Communication9 min read

AAC for Autistic Children: A Parent's Starter Guide

By Verbalyft Editorial TeamEditorial standards

The SLP said "I think we should start looking at AAC," and you nodded like a person who knew what that meant.

Then you sat in the car park and searched AAC for autistic children and got a wall of acronyms. SGD, PECS, core vocabulary, grid size, eye gaze, all of it written by professionals for other professionals. Somewhere in there was a device costing more than your fridge.

You did not ask the real question in the room, because it sounded like giving up: if we do this, does it mean they are never going to talk?

Will AAC stop my child from talking?

No. Research on AAC use in children with limited speech consistently points the other way. Children given a communication system tend to show equal or better speech outcomes than those who are not, and the "shortcut" worry has not held up. ASHA's autism practice portal reflects this: AAC is described as supporting communication development, not competing with speech.

The mechanism is not mysterious. A child who can communicate does it more, and doing it more produces more of everything: more turns, more attempts, more attention to language, often more words. Less practice is not a route to speech.

There is a second reason to stop waiting, and it has nothing to do with speech. Your child needs to be able to say "it hurts", "I want that one", "stop", and "I'm scared" this year, not eventually.

What AAC for autistic children actually means

AAC stands for augmentative and alternative communication. Not a brand, not a device category. Any way of communicating that is not speech, ranked roughly by how much technology is involved.

  • No tech. Gestures, signs, pointing, facial expression. Your child is already doing some of this and it counts.
  • Low tech. Photo cards, a choice board, a laminated core word board. Costs almost nothing, never runs out of battery, works at the beach.
  • Mid tech. Buttons that play a recorded message. Good for one specific job, like calling someone. Not a language system.
  • High tech. A speech-generating device, or an AAC app on a tablet. A full vocabulary the child navigates to build whatever they want to say.
  • Most children who use AAC well end up using several at once. That is multimodal communication, and it is the normal outcome, not an interim compromise. The goal is not one system. The goal is that your child is never without a way to be understood.

    The AAC decision tree: where to start

    Do these regardless of what you choose. Make a low tech board tonight: photograph six things your child wants often and stick them where the asking happens. Get motor, vision and hearing looked at before you commit to any technology, because each one changes access method, symbol size and contrast. And whatever you end up with, have the SLP configure it. None of that delays a real system.

    Now the three questions that actually change your answer.

    1. How much language does your child take in? Not what they say. If your child follows instructions without gestures and reacts to conversation not aimed at them, go straight to a full vocabulary system and hide nothing. If understanding is limited or genuinely unclear, use that same full layout but ask the SLP to mask it down to a handful of visible cells, so buttons never move as more are revealed. Either way, do not start on a six-cell app you will replace in six months.

    2. Low tech or high tech first? If an evaluation is months away or funding is unresolved, low tech is your main system now and high tech is a later upgrade. If the evaluation is booked and funding has a route, go high tech from day one and keep the low tech board as permanent backup. The expensive mistake is not choosing wrong. It is waiting with nothing while you decide.

    3. If high tech, dedicated device or app on a tablet? Choose the device if funding is available, if the family tablet is already contested in your house, or if your child gets stuck in other apps. Choose the app if you are self-funding, if your child is already fluent on your tablet, or if you need something inside a week. Devices are durable and locked to communication. Apps are cheaper up front and easier to lose to Netflix.

    If more than one row below describes your child, take the highest one. They are ordered by which factor decides first.

    Start here ifWhere to startWhy
    Motor, vision or hearing access is unresolvedNothing yet. Book an access assessmentAccess method decides everything below it
    Access fine, but devices get thrown or mouthedLow tech and mid tech first, rugged case before high techReplace-cost fear makes parents ration access
    Access and handling fine, no evaluation booked yetLow tech, six symbols, tonight; high tech at the evaluationFree, immediate, and it informs the appointment
    Evaluation booked or done, child taps accurately and likes screensHigh tech as the main system, low tech board as backupMotivation outweighs most other factors
    Evaluation booked or done, but screens hold no appeal or nobody at home is confident with techLow tech core board, plus one deliberately simple appA system nobody can operate is not a system

    Unreliable speech does not move you down this table. It is a strong indicator for a second channel, not a reason to wait.

