Seven questions to ask before you choose a marketing simulation in 2026

Most marketing professors choose a simulation by default. The textbook publisher bundles one. A colleague recommends what they used last year. The department keeps whatever the previous instructor set up. None of that is a criticism. Evaluating simulation platforms is not what anyone got into academia to do, and most of us have better calls on our time than vendor demos in August.

But the cost of defaulting in 2026 is higher than it was in 2022, because what students need to learn has shifted and the simulations have not always kept pace. What follows is a set of questions, not a recommendation, based on Novela's years of experience building marketing simulations for leading institutions worldwide.

1. Is the curriculum current for 2026, or is it 2018 with a fresh skin?

It is easy, and not unreasonable, for a simulation vendor to keep an old product running and refresh its surfaces. The student interface gets a new colour palette. The dashboard gets a chart that was not there before. The underlying model of how search works is still the one that was built when a results page was ten organic blue links and AI did not feature. Rebuilding a simulation from scratch is a serious undertaking and most vendors do not have the appetite for it. But the world your students are graduating into is moving more quickly than the platforms they are learning on, and that gap is the one worth measuring.

Ask what is actually in the sim, not what is on the marketing page. A simulation built for 2026 should let students work with:

• AI search results, including AI Overviews and the surfaces that have replaced the classic results page

• Organic visibility across answer engines (ChatGPT, Claude, Perplexity, Google AI Mode) rather than Google rankings alone

• Attribution models that do not pretend last-click still describes reality

• AI tools embedded in the student's working day, not bolted on as a separate prompt-engineering exercise
  

A good answer to this question looks like a screenshot of the current simulation. A weaker answer looks like a vague roadmap.

→ Worth a look: our AI Search simulation.

The NEW Novela AI Search (SEO/GEO) Simulation

2. Is AI a topic in the simulation, or is AI the workflow?

Most simulations now mention AI somewhere. The harder question is whether AI is a module the student visits and exits, or whether AI is the way the student does the work of the sim. The difference matters because the labour market your students are graduating into expects them to use AI tools in their daily work, not to have studied them as a separate subject.

Things to look for:

• AI tools available inside the student's working environment, not on a separate page

• Tasks that require the student to use AI to complete them, not tasks that mention AI in passing

• AI behaviour that reflects real model limitations (hallucination, citation, retrieval), not idealised AI

• Assessment that asks the student to explain how they used AI, not just whether they used it
  

A platform where you could remove the AI features without affecting the core curriculum is a platform where AI is a topic. A platform where removing them would break the sim is a platform where AI is the workflow.

→ More on this in our piece on the AI marketing skills students need to learn in 2026.

3. How long does setup actually take before week one?

Some setup is fair. Building a syllabus around a new tool is not a five-minute job, and a vendor who promises otherwise is usually selling something more cosmetic than functional. The real question is the gap between the time a vendor describes on the demo call and the time it actually takes once you have your class list and your dates and your assessment weighting in hand. That gap is the one that quietly kills adoption.

Things to look for:

• A live student environment running in under an hour from sign-up

• Pre-built assignments and assessments mapped to standard syllabi

• A clear path from sign-up to syllabus inside one working day

• A named contact at the vendor who replies in hours, not days
  

If setup eats more time than you have between term planning and the first class, the platform will not survive the semester. That is true even if the platform is otherwise excellent.

→ Our version of that story is in Asked and answered: from sign-up to syllabus in two days.

4. Is the simulation built for marketing students, or for general business students?

There is a difference between a simulation that covers marketing as one of several modules and a simulation built around the actual decisions a marketer makes. The first is broader and works in a generalist programme. The second is narrower and works in a marketing-specific class. Both are reasonable products. The mismatch is when a professor teaching a specialised marketing course adopts a generalist platform and finds it spends two thirds of its time on operations, finance, and HR.

Things to look for:

• Depth in the channels and decisions you actually teach (campaign mechanics, audience theory, channel choice, measurement)

• Cases drawn from marketing rather than from a generic corporate setting

• Vocabulary and reasoning structures that match how marketers talk in practice

• Supplementary content written by marketers, not by general management consultants
  

If the platform feels like it was scaled down from a general MBA simulation, it probably was. If it feels like it was built for the specific course you are teaching, that is usually because it was.

→ How we approach this is in Teaching the principles of marketing through simulation.

5. Can students see why their decisions worked or didn't?

A simulation that ends each round with a leaderboard works as a game. A simulation that explains why a decision changed an outcome can work as a teaching tool. Both have a place in a classroom, and a well-designed game can build engagement that a more analytical platform will not. The question is which one you are looking at, and whether your learning outcomes match what the platform is good at.

Things to look for:

• A diagnostic report after each round, written for the student rather than the instructor

• Causal explanations linking the student's decision to the outcome: what changed, why it changed, what they would do differently

• The ability for the student to replay a decision and see an alternate outcome

• A space for the student to write their own analysis before the system tells them theirs
  

If the only thing the student takes away from a round is their rank, the platform is doing a different job from teaching. That is fine if it is the job you came for. Worth being clear with yourself either way.

→ More on the pedagogy of explained outcomes in From data to decisions.

6. What does grading actually look like after each round?

The least discussed question in any vendor demo, and the one professors live with every week. A simulation that produces clean student data takes work out of the grading process. A simulation that does not generates an additional weekly task on top of the one you already have. Most professors do not ask about this during evaluation because grading is not yet the salient problem. It becomes salient by week four.

Things to look for:

• Per-student decision logs you can read in five minutes per student, not fifty

• Diagnostic reports that map student decisions to your assessment criteria

• Rubric-friendly outputs that align with how you actually mark, not with how the vendor wishes you would

• Bulk-export options for the registrar and for your own records
  

A simulation that gives you cleaner student data than you would otherwise have is one that pays back its setup cost by week six. A simulation that gives you more data of lower quality is one that quietly increases your hours.

7. Who at your institution type is using it, and what do they say?

The most reliable trust signal is a name and an affiliation. A vendor that can introduce you to three professors at institutions like yours, in the last academic year, has earned a place on a shortlist. Anonymous testimonial blocks are less useful than they look, because they do not let you ring a colleague and ask the question you actually want to ask. That is the question whose answer is usually the one that decides the choice.

Things to look for:

• Named professors at named institutions, not "Dr S, R1 university"

• Adopters at your institutional type: R1 research, liberal arts, teaching-focused, undergraduate, MBA, US, European

• Recent adopters, in the last academic year rather than the last decade

• A direct introduction to a colleague you can speak to
  

If a vendor cannot name three professors at your level of institution, it is worth asking why. The answer is sometimes innocent. A new platform with strong early adopters at a few schools is a different thing from one that has been on the market for a decade with no public adopters. The question separates the two.

→ Our Office Hours series is where the professors who use Novela talk about what worked and what did not. A recent one is with John Ploumitsakos at the University of Colorado Boulder.

How to actually run the evaluation

The questions above are not a checklist to send to a vendor. They are a checklist to take into a free trial. Most platforms, including ours, will give you access for the asking. Spend forty-five minutes inside the simulation building a sample assignment as if you were one of your own students. Forty-five minutes is enough to know what is actually running inside the sim, beyond the marketing surface. You will learn more in that one sitting than in any number of vendor calls.

If you would like one of the trials to be ours, the Novela simulations catalogue and the AI Search simulation are the places to start. If you would prefer a teaching guide to read first, that is here. And if you would like to hear from professors who have already used the platform, the Office Hours series is where they talk about it.