Last updated 2026-07-11
TL;DR
Feature matching starts with the child, never the device. You profile a child's motor, language, sensory, and cognitive abilities first, then compare those abilities to the technical features of candidate AAC systems. No single test produces a recommendation. The match gets built across several sessions and real-world settings, ideally with a speech-language pathologist who has AAC experience.
Feature matching is the method the AAC field uses to pick a communication system, and the logic runs backward from how most people shop for anything else. You don't start with a product and ask whether the child can use it. You start with the child and ask which system features line up with what that child actually does, sees, hears, and needs to say. The American Speech-Language-Hearing Association defines it as "the process of comparing an individual's abilities and needs to the characteristics or features of AAC systems and devices." [1] That sounds simple, but in practice it means building a detailed picture of a child across several domains before a single device is ever demoed.
Why go to all this trouble? Because device abandonment is real and well documented. Research in the journal Augmentative and Alternative Communication reports wide-ranging abandonment rates, with some studies putting the figure at roughly a third of device users, often because the device was picked without enough attention to the user's daily environment, motor access, or vocabulary needs. [2] Feature matching is the field's best answer to that problem, and it doesn't happen in a single Tuesday-morning appointment. It unfolds across observations, trials, and conversations with the people who know the child best.
Who should be involved
A speech-language pathologist with documented AAC experience should lead the process. ASHA's scope of practice for SLPs includes AAC evaluation and intervention. [3] SLPs vary widely in their AAC training, though, so it's fair to ask a prospective evaluator how many AAC evaluations they complete per year and whether they have access to a device lending library or trial systems. For children with complex motor needs, such as cerebral palsy or childhood apraxia of speech, an occupational therapist or physical therapist often joins the team to look at seating, positioning, and upper-limb access. A vision specialist should be involved too if there's any question about field of vision or tracking. Parents aren't passive observers in any of this. You're the expert on your child's stamina, frustration patterns, motivations, and what an actual day looks like for them, and a good team treats that knowledge as data, not background noise. Some families pursue an AAC evaluation through their school district under IDEA, which entitles eligible students to assistive technology evaluations at no cost. Others go through private clinics or children's hospitals, where insurance funding often requires a physician referral and a letter of medical necessity, with the evaluation report itself usually serving as the primary documentation for approval.
What gets assessed before any device shows up
This is the part most families don't know about: feature matching starts with a detailed profile of the child, not a catalog of devices. On the language side, the team looks at what communicative functions the child already uses (requesting, protesting, commenting, answering), what modalities they rely on, their receptive language level, and whether they show echolalia (see our article on echolalia for what that means for AAC selection). Cognitive and linguistic testing looks at whether the child can match symbols to referents, understand cause and effect, or categorize, since those abilities determine whether a system with deep vocabulary organization will work or whether a simpler layout is the better starting point. Sensory and perceptual factors matter just as much. Visual acuity, visual field, contrast sensitivity, and hearing all affect which display types and symbol sizes will actually be usable. A child with cortical visual impairment may do better with high-contrast, low-density displays even if they could cognitively handle something more complex. Motor ability assessment often decides more about device selection than anything else: can the child point with a finger, do they have reliable hand control, would they benefit from a keyguard, a head pointer, eye gaze, or partner-assisted scanning? Finally, the team looks at where the child spends their day, who their main communication partners are, and which activities matter most to them. A system that works beautifully in a therapy room but falls apart at the playground or in a noisy classroom is the wrong system. All of this gets documented before a device ever lands in front of the child.
Matching the profile to actual device features
Once the child's profile is built, the team compares it against specific, concrete features of AAC systems. Here's how the main categories line up with assessment findings.
| Feature category | What you're assessing | Example match |
|---|---|---|
| Vocabulary organization | Semantic, activity-based, or core/fringe? | Child uses early language = core vocabulary board |
| Symbol type | Objects, photos, line drawings, text? | Child matches photos but not line drawings = photo-based system |
| Display size and layout | Number of items per page, icon size | Low vision = large display, high contrast, 9 or fewer cells per page |
| Access method | Direct touch, scanning, eye gaze, head tracking | No reliable pointing = explore scanning or eye gaze |
| Speech output | Digitized (recorded) vs. synthesized (text-to-speech) | Literacy emerging = synthesized preferred for generalization |
| Portability and durability | Weight, case, water resistance | Active child, outdoor time = ruggedized, lighter device |
| Customization | Can vocabulary be added or reorganized? | Growing child = high customizability needed |
| Partner dependency | Independent use vs. aided input required | Beginning AAC user = assess whether partner will do aided language input |
This table isn't the whole picture. Systems also differ in their internal symbol libraries, the language levels built into preprogrammed pages, and their software ecosystems. Some run on dedicated devices, others are apps on commercial tablets, and neither category is inherently better: the match is what matters. [1] For children with apraxia of speech, motor speech differences heavily influence which access method and which output type will give the child the most independence.
