Contents

    People looking for a course no longer ask only Google. Four out of five people in Switzerland use AI, and more than half search specifically via AI platforms. Whether ChatGPT, Gemini or Google AI Overviews name your programme depends on three things: course pages with clear facts, the portals and lists the assistants read, and a measurement that shows where you stand. In our own measurement of 134 AI answers on continuing education in Zurich, 63 percent of cited sources were provider websites and 20 percent were education portals and review sites.

    This guide is for marketing leads at continuing education providers, academies and private schools in Switzerland: how you show up in AI answers, what you report to management and where to start when marketing so far has been a collection of one-off activities.

    How prospective learners look for a course today

    Continuing education is a mass market in Switzerland. According to the Federal Statistical Office, 45 percent of the population took part in continuing education in 2021. How these people search is shifting fast. In the IGEM-Digimonitor 2026, 80 percent of respondents say they use AI, 20 percentage points more than a year earlier. 57 percent search specifically via AI platforms, more often than via YouTube (34 percent) or social media (28 percent). 67 percent use ChatGPT at least occasionally.

    There are no robust Swiss figures yet on AI use in education decisions. There are from the US: in a study by UPCEA and Search Influence among 760 adult prospective learners, half use AI search at least weekly. 79 percent read Google's AI summaries, and 56 percent are more likely to trust institutions cited there. At the same time, 77 percent consider the institution's own website the most dependable source. Among high school students choosing a college, EAB measured 46 percent AI use, up from 26 percent in spring 2025. 18 percent removed a college from consideration because of AI answers.

    The search path therefore often looks like this: the person asks ChatGPT or Google which qualification will move them forward, gets providers suggested, checks them on portals and review sites, and only then lands on your course page. If you are missing at the first step, you are not compared at the second.

    Who AI names for continuing education in Zurich: our measurement

    On 16 November 2026 we asked six systems 24 questions in German about continuing education in Zurich: ChatGPT, Claude, Gemini and Perplexity with web search, plus Google AI Mode and Google AI Overviews with location Zurich. The questions cover six course categories, HR, project management, marketing, leadership, finance and digital, plus general questions on recognition, funding and reviews. Each question ran once per system. That produced 134 answers, because Google showed no AI Overview for nine questions.

    This is a one-day sample, not a quality ranking. It shows who the assistants name and which pages they read to do so.

    Who gets named

    For the 14 comparison questions, such as "Which providers for the HR specialist course with federal diploma in Zurich are good?", answers usually named four or more providers. The most frequent:

    ProviderComparison answers (of 76)Named by
    KV Business School Zürich49all six systems
    HWZ University of Applied Sciences in Business Administration Zurich33all six
    ZHAW30all six
    AKAD20all six
    Bénédict18all six
    MBSZ15five

    The qualification decides who appears

    For questions about a federal diploma (eidgenössischer Fachausweis), the assistants mostly name private schools. KV Business School Zürich appeared in all ten comparison answers on finance and fiduciary courses and in 11 of 12 on HR. For the marketing diploma, MBSZ led with 10 of 11. For CAS programmes and digital topics such as data science or digital transformation the picture flips: 11 of 11 answers named HWZ and 10 of 11 named ZHAW, KV Business School only once.

    For your measurement this means: split the prompt set by type of qualification. To the assistants, federal diplomas and CAS programmes are two different markets with different favourites.

    Where the information comes from

    128 of the 134 answers cited sources, 604 in total, counted per domain and answer:

    • 63 percent were provider websites, 84 different ones. The most cited were kv-business-school.ch (in 54 answers), zhaw.ch (35) and fh-hwz.ch (32).
    • 20 percent were education portals, directories and review sites. Leading were ausbildung-weiterbildung.ch (37 answers), berufsberatung.ch (26) and eduwo.ch (14). Just over half of the answers with sources relied on at least one such site.
    • 8 percent were associations, exam bodies and quality labels, 6 percent authorities such as SERI (SBFI) and cantonal career services.
    • Media and social networks hardly mattered: six links in total across 134 answers.

    The assistants take concrete facts from these pages. 34 of the 76 comparison answers stated costs in francs, 24 mentioned the federal contributions. Whatever is not written as text on the course page cannot appear in these answers.