    Nobody should have to earn a device

    You may be told your child must first demonstrate certain skills: matching pictures, following instructions, sitting still, proving intent. That is the prerequisite model, and the field has largely abandoned it. Nobody made your speaking child prove readiness before you talked to them. Making access conditional on skills the child would develop by having access is a loop with no exit.

    One version is a skills checklist gating the evaluation. The softer and far more common version: the device is granted, then kept in a cupboard, brought out for therapy, taken away at mealtimes and in the car and at Grandma's. A communication system available for twenty minutes a day is a worksheet. It goes everywhere the child goes.

    Why the SLP configures it, and what to ask

    Whether a system works or gets abandoned is decided almost entirely by how it was set up, and the right answers are not intuitive. What a good setup gets right is core vocabulary, the small set of words that make up most of what anyone says. Want, more, go, stop, help, that, my, no, different, again. Unglamorous, and they are what lets a system say things you did not anticipate. A grid full of food photos can request lunch. It cannot say "not that one", "I don't like this", or "go away".

    Questions worth asking your SLP directly:

  • What is on page one, and why those words? If page one is all nouns, ask why.
  • How does this grow? You want the same layout expanding, not a rebuild in a year. Button positions must stay put, because selection becomes motor memory.
  • Who trains the people around my child? School staff, grandparents, you. The most skipped step, and the most common cause of failure.
  • How will you model on it? The adult should be using the device to talk too, pointing to symbols while speaking. Children learn AAC by watching fluent users. Ask for a demonstration to copy.
  • What if my child uses it to say no, to swear, or to refuse therapy? Those should get honoured. If the answer involves removing access, that is a serious flag.
  • Funding, briefly and honestly

    Routes vary by country and by plan, so treat this as a map, not a guarantee.

  • Insurance (US). Devices are often covered as durable medical equipment, but need SLP documentation: an evaluation, device trials, a statement of medical necessity. The paperwork is the barrier, not the policy.
  • School district (US). If a device is needed to access education it can be written into an IEP, with access specified across all settings. See [IEP communication goals for autism](/blog/iep-communication-goals-autism).
  • Health service or charity funding. Specialist assessment services and grant programmes exist outside the US. Your SLP will know the local route.
  • Vendor loan and trial schemes. Most established vendors run them. Trial before you buy, with your child, in your house.
  • If you are told no, ask which criterion was not met, in writing. A verbal "we don't do that here" is not a decision you have to accept. A written one gives you something to appeal.

    What the setup actually costs you

    Expect weeks of configuration, photo-taking and page-building to get the vocabulary reflecting your family rather than a generic template. Expect to be worse at the device than your child within a month. Expect a stretch, often around week three, where nothing is happening, your child is tapping randomly, and you feel foolish.

    That stretch is where most systems die. What gets families through is modelling: using the device yourself to say ordinary things all day, with no expectation that your child uses it back. If you are tapping symbols while you talk about breakfast and nobody is being tested, you are doing it right.

    A rejected system is data about the configuration, not proof that AAC has failed. Take it back to the SLP and change it.

    One disclosure, since it is relevant here: we make a speech and language practice app. It is not AAC and will not function as a communication system. If your child needs a way to be understood today, an app like ours is not the answer to that. Practice tools and communication systems solve different problems, and conflating them wastes time you do not have.

    FAQ

    At what age can a child start AAC?

    There is no minimum, and no developmental prerequisite that must be met first. Systems are introduced to toddlers routinely. If a child has something to say and no reliable way to say it, that is the criterion.

    Is PECS the same as AAC?

    PECS is one specific picture-exchange programme; AAC is the whole category. It suits some children and is limited for others, because it is built around requesting. Ask what your child would say on it beyond asking for things.

    How long before I know whether the system is working?

    Count initiations, not words. Pick a week about a month in and tally how often your child starts something themselves: a request, a refusal, a comment, a protest. Run the count again three months later. Rising initiation is the signal the configuration fits, and it moves before anything resembling vocabulary does.

    Related Reads

  • Helping a Nonspeaking Autistic Child Communicate
  • Will My Autistic Child Ever Talk?
  • Teaching an Autistic Child to Request
  • Try Verbalyft Free

    AI-powered speech therapy that makes practice feel like play. No credit card required.

    Start Free Trial