Skip to content
Say Yay
Start now

For speech-language pathologists

The home practice your families will actually do.

Thirty to ninety seconds a round, no score, and nothing that tells a child they said a word wrong. It runs alongside your plan and never competes with it.

What a family actually sees

  • A round green monster smiling just after copying a sound back, with the next sound to make printed large below him and a line describing the shape the mouth makes.
  • A picture to name and two bins to sort it into by the sound it starts with, each showing how many have gone in.
  • Two pictures labelled first and then, to be named in that order, with what it heard underneath.
  • Two pictures of a before and an after, the question they raise underneath, and a sentence with its middle left blank to fill in.

The decision that matters clinically

A child who cannot be understood still moves the story forward.

What the child meant decides what happens next. How clearly they said it never does. The two are judged separately and gated separately, so a sample can be perfectly clear in meaning and unusable clinically — a television in the room, a sibling talking over them — and the world still advances. That sample is set aside as unscored. It is never treated as a failure, because nothing here can fail.

Hold the meaning fixed, sweep the pronunciation result across its entire range, and the story does the identical thing. nine behavioral tests assert that, and the build does not ship if one of them stops being true.

Coverage

Six domains, not articulation alone.

  • Being understood

    12 games

    saying single words clearly, hearing the first sound in a word, where the tongue and lips go, hearing rhymes, hearing the beats in a word, keeping the quiet syllable in

  • More words

    12 games

    colours, sizes and amounts, naming things, words for where things are, copying a word back, making sound on purpose, number words, words for what things are doing

  • Putting words together

    9 games

    asking for what they want, one and more than one, two things in one sentence, describing something, joining two ideas, talking about what already happened

  • Listening and following

    2 games

    giving and following directions, following two steps in a row, what happened first, then next

  • Back-and-forth

    6 games

    taking turns, hello, goodbye and please, using language with people, trying again when not understood, answering who, what and where, remembering and retelling, saying why, telling what happened

  • Willing to try

    3 games

    loud and quiet on purpose, the tune of a sentence, leaning on the word that matters, which part of a word is loud

Reinforcement language

Your client will never hear any of these words.

“You did it!” celebrates the act, and is allowed. “You got it right!” passes a verdict on the production, and a verdict is a score. Lines of the second kind are refused before they can be spoken — not as a writing guideline somebody might forget, but as something the product has no way to say.

The same applies to what the game is able to do at all. There is no way for anything in it to put a score on the screen, show a correction, display the word it was hoping for, or ask a child to try again. Those are not discouraged. They are absent, and a line of character dialogue that asked for one would be discarded before the game saw it.

Using it with a family

What to tell a parent on your caseload.

  • Rounds are 30 to 90 seconds.

    The intake asks how long a bad day allows, five, ten or fifteen minutes, and builds the week to that rather than to an ideal one.

  • Missing two weeks costs nothing.

    There is no streak, no reset and no number that goes down. A family who stops and comes back finds exactly what they left.

  • It says out loud that it is not you.

    A parent who says they are seeing a therapist, or waiting for one, gets a plan that says in writing this runs alongside therapy, not instead of it.

  • It will not compete with your plan.

    There is no clinician login, no way to hand the scheduler a target sound, and no world mapped to one. Drilling a specific sound stays yours. This keeps them talking between sessions.

Where the limits are

What it will not do.

  • It does not diagnose.

    No score, level, percentile or severity exists anywhere in it. The intake outputs a practice plan, never a result about the child.

  • Coaching copy and the viseme set are not signed off.

    Progression logic and every mouth shape need licensed SLP review before release. That review is not complete.

  • The developmental-norms table is an engineering default.

    It stops the scheduler targeting a sound a child is too young for. No claim about children is made anywhere on this site.

  • No audio is recorded or transmitted.

    Nothing in the product accepts a recording, and no recogniser is connected to one.

  • It is not a screening tool.

    If a parent asks whether the app said something was wrong: it did not, because it cannot.

  • There has been no external validation.

    No trial, no published outcome data, no peer review. Treat it as a motivation surface, not as evidence.

Working with us

Send it to a family on your caseload.

If a claim here is imprecise, or a game asks a child to do something it should not, we would rather hear it from you than from a parent.
See the plan a family getshello@sayyay.app

Every absence described above is also listed with the place in the code it lives, if you would rather check than take our word for it, and what is coming next is published as well.