Running the trial phase
Once you've narrowed things down based on the child's profile, you trial the top two or three systems with the child directly, and trials need to last long enough to mean something: usually two to four weeks per system in everyday settings, not a single clinic visit. During that time, the SLP or family tracks how quickly the child locates vocabulary, how often they initiate rather than just respond, whether error rates drop over time, and how communication partners interact with the system. You're also watching for fatigue, frustration, and whether the child reaches for the device unprompted. Many states and AAC vendors run device loan programs built for exactly this stage. ASHA maintains a list of state assistive technology programs, most of which are federally funded under the Assistive Technology Act and lend devices at no cost. [4] Ask your SLP about this before paying anything out of pocket for a trial device. Data collection doesn't need to be elaborate. A simple tally sheet tracking initiations, vocabulary used, and access errors across 10 to 15 minute sessions gives you enough to compare systems honestly, and parents or teachers can handle the observation with the SLP guiding the process. If a child has been through early intervention, those prior records, especially existing communication assessments, are useful context and should travel with the child's file.
Why there are no prerequisites for AAC
Some professionals, and unfortunately some school-based teams, still tell families a child has to demonstrate certain cognitive or language skills before they can have AAC. Current evidence doesn't support that, and neither does ASHA. ASHA's position states that "Individuals of any age, type, or severity of disability may benefit from AAC" and that there are no prerequisite cognitive or language skills required for an AAC trial. [3] This isn't a new idea: a frequently cited review in the Journal of Autism and Developmental Disorders found no evidence that symbolic understanding, cognitive ability, or oral language level should be treated as prerequisites. [5] In practice, this means feature matching and AAC trials can begin with a very young child, a child with significant cognitive delays, or a child who is still communicating pre-symbolically. The system selected will look different for each of them, but the right to a trial doesn't change. If a school team tells you your child isn't ready for AAC, cite ASHA's position directly and request a formal assistive technology evaluation under IDEA: that request triggers procedural timelines the district has to follow.
How does feature matching differ for children with autism versus other diagnoses?
The process itself stays the same no matter the diagnosis, but the profile that comes out of it looks different for a child with autism spectrum disorder than it would for a child with cerebral palsy or Down syndrome.
Autistic children often show stronger visual-spatial processing than auditory processing, so symbol-based and visual supports tend to work better than systems that rely on auditory prompting. Many autistic AAC users also have real strengths in reading or pattern recognition, and that shapes which symbol type makes sense. Echolalia, whether immediate or delayed, also changes how you think about language on a device: if a child's functional communication is mostly echoed phrases, an SLP with AAC experience can build vocabulary from what the child already says rather than starting from scratch.
Sensory sensitivities matter as well. Some children find certain device sounds unpleasant, and that alone can determine whether they'll actually use the system. Volume control, voice type, and vibration feedback are all worth checking during trials.
Motor profiles vary more than most people expect, too. Some children have strong fine motor control and can direct-access a densely packed grid without trouble. Others have motor planning difficulties in line with apraxia of speech, which affects both their spoken output and how they physically access a device.
There's no single autism AAC profile. Two children with the same diagnosis can end up with completely different feature matches, and that's exactly the point of doing the process properly.
What does a feature matching report look like, and how is it used for funding?
A full feature matching assessment ends with a written report. A good one summarizes the child's communication profile across every domain assessed, lists the systems trialed along with the data from each trial, ties the recommended system back to specific features of the child's profile, and includes an implementation section covering vocabulary and partner training.
That report does two jobs: it gets the team started with the right system and vocabulary, and it's usually the main document submitted to insurance or Medicaid for funding. Medicare and Medicaid have specific coverage categories for speech-generating devices; for children, Medicaid coverage falls under EPSDT, which in most states requires covering medically necessary assistive technology regardless of cost. [6] Private insurance requirements differ by plan and state.
A report should never name a specific brand without explaining why that brand's features fit this particular child. A report that simply says "the child needs a Tobii" or "the child needs a Snap Core system" with no rationale behind it is weaker for funding purposes and probably doesn't reflect real feature matching.
Schools sometimes use this same report for an IEP assistive technology goal and may fund a device under IDEA, though rules vary on whether a school-funded device can go home with the child. Some families end up pursuing funding from both sides (school for educational use, Medicaid for home use), often with help from an AAC advocate.
What are the most common mistakes families see during AAC assessments?
The most common one is an assessment that's really just a product demo: a clinician brings in a device or two, shows the child how it works, watches for twenty minutes, and writes up a recommendation. That's not feature matching, that's guessing with props.
Another problem is leaning too hard on standardized cognitive tests as gatekeepers. A child who scores poorly on a nonverbal IQ test in a clinical room may show very different abilities in their own environment. A real feature matching process weighs naturalistic observation more heavily than structured testing.
Families also run into evaluators who don't know the full range of AAC options, especially low-tech ones. Paper-based boards and PECS-style systems often get treated as a fallback, when for some communicators they're a legitimate long-term solution. Low-tech options belong in the comparison alongside electronic devices, not as an afterthought.
An assessment without an implementation plan is only half finished. It doesn't matter which device fits if the teachers, parents, and aides around the child don't know how to model it. Aided language input, where communication partners use the device themselves to show the child how language works on it, is one of the best-supported strategies for building AAC use, and any solid report should recommend it. [7]
How do you document progress after the feature match is made?
