The email arrives with a PDF attached and a subject line that says "Annual review — goals". You open it on your phone in the school parking lot. Six sentences, each built like a clause in a lease. Your child will produce two-word combinations in 80% of opportunities across three consecutive sessions given a visual prompt.
You read it twice. It sounds serious. It sounds professional. And you have no idea whether it is any good.
Almost every article about speech therapy goals for autism hands you a list of sample goals. That was never the problem. You can find a hundred in ten seconds. The problem is that you cannot tell whether the six on your child's sheet are the right six, or whether one of them is quietly asking your child to be less visibly autistic rather than more able to communicate.
So this is not a list of goals. It is how to judge one.
What makes a good speech therapy goal for an autistic child?
A good goal names a specific communication behavior, in a specific situation, at a stated level of independence and frequency you could verify yourself if you sat in the room with a tally sheet. Speech therapy goals for autism should also be functional. Meeting the goal should change something in your child's actual Tuesday. And they should count communication in whatever form your child uses, not speech alone.
Decoding the jargon
The vocabulary is not there to confuse you. It is genuine clinical shorthand, and the precision only helps you if you know what each phrase commits the therapist to.
| Phrase on the sheet | What it actually means | What to ask |
|---|---|---|
| independently | No adult prompt of any kind, including a raised eyebrow, a held-out hand, or an expectant pause | "What counts as a prompt here? Does waiting expectantly count?" |
| with 80% accuracy | Eight out of ten attempts | "Eighty percent of what? Who counts, and out of how many?" |
| in 4 out of 5 opportunities | The same idea with the denominator stated, which is better writing | Nothing. This one is doing its job. |
| across 3 consecutive sessions | A stability rule, so one lucky Thursday cannot count as mastery | "Consecutive across how long — three weeks or three months?" |
| given a visual prompt | A support that is built into the goal permanently | "What is the plan for fading the visual, and by when?" |
| given a verbal model | The adult says the word first and the child copies it | "Is there a version of this goal without the model? Imitation is not the same as use." |
| in a structured setting | The therapy room, at a table, with the therapist | "Where else does this need to happen for it to count?" |
| spontaneously | Your child starts it. No lead-in, no question, no expectant look | Nothing. This is the most valuable word on the whole sheet. |
| as measured by clinician observation | The weakest data source available | "Can I see the actual data sheet rather than the summary?" |
| will tolerate | A compliance word that has wandered into the wrong document | "What is my child learning to do here, rather than learning to put up with?" |
One more thing to hunt for, often missing: the baseline. A goal without a starting number cannot show progress, because there is nothing to have moved from. If the sheet says your child will request in 8 of 10 opportunities, and nobody wrote down that they currently manage 1 of 10 with a full physical prompt, then in nine months everyone will be arguing from memory. Ask for baselines in writing before you sign.
A weak goal and a strong goal, side by side
Here is a goal of the kind that appears on real paperwork constantly:
Weak: Child will improve expressive language skills with 80% accuracy as measured by therapist observation.
It reads fine at a glance and is close to meaningless. "Expressive language skills" is a category, not a behavior. Eighty percent of an unnamed denominator is a number pretending to be data. There is no setting, no prompt level, no baseline. A goal like this cannot be failed, which means it also cannot be met — it can only be reported as "progressing" indefinitely.
Now the same underlying intent, written properly:
Strong: During a preferred play routine, Child will request a desired object or action using a spoken word, a sign, or their AAC device in at least 8 of 10 opportunities, with no adult prompt beyond a five-second expectant pause, across three consecutive sessions and in at least two settings including home. Baseline: 2 of 10 with a full physical prompt.
What makes the second one strong:
If your child's goals do not survive that comparison, it is usually a reason to ask for a rewrite rather than to distrust the therapist. Most will do it willingly, because tight goals make their job easier too.
The goals that are compliance wearing a communication costume
Some goals sitting in the speech and language section are not about communication at all. They are about your child being easier to manage.
Watch for goals shaped like these:
Here is the test, and it runs on every line of the sheet in about ninety seconds. If my child met this goal perfectly, what could they do next year that they cannot do now? If the honest answer is "be easier to teach", it is a compliance goal. If the answer is "tell someone their ears hurt", it is a communication goal.
The goal most often missing is refusal. A reliable way to say no is a safety skill and a dignity skill. A word, a sign, a symbol, a push of the device: any of them count, and it belongs on the sheet as explicitly as requesting does. ASHA's autism practice portal treats the full range of communicative functions as the target, not speech production alone.
How do I know if speech therapy is working?
You know it is working if someone can show you a number today and the number it started from, and the second is bigger. If nobody can produce both numbers, nobody knows whether it is working, including the therapist.
One reporting period of flat data is noise. Three is a wasted year. Two is the moment to ask hard questions, and these are the ones that get useful answers. Take them to the meeting on paper.
1. Can I see the raw data by month rather than the summary? Summaries smooth. Raw tallies show whether the line is flat, wobbly, or slowly climbing.
2. What was the baseline when this goal was written, and what is the number today? If nobody can answer this, you have found the problem.
3. Is the goal too big? Ask what the next smallest step would be. Often a goal has not moved because it skipped two rungs.
4. What do you think the barrier is: motivation, comprehension, motor planning, or regulation? The highest-value question here, because those four have four completely different fixes, and a therapist who has not decided between them is guessing.
5. Have we tried a different modality? If a spoken-word goal has not moved in two reporting periods, the question of AAC is overdue, not premature.
6. What would you need to see to conclude this is the wrong goal, and what would replace it? Ask it exactly like that. It gives the therapist permission to say what they may already think.
If you only get two, ask two and four. The baseline question exposes whether anyone is measuring; the barrier question exposes whether anyone has a theory.
A therapist who welcomes this list is a good sign. One who is visibly offended by it is also telling you something. If a rewrite is refused twice, ask for the refusal and the reason in writing, and ask what evidence would change their mind. In a school setting that request has teeth; in private practice it is your cue to get a second opinion.
Tracking speech therapy goals for autism at home
You do not need training to gather evidence. Pick one goal and keep a tally where you will see it, like the side of the fridge: date, what your child wanted, how they asked, whether an adult prompted. Two weeks of that is a real counter-argument to "we are seeing lovely progress in sessions", and just as useful when it agrees with the therapist. Anything that logs practice attempts does the same job. Verbalyft shares its records with the therapist for that reason, but a fridge tally works just as well. The point is arriving with something dated.
This is also what makes a two-setting goal writable. A therapist cannot promise data from your kitchen. You can.
FAQ
How many goals should my child have at once?
No fixed correct number, but a sheet with eight or nine communication goals usually means none is getting real session time. Ask which two or three are the priority, and what that means in minutes.
My child's goals only mention speech. Can they include AAC?
They can and, if your child uses or is trying out AAC, they should. A goal written as speech-only quietly stops counting your child's most reliable way of communicating, which distorts every number that follows.
Should eye contact ever be a goal?
Attention and orientation can reasonably be targets when a child genuinely misses information because of them. "Eye contact for four seconds on command" is something else. Ask what the eye contact is in service of, and whether the same purpose could be met by a less costly behavior.