Useful guidance on littleWords has to respect neurodivergent kids and exhausted families at the same time. The right plan is gentle, repeatable, and clear about when an SLP should guide the next step.
Let me get the conflict of interest on the table. I am the dad of a four-year-old autistic daughter with a moderate expressive language delay. I spent about three years trying to find a speech app that worked for her. I tried eight of them. None did what I needed. So I built one myself. It’s called LittleWords, and yes, this post is partly about a product I made.
But the real reason I’m writing this is simpler: I want to give other parents the diagnostic framework I wish someone had handed me in 2021. I’ll name apps. I’ll be specific about what failed and why. Not to trash anyone’s work, but because vague app reviews help nobody, and the parents shopping for these tools at midnight are too tired for vague.
The Night I Stopped Trusting App Store Ratings
In March of last year, my friend Rana in Austin texted me a screenshot. Her son Elias, age five, nonverbal until about three and a half, had been using one of the top-rated speech apps for six weeks. “He scored 94% on Level 3!” the dashboard announced. Rana was thrilled. I asked her one question: “Is he actually saying more words at dinner?”
Long pause. Then: “No. He just got really good at tapping the right picture.”
That moment crystallized something I’d been feeling for months. The metrics these apps optimize for and the outcomes parents actually care about are frequently two different things. Elias could ace a picture-matching drill. He still couldn’t ask for more juice. A 94% score was meaningless. Worse, it was reassuring in a way that delayed Rana from pushing harder on his SLP referral.
That gap between dashboard performance and real-world speech is where most of these apps fail, and it’s where my rubric starts.
Five Criteria That Actually Matter
I developed these through trial and error. Mostly error.
Pause tolerance. How long does the app wait for the kid before it moves on or re-prompts? My daughter needs at least five seconds. Most apps wait two, which is standard for adult voice products and absolutely wrong for a kid with processing delays.
Approximation acceptance. If my kid says “ca-keh” for cracker, does the app recognize that as cracker or mark it wrong? Approximations are how kids learn language. Penalizing them is bad design dressed up as rigor.
Interest-driven flow. Does the app follow what my kid wants to talk about, or does it force a preset curriculum? Forced curricula do not work for most autistic kids. Full stop.
Pressure design. Stars, levels, streaks, scoring, time limits. All pressure cues. Pressure shuts down language production in kids who are already anxious about speaking. This is not controversial in the SLP literature; it’s just ignored in app design.
Sensory and prosody. Is the voice flat or expressive? Are the animations calm or seizure-inducing? Is the audio mix appropriate for a sensory-sensitive kid?
That’s the rubric. Now the apps.
The Pediatrician’s Flyer App
You know which one. The pediatrician hands you the brochure. The app has cartoon characters and badges. Good intentions, clean visuals, developmentally appropriate vocabulary.
It failed every single one of my five criteria. Two-second pause window. No approximation tolerance. Forced curriculum. Stars and stickers everywhere. Hyperactive animations.
My daughter lasted four days. She started melting down when I opened the app. We deleted it.
The Forty-Dollar Dolphin
Sleek branding. Targeted ads to special needs parents. Production value through the roof. Responsive customer service.
The dolphin character was too animated. The voice was condescending. The “lessons” felt like corporate onboarding for small children. There was a progress dashboard I was supposed to monitor as a parent, which is a red flag I’ll come back to.
The kid hated the dolphin. I hated paying forty dollars a month. Three weeks. Canceled.
The University AAC App
Free, from a major university. Symbol-based, PECS-style, limited audio. Clean interface, calm design, appropriate audio. Actually a solid tool.
Here’s the thing: it wasn’t really a speech app. It was an AAC introduction app. We needed something for spoken language practice. My fault for trying it for the wrong use case.
I want to be clear. I’m not arguing AAC is inappropriate. AAC is foundational for many autistic kids. But the AAC app and the spoken-language practice app are different products for different needs. I needed the second kind.
The Everything-App With the Bear
One of those apps bundling literacy, math, and “social-emotional learning.” Massive subscriber base. Heavy advertising. Reasonable price. Real breadth.
