Put your wallet away and take a breath.
AI optimisation is the new marketing buzz – the idea being that because people increasingly rely on AI searches and AI results at the top of Google searches to inform their decision making, you need to deposit a six figure sum in the bank account of a Svengali of AI wrangling to remain visible and in the consideration set of future students.
However, a quick mystery shopping effort reveals that the communication problem is often not the fault of AI – or anything to do with AI. The problem is often that your webpages are not clear and consistent on why students should bother looking in your direction in the first place. Of course, this is part of the AI optimisation process – but there is not much point doing the rest of the work until your basic messaging is right.
This is important in a way that it wasn’t last year, because ChatGPT and Google’s AI search both appear to be taking up a counselling/ decision-making role when asked which university to choose, in a way that they weren’t last year.
Last year I conducted a range of searches asking various gen AI platforms to help me choose between different universities, different TAFEs and different courses – and the results were pretty beige – with few responses providing a nudge any direction and instead just summarising web searches.
In contrast, 2026 heralds a new dawn of outsourced decision making – with opportunities to outsource choosing an institution or course to the unseen megaminds spitting out answers through gen AI.
Here’s some examples:
I asked Google AI “Is Deakin’s Bachelor of Psychological Science better than La Trobe’s,” expecting that the program would hedge its bets, not choosing one or the other, as it did last year, but instead it spat out, “Both Deakin University and La Trobe University offer highly regarded, APAC-accredited programs. In 2026, the "better" choice depends on whether you prioritize fast-tracked career entry (Deakin) or research-led community engagement (La Trobe).”
I thought that was pretty interesting, and also that La Trobe would be unlikely to be happy to be characterised as inferior on fast-tracked career entry and focusing on research-led community engagement – a kiss of death if you want to usher an extra few hundred 18 year-olds in the door. It also misses the AI-first mindset spruiked in Theo Farrell’s latest strategy (but obviously not in the university’s web pages).
I also asked Google’s AI search for a comparison on which Bachelor of Laws degree was better at ACU, Deakin or La Trobe, and it responded with a range of useful considerations, but prioritised ATARs and university rankings in relation to desirability (QILT data and other metrics of course employment rates and satisfaction were not referenced). When I then asked ChatGPT which would get me a “good job,” the answer was Deakin, based on broader electives, clinical and internship pathways, employability metrics, networking and industry.
When I asked whether I should enrol in TAFE or university to get my first job in the building and construction industry, Google’ AI summary told me TAFE was the better option, “because it provides the hands-on, practical skills and certifications (like the mandatory White Card) that employers in trades and technical roles highly value. University is more suited for academic, research-based careers or professional roles like engineering or architecture.”
When I asked if I could save money by doing a TAFE qualification before enrolling in a nursing degree, Google AI said “Yes, you can potentially save a significant amount of money by completing a TAFE Diploma of Nursing before pursuing a university nursing degree.”
This could be very useful, but needs to be paired with pathway information – which ChatGPT provided, giving me Victoria-specific information without asking for it.
Takeaways
- Gen AI tools will be adopted by school career practitioners and established as online career planning tools in coming years, whether the tertiary sector likes it or not.
- Any course convenor wanting to grow enrolments over the next 12 months should do their own mystery shopping. It takes no special skill, and only a few minutes, but does require a mindset change. Pose as a prospective student and see how your course and institution is presented vis-à-vis competitor programs. Once you have done that, don’t shoot the messenger / AI Bot. This is not an AI issue. AI simply exposes the issues of not just poor communication, but poor strategy and positioning, failing to clarify key points of differentiation that matter to prospective students.Many universities and TAFEs have struggled with genuine and effective differentiation for decades. Many courses are extremely similar, and will only be presented as different based on differences scraped from a range of webpages.
- Third-party forums, such as Reddit and Whirlpool are influential. Students and alumni give their views in various forums, opining on which course is best have a significant influence on results with some AI platforms, even when they are quite old – we saw some comparisons quoted from 2013. This, of course, is easy to game by putting up a range of fake positive reviews from spam accounts – and so gaming AI is likely to become just as significant as gaming rankings systems in the near future.
AI tools on their own clearly do not provide the best advice that students can get, and there should be many concerns about a proliferation of tools which pretend to advise students on choosing a path based on spurious criteria and inadequate descriptions of course offerings. However, this analysis does provide a window into what a quick web scrape can reveal. Those ancient course pages at the bottom of your web update priority list suddenly aren’t so trivial anymore.
For Universities and TAFEs, the biggest change of AI is not exposing different levels of sophistication in technical management, but exposing the lack of clarity and audience-relevant differentiation in the way institutions present themselves.