    Three questions, three lessons

    • "How do I tell whether a continuing education provider in Switzerland is recognised and reputable?" All six answers named eduQua and federal recognition via SERI. A provider that is certified but does not state it as text on its website leaves out exactly the argument being asked for.
    • "Which provider in Zurich has the best pass rates for the marketing specialist exam?" All five answers named providers that publish their pass rate on their own website, with a percentage and a year. Several answers checked the figures critically: which year, how big was the class, who is counted, what is the Swiss average? A rate with these details gets quoted, a rate without context gets qualified.
    • "Which continuing education schools in Zurich have the best reviews?" Here the assistants mostly read review sources: ausbildung-weiterbildung.ch and qualitaetstest.ch, the site of the Swiss Institute for Quality Tests, in four of five answers each, ranq.ch in three.

    We counted mentions using a list of roughly 40 frequently named providers, so less frequently named schools are not fully captured in the table. The source count covers every cited page.

    Where the assistants get their recommendations

    For our AI visibility study we collected 1,515 answers from ChatGPT, Gemini and Claude to purchase-intent questions across twelve Zurich industries and traced 14,135 mentions back to their source. Continuing education was not among them, hence our own measurement above. But the three routes to a mention showed up in all twelve industries, only their shares varied:

    • 40 percent of mentions came from model memory. The model knew the company from training and opened no page. Well-known names benefit most.
    • 39 percent came from the named company's own website. The assistant read the page and took over its arguments.
    • 18 percent came from third-party sites: lists, portals and comparisons. Companies on a list page the assistant read were named 8.2 times as often as companies missing from it.
    • Pages with an FAQ were cited 1.6 times as often as pages without.
    • About three quarters of recommended companies were not in Google's top 10 for the same question. A ranking does not guarantee a mention.

    For a continuing education provider this means: you can hardly influence the memory route in the short term. The website route and the third-party route you can. How the criteria work in detail is in How do I get recommended by ChatGPT?.

    What a course page needs so that AI can recommend it

    An assistant recommends a programme with reasons: "part-time, two semesters, federal diploma of higher education, starts in March, in Zurich and online". It does not invent those reasons. It takes them from your page. If they are not there as text, it cannot cite them. Every course page therefore needs a fact block that can be read without opening a PDF:

    ItemWhat exactly belongs on the page
    QualificationThe exact title: federal diploma (eidgenössischer Fachausweis or Diplom), HF diploma, CAS, school certificate. Plus who awards it.
    RecognitioneduQua or other certifications of the institution, listing on the SERI notification list, recognition by associations. With a note on what the recognition covers.
    Duration and scopeSemesters, lessons or ECTS, teaching days per week, self-study.
    FormatOn-site, online, hybrid, part-time, full-time.
    Start and locationConcrete start dates and course locations, not "rolling".
    PrerequisitesPrior education, work experience, admission to the exam.
    CostsCourse fee, exam fee, materials, and how much of it federal subsidies can refund.
    OutcomeWhich roles and responsibilities the qualification opens up, and what comes next.
    FAQThe ten questions your advisory team hears most often, with short answers.

    The cost question deserves particular care. For courses that prepare for a federal professional exam, the Swiss Confederation refunds 50 percent of the eligible course fees, up to eligible fees of CHF 19,000 for federal professional examinations and CHF 21,000 for advanced federal professional examinations. The condition is that the course is on the notification list (Meldeliste) of the State Secretariat for Education, Research and Innovation (SERI). Providers enter their courses there by 15 November of the course year. If this applies to your programme, put it on the course page with figures, so that an assistant can cite it.

    Two rules apply to all of this. First: one page per programme. An overview page with twelve programmes gives the assistant one sentence about each. Second: the same data everywhere. If your website says the course starts in March, a portal still shows last year's date and your Google listing has an old address, the assistant sees contradictory facts. That weakens every recommendation.