Nothing in it was designed for kids with speech delays, let alone autistic kids. Voice prompts too fast. Choice timers too short. The “you got it!” cheering was overwhelming for my daughter’s sensory system. The bear was, frankly, a lot.
Two weeks. Gone.
The SLP Drill Tool
A clinical tool that SLPs use in sessions, available to parents at a premium. The clinical rigor was visible. Vocabulary lists aligned with actual SLP curricula.
But the format was drill. Show a picture, prompt the word, score the production. For some kids in clinical contexts, this works. For my daughter in our living room at 7 p.m. after a long day at preschool, it was the opposite of what she needed. She would not engage.
One month. Stopped.
The Generic AI Chatbot
A general-purpose AI chatbot marketed as a learning companion for kids. Not designed for speech delays. The conversational openness was genuinely interesting. The bot would let my daughter talk about whatever she wanted. That was the right format.
Everything else was wrong. The bot couldn’t understand approximate speech. It routinely misinterpreted her words and responded to the wrong thing. It also wasn’t COPPA-compliant in a way I was confident in, and it stored conversation data on servers I didn’t control. Pulled it after about a week for privacy reasons alone.
The conversational format was right. The execution was wrong. I kept thinking about that.
The Indie Fox App
Smaller indie app. Reasonable price. Designed for kids with speech delays but not specifically for autistic kids. The pace was slower than most. The fox character was calm. Visual design was uncluttered.
But the curriculum was still forced. The app moved my daughter through lessons whether or not she was engaged. The fox would prompt her with the same phrase repeatedly, which she found frustrating. Pause window still too short.
Six weeks. Diminishing returns. Stopped. Of all eight apps, this one came closest. The spirit was right; the mechanics weren’t.
The “Evidence-Based” Startup With the 40-Minute Onboarding
Most polished of the bunch. The onboarding process was thorough, almost impressively so. The marketing copy was loaded with citations.
The personalized plan it generated was, in practice, just a sequence of drills in a nicer wrapper. The “evidence base” consisted of citations to general speech-development research, not research on autistic-kid-specific outcomes. There’s a meaningful difference between “studies show repetition aids vocabulary acquisition in typically developing preschoolers” and “this specific approach works for autistic kids with expressive language delays.” The first does not imply the second.
The onboarding promised more than the product delivered. We stopped after two months.
What I Built, and Where It Falls Short
After app eight, I was out of options. I’m an engineer. I built LittleWords for my daughter first. Then for other parents’ kids. Then as a product, currently in waitlist phase with a Founding Family $49 lifetime offer.
The app is a conversational AI companion called Buddy. The kid talks. The app waits as long as needed. The app responds in a warm, prosodic voice. The kid talks again. No scoring. No leveling. No streaks. No timer. The character follows the kid’s interest. Approximations are accepted as the word the kid meant. Kid data is COPPA-compliant and never stored or sold.
It’s not an AAC replacement. If your kid uses or needs AAC, that’s a different category of product entirely. LittleWords is for kids who can produce some words and need a low-pressure conversational practice partner. Target age is roughly three to ten. Target use case is autistic and speech-delayed kids who don’t do well with drill-based apps.
How does it score on my own rubric?
Pause tolerance: eight seconds default, adjustable. Approximation acceptance: yes, with a custom layer that maps kid approximations to intended words. Interest-driven flow: the character follows what the kid talks about. No curriculum forcing. Pressure design: zero gamification. No stars. No levels. Sensory and prosody: voice is expressive, animations are calm, audio mix designed with SLP input for sensory sensitivity.
Five for five on my own rubric. That was the bar I built to, because it was the bar I couldn’t find anywhere else.
Now, the weaknesses. I want to be honest about these because I spent eight apps’ worth of frustration on products that hid their limitations.
The app is not designed for kids under three. The conversational format requires some emerging language. If your kid is producing zero or near-zero spontaneous words, this isn’t the right tool yet. Start with snack-time speech practice, sign language, and an SLP. Add the app later.
It’s not a diagnostic tool. It will not tell you if your kid is autistic. Get a real evaluation.
It’s not a substitute for an SLP. If your kid is in therapy, stay in therapy. If your kid is on a waitlist (and I know how brutal those waitlists are), the app can help fill the gap, but it doesn’t replace clinical work.
