many uses.

For a great many jobs, that is exactly what an institution needs. None of what follows is a knock on Jisc Online Surveys as a survey tool.

Where it falls short for course evaluation specifically

Course and programme evaluation is a specialised job, and a general survey builder leaves the hard parts to you:

Static instrument, no probing. A respondent sees a fixed Likert grid and a comment box. When a student writes "the labs were disorganised," nothing asks which labs, when, or what would have helped. The most useful detail is exactly what a static form never captures.
Manual qualitative analysis. Open-text comments arrive as a spreadsheet column. Someone — often an over-stretched director of teaching — must read, code, and theme hundreds or thousands of comments by hand every cycle. At scale this is where evaluation insight quietly dies.
No bias-aware standardisation. Free-text and Likert data carry well-documented biases, and a generic survey tool offers no moderation layer to standardise how feedback is elicited across courses and cohorts.
Not built around the course/instructor hierarchy. Dedicated evaluation platforms model courses, instructors, terms, and anonymity-threshold reporting (suppressing results below a minimum response count). In a general survey tool you rebuild that structure by hand each cycle.
No closing-the-loop tracking. There is no native way to record what action a feedback theme triggered, or to produce "you said, we did" evidence — the very thing accreditation reviewers (QAA, ESG, NVAO and peers) increasingly ask for.

In short: Jisc Online Surveys collects the data competently. Turning that data into standardised, analysed, audit-ready evaluation evidence is left entirely to your staff.

How Koji approaches the same job

Koji is purpose-built for course and student evaluation and is designed around the parts a survey tool leaves manual:

AI-moderated conversational interviews. Instead of a static grid, each student has a short, adaptive conversation that asks evidence-seeking follow-ups — turning "the labs were disorganised" into a specific, actionable account.
Automatic thematic analysis. Every open response is themed and aggregated automatically, so a director of teaching reads a structured synthesis rather than a 2,000-row spreadsheet. (For the evidence on how reliable LLM thematic analysis can be, see our methodology piece [linked below].)
Bias-aware, standardised moderation. The same calibrated interview logic runs across every course, so feedback is elicited consistently rather than depending on who wrote which survey.
Formative and summative in one place. Mid-semester formative pulses and end-of-term summative evaluation share one engine, so problems surface while there is still time to fix them.
Closing-the-loop action tracking. Themes link to recorded actions, producing the longitudinal, "you said, we did" documentation that quality cycles and accreditors expect.
EU data residency and GDPR by design. Data is handled under EU/GDPR terms, which matters for EU/EEA institutions weighing data-residency questions.

Side-by-side comparison

Dimension	Jisc Online Surveys (BOS)	Koji
Primary design	General-purpose survey builder	Purpose-built course/student evaluation
Question format	Static Likert + open-text, with branching/piping	Adaptive AI-moderated conversation with follow-ups
Follow-up probing	None (logic is pre-scripted)	Yes — intelligent, evidence-seeking probes
Qualitative analysis	Manual coding in Excel/SPSS	Automatic thematic analysis at scale
Bias-aware standardisation	Not built in	Standardised, calibrated moderation across courses
Formative + summative	Possible, but built by hand each time	Native, shared engine
Closing the loop / action tracking	None	Built-in "you said, we did" tracking
Course/instructor hierarchy + threshold reporting	Build it yourself	Native
Data residency	UK-hosted (ISO 27001)	EU data residency, GDPR by design
Accessibility	WCAG 2.2 AA	Accessibility-focused conversational UI
Pricing model	Tiered via CHEST/Jisc agreements; list price not public	Institutional; contact for pricing
Best for	Low-cost, versatile closed-question surveys	Deep, analysed, audit-ready course evaluation

Honest take: when Jisc Online Surveys is the better choice

Be fair to the incumbent. Jisc Online Surveys is genuinely the right tool when:

Your institution already licenses it and your evaluation need is a simple closed-question form.
You want one low-cost tool for many survey types — research ethics surveys, staff pulses, event feedback — not just course evaluation.
Your workflow depends on raw data export to SPSS for your own statistical modelling.
You have a tight budget and the conversational depth, automatic theming, and action tracking simply are not priorities this year.
You specifically need a UK-hosted processor and are comfortable managing the qualitative analysis in-house.

If that describes you, Jisc Online Surveys is a sensible, defensible choice and you do not need to switch.

Where Koji wins

Koji is the stronger fit when the evaluation — not just the survey — is the point: when you want richer, probed feedback; when manual comment-coding has become unsustainable; when you need standardised, bias-aware collection across many courses; and when accreditation or quality-cycle reviews demand documented closing-the-loop evidence rather than a stack of unanalysed spreadsheets.

One engine, two products

Koji for Education runs on the same AI interview engine that powers the main Koji platform at koji.so for general user, customer, and market research. If your institution also runs alumni, employer, or staff-experience research, the conversational-interview approach transfers directly — the education product is simply that engine tuned for the course-evaluation quality cycle.

Bottom line

Jisc Online Surveys is a trustworthy, accessible, UK-hosted survey builder, and for simple closed-question feedback it is hard to fault on value. But course evaluation is more than data collection — it is collection, analysis, standardisation, and action. Koji is built for all four, which is why institutions that have outgrown manual comment-coding and want audit-ready evidence tend to move beyond a general survey tool. Match the tool to the job: a form for forms, a purpose-built platform for evaluation.

What one evaluation cycle actually looks like

The difference between the two tools is clearest in the work after the survey closes. In a typical end-of-semester cycle with Jisc Online Surveys, an administrator builds (or copies) a survey per module, distributes links, chases response rates, then downloads a spreadsheet per unit. The closed questions summarise themselves; the open-text comments do not. Someone reads them, tries to code recurring issues consistently across dozens of modules, writes a summary, and — if there is time — drafts recommendations. Whether any of those recommendations were acted on, and whether students noticed, is rarely captured anywhere the next review can find it.

With Koji, the same cycle runs differently: a standardised interview goes out for every unit, students have a short adaptive conversation rather than tick a grid, and by the time the window closes the open responses are already themed and aggregated. The coordinator reviews a synthesis, links the significant themes to actions, and those actions and their follow-up sit in the same record that next year''s cohort report draws on. The data-collection step looks similar; the analysis-and-action steps — where most institutions lose time and evidence — are largely automated.

Migration and adoption

Switching tools is a real cost, and it is fair to weigh it. Institutions moving from a general survey tool to a purpose-built evaluation platform usually do so module-by-module or faculty-by-faculty rather than all at once, often running a pilot alongside the incumbent for a semester. Because Koji handles the course/instructor structure and reporting natively, the administrative setup that a general tool repeats every cycle is configured once. If your only barrier to richer evaluation has been the manual analysis workload, that is precisely the cost a purpose-built engine removes.