How to Run Speaking Practice with a Large Class

The challenge every language teacher knows

You have 30 students, 50 minutes, and a speaking objective. The textbook says "discuss in pairs" but doesn't mention the logistics: who pairs with whom, what happens when they finish in 90 seconds, how you monitor 15 simultaneous conversations, or what to do with the three students who refuse to talk.

Teachers face the same problem in every large class: speaking practice scales badly when only one student talks at a time. YapYapGo is a classroom speaking practice tool for ESL teachers that runs the whole room in simultaneous pairs from a single projected screen.

Running speaking practice with a large class isn't just harder - it requires a fundamentally different approach than small-group teaching.

Start with the pairing problem

The single biggest time sink in large-class speaking practice is organising who works with whom. Teachers typically default to one of three methods: "turn to your neighbour," teacher-assigned pairs, or letting students choose. Each has serious drawbacks.

Neighbour pairs mean students always work with the same person. Within a few weeks, they've exhausted their conversation strategies and fall into repetitive patterns. Teacher-assigned pairs eat five minutes of class time every lesson. That's over three hours of lost teaching time per semester - just on logistics. Student choice creates cliques, leaves shy students isolated, and often results in L1-dominant pairs who default to their shared first language.

The solution is randomised rotation. When pairs are shuffled automatically before each activity, students encounter different partners in every lesson. This builds classroom cohesion, prevents comfort-zone stagnation, and ensures every student gets exposure to different speaking styles and ability levels.

Structure the time, not the conversation

A common mistake with large classes is over-scripting the speaking task. When you give students a rigid dialogue to follow, the activity becomes a reading exercise, not a speaking exercise.

Instead, structure the time and let the conversation flow naturally:

• Give students a clear question or topic
• Set a visible timer so everyone knows how long they have
• Let pairs manage their own conversation
• Rotate to a new question when the timer ends

This approach works because it removes the teacher as a bottleneck. You're not directing 15 pairs individually - the timer and question prompt do that for you, freeing you to circulate, listen, and give targeted feedback.

Use vocabulary as scaffolding, not a crutch

Showing suggested vocabulary during speaking tasks is a balancing act. Too much support and students read rather than speak. Too little and lower-level students freeze up.

The key is making vocabulary available but not visible by default. Students who need it can access it; students who don't aren't distracted by it. This respects the natural variation in ability levels within any large class.

Monitor without hovering

With 15+ pairs talking simultaneously, traditional monitoring - standing next to one pair and listening - means you hear less than 7% of the total speaking time. Instead, adopt a scanning pattern: move through the room in a predictable circuit, spending no more than 30 seconds with any pair. Listen for patterns and common errors rather than trying to catch every mistake.

Save detailed feedback for after the activity. During the activity, your job is to keep energy high and participation universal.

The bottom line

Large-class speaking practice works when you solve the logistics first. Automate the pairing, structure the time, provide optional scaffolding, and circulate rather than hover. The conversation takes care of itself.