There’s no parent dashboard. This is deliberate. I don’t want parents quantifying their kid’s progress in app metrics. Rana’s son scored 94% and still couldn’t ask for juice. The progress is in the kid, not in the dashboard. If you want detailed tracking, your SLP can provide it.
English only as of this writing. We’ll expand.
These are real limitations. I want you to know them before you join the waitlist.
What I’d Tell Another Parent Shopping for a Speech App
Eight things, distilled from three expensive years.
Use my rubric or build your own. Pause tolerance, approximation acceptance, interest-driven flow, pressure design, sensory and prosody. Score every app on these five dimensions before you commit money or (more importantly) your kid’s emotional bandwidth.
Watch your kid, not the screen. The kid’s body will tell you if an app is working. Avoidance, meltdowns, throwing the tablet: those are signals. Engagement, return visits, voluntary use: also signals. Trust the body over the progress bar.
Beware of dashboards. Apps that emphasize parent dashboards are often optimizing for parent engagement, not kid outcomes. Dashboards are like the calorie counter on the elliptical machine. They make you feel like something is happening. Sometimes something is. Often it isn’t.
Beware of free trials that autoship. The high-priced subscription apps target desperate parents. Read the cancellation terms before you enter your card number.
Beware of “evidence-based” without specifics. Anyone can claim evidence-based. Ask what the evidence is. Read the citations. In this product category, most of them won’t hold up to even casual scrutiny.
Beware of one-size-fits-all curricula. Your kid is not the average kid. (No kid is the average kid, but especially not yours.) Apps that force a linear curriculum will not adapt.
Pay attention to privacy and COPPA. Kid data is sacred. Apps that aren’t COPPA-compliant should not be on your kid’s tablet. Period.
Try one app at a time. Don’t stack. You won’t be able to tell what’s working if everything is happening at once. Give each tool at least three weeks of consistent use before you evaluate.
The Honest Ending
Eight apps. Three years. One that I eventually had to build because I couldn’t find what I needed.
If you’re in the same spot I was, my hope is that this post saves you some of the time I lost. The rubric is real. The privacy concerns are real. The need for low-pressure, interest-driven, conversational tools for autistic and speech-delayed kids is real, and the market is not meeting it well.
If LittleWords is the right fit for your kid, I’d love to have you on the waitlist. If it’s not, I genuinely want you to find what is. The kid in front of you deserves the right tool. Spending another year on an app that’s fighting your kid’s neurology isn’t worth it. Spending another year on the right one is worth everything.
That’s the honest review. Take what’s useful. Ignore what isn’t. Watch your kid. Trust your gut. The right tool exists, or, like me, you might end up building it yourself.
Frequently Asked Questions
Is LittleWords a replacement for speech therapy? No. It’s a practice tool, not a clinical intervention. If your child is working with an SLP, keep that relationship. If you’re on a waitlist, the app can provide meaningful practice in the meantime, but it doesn’t replicate what a trained therapist does in a session.
What age range is the app designed for? Roughly three to ten. The conversational format requires some emerging spoken language, so kids who aren’t yet producing any spontaneous words will likely not benefit from the current version.
Is it safe for my kid’s data? The app is COPPA-compliant. Conversation data is not stored, sold, or used for training. This was a non-negotiable design requirement from day one.
Does it work for kids who aren’t autistic but have speech delays? Yes. The design principles (long pauses, approximation acceptance, no pressure cues) benefit any kid with an expressive language delay. It was built with autistic kids as the primary audience because that’s my daughter’s profile, but the approach generalizes.
How is this different from a regular AI chatbot? Three main ways: the speech recognition layer is tuned for child approximations rather than adult speech, the interaction design removes all pressure and gamification elements, and the privacy architecture is built for COPPA compliance from the ground up. A general chatbot has none of these.
What does it cost? Currently in waitlist phase with a Founding Family $49 lifetime offer. No subscriptions, no autoship, no free-trial traps.
Can I use it alongside AAC? Absolutely. LittleWords is not an AAC replacement and doesn’t compete with AAC tools. If your child uses AAC, that’s a separate and important part of their communication system. This app focuses specifically on spoken language practice for kids who have some emerging verbal output.
