Every week, another school buys an AI tutoring platform. Half of them regret it within six months. The other half transform their academic outcomes. The difference isn't luck — it's knowing what to look for before you sign a contract.
Start with the problem, not the tool
Before evaluating platforms, be specific about what you're trying to solve. "Students need more homework support" is too vague. "Our Grade 9 algebra students are falling behind, and we don't have enough math TAs to cover afternoon office hours" is a problem a tool can actually address.
Schools that buy a platform without a clear problem usually end up with a tool no one uses.
Criteria that actually matter
Pedagogy first, AI second. Does the platform teach the way your teachers teach? Does it align with your curriculum? Does it encourage active learning or just answer-giving? A beautiful AI that undermines your instructional model is worse than no AI.
Curriculum alignment. If you teach IB, AP, national standards, or something specific — make sure the platform is aligned out of the box. Shoehorning a generic tool into your curriculum is a constant headache.
Teacher dashboard quality. Can teachers see what students are working on, where they're struggling, and what topics to reteach? A platform without a strong teacher view is a platform that operates outside your instructional loop.
Safety and content filtering. Non-negotiable for K-12. The platform must restrict conversations to appropriate topics, filter content aggressively, and give admins visibility into any concerning interactions.
Privacy and compliance. FERPA, COPPA, GDPR as applicable. Ask for written commitments — not marketing language. Where is data stored? Is it used for training? Can students' work be deleted?
Integration with your existing stack. SSO with Google/Microsoft/Clever. Grade passback to your LMS. Roster sync. Without integration, adoption dies.
Parent visibility. Can parents see progress without logging in as their child? Clear reporting is a huge adoption booster.
Language support. Do you have ELL students? International students? Does the platform support their first languages for complex concepts?
Questions to ask every vendor
- Can I see an actual teacher dashboard demo, not just the student view?
- How do you prevent the platform from just giving students the answers?
- What happens when students try to use it for off-topic conversations?
- What reference schools can I call — specifically schools like mine?
- What's your typical onboarding timeline and what resources do you provide?
- What does renewal look like in year two?
Pilot before you commit
Run a structured 6-8 week pilot with one grade level or department. Define what success looks like before you start. Measure before and after. Survey teachers. Survey students. Then decide.
Any vendor unwilling to support a real pilot is a vendor you should walk away from.
Red flags
- Vague answers about how the AI handles sensitive topics
- No clear data privacy commitments
- Single customer case studies repeated across pitches
- Pricing that balloons at renewal
- No dedicated teacher training
The bottom line
An AI tutoring platform can materially improve outcomes in your school — but only if you pick one that fits your pedagogy, your curriculum, and your operational reality. iTutor supports schools through this exact process: curriculum-aligned content, teacher dashboards, strict content filtering, and real pilot support. The platforms that last are the ones that integrate into how you already teach.