    Which third-party sources matter in Switzerland

    In our measurement the assistants mostly read provider websites, plus ausbildung-weiterbildung.ch, berufsberatung.ch and eduwo.ch. Which pages matter in your category is shown by a measurement of your own. These sources are almost always relevant for continuing education in Switzerland:

    • eduQua. The Swiss quality label for continuing education providers, run by the Swiss Federation for Adult Learning (SVEB). More than 1,000 schools, institutes and academies are certified, and in many cantons the certificate is a prerequisite for public funding. If you are certified, state it on your website as text, not just as a logo.
    • SERI notification list. The list of preparatory courses for federal professional exams. It decides eligibility for federal subsidies and therefore belongs as a fact on every relevant course page.
    • berufsberatung.ch. The portal of the Swiss Service Centre for Vocational Education and Training and Career Guidance (SDBB), with information on occupations, schools, training and continuing education. The SDBB has fully relaunched the portal. Check that your offers are listed correctly.
    • ausbildung-weiterbildung.ch. An education portal that has listed programmes and courses by topic and location since 2004. By its own account, more than 100,000 prospective learners use it each month and around 700 providers advertise there. It hosts more than 12,000 manually checked reviews rating organisation, content, lecturers, teaching method and infrastructure, plus a recommendation rate.
    • eduwo.ch. A platform for discovering and comparing continuing education in Switzerland, with more than 300 partner schools by its own account and a course finder.
    • Associations and exam bodies. Check whether the organisation responsible for your exam lists preparatory course providers. A listing there is independent confirmation of your offer.
    • Guides and comparison articles. Trade media and career portals that compare programmes. Which of them get cited varies a lot by topic.

    How to get into such sources legitimately is described in Digital PR for AI visibility: first the source map, then targeted work on the pages that are actually read.

    Measuring: prompts, brands or keywords?

    Education providers ask this often, usually adding that management wants quarterly figures. The answer is clear: the prompt is the unit of measurement, the brand is the metric, the keyword remains an SEO tool.

    People do not ask ChatGPT "continuing education HR Zurich". They ask "I've worked in HR for three years and want to get the federal diploma, part-time in Zurich. Which schools are there?". A good prompt set reflects questions like these, structured by course category and decision stage:

    StageWhat the person wants to knowExample prompt
    OrientationWhich qualification moves me forward?"Which continuing education do I need to become a team leader?"
    ComparisonWhich providers are there?"Which providers for the federal HR diploma in Zurich would you recommend?"
    ReassuranceIs this provider good and affordable?"Is [your institute] recognised? What does the programme cost, and are there federal subsidies?"

    Five to ten prompts per course category are enough. A provider with six categories ends up with 30 to 60 questions. You run them regularly against ChatGPT, Gemini and Google AI Overviews, several times, because the answers fluctuate. Three things are counted:

    1. Share of mentions: your share of all provider mentions across the prompt set, compared with three to five defined competitors.
    2. Citations: how often your website appears as a source, and which pages.
    3. Competitors' sources: which portals, lists and articles the assistants read when they name others.

    The quarterly report to management fits on one page: share-of-mentions trend per category and system, biggest gains and losses, the most important sources, AI traffic and enquiries from GA4, and actions taken and planned. How to build the setup and the AI channel in GA4 is covered in detail in Measuring AI visibility.

    When competitors are named more often: finding the causes

    If competitors are named much more often, do not start producing content straight away. First assign every competitor mention to one of the three sources:

    • From memory: the assistant names the competitor without a source. Little helps in the short term. In the long term, mentions in media, associations and trade articles that feed into future training data help.
    • From the competitor's website: their course page answers the question better. Compare fact block, FAQ and level of detail with your own page.
    • From a third-party page: the competitor is on a list or portal the assistant reads, and you are missing. This is often the case you can fix fastest.

    Only this assignment shows whether you need to work on course pages, third-party sources or awareness. A partner for this analysis should deliver causes per prompt, not just a hit rate.

    The first 45 days for a new marketing lead

    Taking over marketing at a continuing education provider often means inheriting a patchwork: a few portals, some Google Ads campaigns, social media, a newsletter, trade fairs, and nobody knows exactly which of it brings enrolments. In its 2025 provider survey, SVEB describes strong competitive pressure that intensifies advertising for providers' own programmes. According to SVEB, a good third of continuing education organisations already use AI, mainly in marketing and communication. More one-off activities will not solve the problem. A clear sequence will.

    Days 1 to 15: inventory

    • Put every running activity, portal, agency and budget into one list, with costs and measurable results.
    • Check the course data: do qualification, duration, dates, costs and recognitions match across website, portals and Google Business Profile?
    • Check GA4: are enrolments and enquiries tracked as key events, is there an AI channel, does the form ask how people found you?
    • Build the prompt set and run a first baseline measurement. Define your competitors.

