Quick Answer
Following a child's lead in speech therapy does not mean abandoning your goals. It means changing your delivery method. Instead of creating artificial contexts to practice targets, you embed goals into activities the child is already engaged in. This requires strong clinical observation, flexible planning, and a shift from "director" to "partner." Children typically make faster progress because motivation and engagement drive learning.
In This Post
The Fear That Keeps SLPs From Going Child-Led
I hear it all the time. In my DMs, in workshops, in the comments on my posts. "I want to follow the child's lead, but I'm afraid we'll just play and nothing will happen."
I understand that fear because I lived it. Early in my career, I had sessions planned down to the minute. I knew exactly what we were doing, in what order, and how many trials I needed. And if a child went "off script," I redirected them back.
Because if I wasn't eliciting targets in a structured way, how could I be sure anything was happening?
Let's be honest about where that fear comes from. We were trained in programs that taught us to value control, measurement, and standardized procedures. Our supervisors counted trials. Our professors talked about "opportunities to respond." The entire framework assumed that learning happens when we create it and the child performs on cue.
So when someone says "just follow the child's lead," it can feel like being told to do the complete opposite of what we were taught. But it doesn't have to feel chaotic!
Why "Following the Lead" Doesn't Mean "No Plan"
I want to be really clear about this, because I think it's the biggest misconception about child-led therapy. You still plan. You absolutely still plan. The difference is what you plan for.
In a therapist-directed model, you plan activities, the specific toys/games/worksheets, the order, the number of repetitions, the reinforcement schedule. The plan is a script, and the session is a performance.
In a child-led model, you plan your targets (e.g. we're going to focus on requesting and protesting), you prepare your environment, and you plan your strategies. But you don't plan the specific activities, because the child is going to tell you what those are.
Think of it like this: In therapist-directed therapy, you build the road and the child walks it. In child-led therapy, the child picks the direction and you lay the road as you go. You still know where you need to end up. You're just letting the child choose the route.
Flexible planning is harder than rigid planning. It requires deeper knowledge of your goals, more creative clinical thinking, and stronger observational skills. If anyone tells you child-led therapy is the "easy" way, they haven't actually done it well.
The Partner Mindset: Your New Role in the Room
The biggest internal shift you'll make in this process is moving from director to partner.
As a director, your job is to orchestrate. You decide the activity, the pace, the sequence. You know where you're going and you steer the child there.
As a partner, your job is to join. You enter the child's world. You become a participant in whatever they've chosen, and from inside that activity, you create the conditions for language and communication to emerge.
Child-led therapy is not passive. In fact, it demands more clinical skill because you have to be responsive and intentional at the same time. You're tracking what the child is doing, anticipating where language might emerge, setting up the environment to support their goals, and modeling in real time.
A Step-by-Step Framework for Child-Led Sessions
1 Observe Before You Act
Before you do anything else, watch. Give the child 2–3 minutes of unstructured time and observe: What do they gravitate toward? How do they engage with materials? What is their communication like when no one is making demands? This observation is clinical data and its hugely important.
2 Join Their Activity on Their Terms
Enter the activity as a participant, not a facilitator. If they're building with blocks, start building too. If they're moving cars around a track, get a car. Don't redirect. Don't add demands. Simply join in and observe how they respond. This is how you build the trust and joint attention that makes everything else possible.
3 Map Your Goals to What's Happening
Now that you're in the activity, start identifying where natural opportunities exist. If your goal is requesting, when would be a good time to model a request in this activity? If your goal is describing, what is there to describe? If your goal is turn-taking, how can you create natural turns? You're not forcing the goal into the activity, but instead you're adding value in where it makes the most sense.
4 Model, Expand, and Wait
These are your three primary tools. Model relevant language within the context of the activity. Expand on what the child says or does. For example, if you're playing with blocks, maybe you knock *your* tower down. Wait, and genuinely wait, for the child to respond, initiate, or communicate. The wait is where the magic happens, but most of us don't wait long enough to see it!
5 Stay Flexible as the Activity Shifts
Ideas shift! Blocks become a crashing game. The cars get abandoned for something across the room. This is not a problem! Follow the child. Each transition is an opportunity to observe, join, and provide value in a new context. Multiple contexts = stronger generalization.
Embedding Goals in Real Time: What It Actually Looks Like
Let's make this concrete. Here's a real session example:
Goal: Increase spontaneous requesting using core vocabulary (more, want, help, stop)
Child-selected activity: Filling and dumping a container of small toys
- You join the activity and start filling the container too
- You "accidentally" run out of toys and pause, waiting for a request
- You model "more toys" and add to the container
- You put the lid on too tight, waiting for "help"
- You dump the container without warning, waiting for "stop" or a protest
- You narrate everything: "filling… full… dump… all gone…"
You got multiple opportunities to target your goal. The child never left their preferred activity. There were no flashcards or worksheets. Instead you acted as a thoughtful partner creating natural communicative opportunities.
Tracking Progress Without Killing the Flow
Data collection is where many SLPs feel like child-led therapy falls apart. And I get it... if you're used to counting trials on a clipboard, this feels incompatible.
But data collection doesn't have to mean trial counting. Here's what works in child-led sessions:
- Frequency tallies in real time: A small tally sheet on your knee or a counter in your pocket. Count spontaneous productions, prompted productions, and communication attempts.
- Post-session narrative notes: 2–3 minutes after the session, write what you observed. This is often richer clinical data than trial counts.
- Video samples: Short clips (with consent) that capture authentic communication in context.
- Communication temptations: Set up specific conditions (like the lid scenario above) and note whether the child responded, the level of support needed, and the quality of the response.
Remember: Data is evidence of learning. Child-led therapy generates real evidence, it just looks different than a traditional + or - data sheet. Your job is to collect it in a way that captures what's actually happening.