    Days 16 to 30: fix the foundations

    • Rework the course pages of the five to ten programmes with the largest revenue share: fact block, FAQ, outcome, costs.
    • Align the data on all portals. Correct or remove outdated listings.
    • Make recognitions visible: eduQua, notification list, association recognition, as text with an explanation.
    • Pause activities with no measurable contribution and shift the budget.

    Days 31 to 45: third-party sources and reporting

    • Derive the source map from the measurement: which lists, portals and articles do the assistants read in your categories?
    • Work on the three to five most important sources: listing, profile, alumni reviews, contact with editors.
    • Write guide content for the orientation questions, for example "Federal diploma or CAS: which suits whom?".
    • Run the second measurement and deliver the first report to management, with a roadmap for the next quarter.

    The rule behind it: whatever is close to enrolment and takes little effort comes first.

    Beyond the funnel: a framework for education decisions

    The funnel thinks in ad contacts. Continuing education, though, is a decision that takes months, often with an employer and funding involved. A framework along the decision fits better, close to the idea of "jobs to be done":

    1. Trigger: promotion, job change, new requirement, dissatisfaction.
    2. Orientation: which qualification gets me there?
    3. Comparison: which providers, which formats, which dates?
    4. Reassurance: recognition, costs, federal subsidies, reviews, experiences.
    5. Enrolment and completion: advice, placement, start.
    6. Loop: alumni review, recommend and come back for the next qualification; employers send more staff.

    Each stage has typical questions people ask ChatGPT and needs a page that answers them. The same framework structures your prompt set, so content, measurement and reporting follow one logic.

    Common mistakes

    • Facts only in a PDF. Brochures are hard for assistants to access. Whatever matters belongs on the course page as text.
    • "Costs on request". If you name no costs, you give the assistant nothing to cite when people ask about costs. A range with a note on federal subsidies is often enough.
    • Pass rate missing or without context. In our measurement, the question about pass rates mostly surfaced providers that publish theirs. State the year, the number of participants and the Swiss average alongside it.
    • Unclear recognition. "Certified" without saying by whom and for what convinces neither people nor machines.
    • Keywords and screenshots instead of a prompt set. Ranking reports and single answers do not show whether ChatGPT reliably recommends you.
    • Leaving reviews to chance. Ask alumni specifically for reviews on the portals that are read in your category.

    What we do for continuing education providers

    Hierarchy is a full-funnel marketing agency in Zurich, specialised in SEO, AI visibility and paid ads. For the continuing education sector we ran the measurement above in November 2026. For clients we work with the same method as in our AI visibility study: a prompt set per course category, a source analysis per answer and a baseline measurement against your competitors. This becomes a prioritised strategy that we implement ourselves: course pages, FAQs, structured data, portal profiles, third-party sources and the quarterly report to management. More under generative engine optimization. If your team wants to run the measurement itself, we train it in workshops. Where you stand today is shown by the free AI visibility check.

    Conclusion

    Providers get named when their own course pages carry clear facts and they appear on the portals and lists the assistants read. You measure this with a fixed prompt set by category and decision stage and report it as share of mentions in a quarterly trend. For a new marketing lead this is also the way out of the patchwork: inventory and baseline first, then course pages and portal data, then third-party sources.

    Sources

    Tenzin Langdun

    About the author

    Co-Founder of Hierarchy · SEO, GEO & AI Expert

    Tenzin Langdun is a co-founder of Hierarchy, where he leads SEO and GEO: how visible companies are in Google, in Google's AI answers and in assistants such as ChatGPT, Gemini and Claude. He also leads the agency's own data studies, including the Zurich AI Visibility Study of 1,515 AI answers about 5,110 companies. He brings over 10 years of marketing experience across strategy, paid acquisition and SEO, and has worked at KPMG, EY, Siemens, Adnovum and LGT, in AI, cybersecurity, audit and consulting, and in software engineering for the insurance sector and for banking and finance. He holds a BSc from ZHAW and an MSc in Artificial Intelligence from the University of Bath. His research on AI-based thyroid cancer detection was published in Scientific Reports (Nature Portfolio) in 2024 and won the Siemens Excellence Award 2024 and the ZHAW Lab Sciences Award 2024.