Choosing a system is the beginning, not the end. Once it's in place, ongoing data collection tells you whether the match was right and when it's time to revisit it.
The Communication Matrix, developed by Charity Rowland at Oregon Health and Science University, gives families and clinicians a structured way to track communication across seven levels, from pre-intentional behavior up through language use. [8] It's free online and takes about 20 minutes. Running it at baseline and again every few months gives a clear picture of whether communication is actually growing.
Worth tracking specifically: how many different vocabulary items the child uses per session, how many utterances they initiate on their own (not prompted), whether message length is increasing, and whether they're using the device across different settings and with different people. Movement in these numbers is a good sign the match is working.
If progress stalls, the first question is almost always about implementation, not the device itself. Is it charged and available all day? Are the adults around the child actually modeling with it? Is the vocabulary organized around what the child wants to say? Most AAC specialists will troubleshoot implementation for several months before concluding the device itself is the wrong fit.
Some children outgrow their original feature match as they develop. A child who started on low-tech boards may gain the motor control for direct touch on a dynamic display. A child who started with photo symbols may develop the symbolic understanding to move to line drawings or text. Feature matching isn't a one-time decision, it's something you return to whenever a child's abilities, environments, or needs change significantly.
Families looking for a way to build vocabulary between therapy sessions can try Little Words, which has a structured quiz to help pin down communication goals and figure out where a child is starting from.
What should parents ask for when requesting a feature matching evaluation?
When you contact a clinic, school, or insurance plan about an AAC evaluation, being specific gets you further than a general request will.
Ask for a full AAC evaluation using a feature matching approach: direct trials of at least two candidate systems in everyday settings, both low-tech and high-tech options included in the comparison, a written report with a funding rationale, and a plan for training the people who communicate with your child.
If you're going through a school district under IDEA, put the request in writing and date it. The district's evaluation timeline starts from that date, typically 60 days under federal guidelines, though this varies by state. [9] Keep copies of everything you send.
Families working with online speech therapy should confirm the remote SLP has access to device loan programs and can coordinate with a local team for the hands-on trial phase. A remote SLP can contribute a lot to the assessment, especially the observation and report-writing, but the actual device trials are harder to run at a distance.
Little Words' quiz at littlewords.ai/start can help you organize what you've noticed about your child's communication before that first evaluation appointment, which tends to make the appointment itself go faster and further.
Common questions about feature matching and AAC
There's no minimum age for a feature matching AAC assessment. ASHA's position is that AAC evaluation is appropriate for individuals of any age or ability level, and in practice many children are evaluated during early intervention, before age three, especially when there's a significant communication delay. Earlier trials mean earlier access to communication, and that matters for language development.
A thorough assessment, covering the initial profile, device trials, and report writing, usually takes four to eight weeks from the first appointment to the final report. The trial phase alone should run at least two to four weeks per system in real-world settings. One-day assessments that skip trials aren't considered best practice, and they tend to produce weaker documentation for funding.
Your child doesn't need to speak at all to use AAC. It's appropriate for children who are non-speaking as well as those with some speech that isn't enough for their needs, and children who are minimally verbal or non-verbal are often the clearest candidates. Research also shows AAC doesn't inhibit speech development: in most studies, it supports it.[7]
Low-tech systems, such as paper-based communication boards, PECS books, and printed core vocabulary boards, have no electronics, no charging, and no breakage risk. High-tech systems are electronic speech-generating devices, either dedicated hardware or tablet apps. Feature matching looks at both categories: some children do best with low-tech permanently, while others use it as a backup to a high-tech system.
A school can't refuse a written request for an assistive technology evaluation if your child is eligible under IDEA and AT might help their education. If a district refuses anyway, it has to provide prior written notice explaining why, and parents can then request mediation or a due process hearing. Citing ASHA's no-prerequisite position and IDEA's AT mandate in your written request strengthens your case.
Many private insurance plans cover AAC evaluations when they're billed as a speech-language pathology evaluation, especially with a physician referral, and Medicaid covers these evaluations for children under EPSDT in most states. Medicare has its own SGD coverage criteria for adults. Get pre-authorization when you can, and ask your SLP to spell out medical necessity clearly in the referral.
Aided language input means communication partners use the child's AAC system themselves to model language throughout the day, rather than only prompting the child to use it. It's one of the most evidence-backed strategies for building AAC use,[7] and a feature matching report that skips implementation guidance for it is incomplete.
To tell whether the selected system is working, track how many different vocabulary items your child uses independently, how often they initiate communication instead of just responding, and whether they use the device across different settings. The Communication Matrix (free at communicationmatrix.org) gives you a structured way to measure that growth every few months.
Feature matching can and often should be repeated for a child who already has a device. Kids change. If a device was chosen two or three years ago without a real feature matching process, or if your child's abilities have shifted, a reassessment makes sense, and it's also the right move when a funded device is due for replacement, since insurers often want updated documentation.
Parents bring information no clinical assessment can capture: daily routines, motivations, frustration patterns, environments, and who the child communicates with. You should be present for observation sessions, help build the communication profile, and take part in training on whatever system gets recommended. If an SLP sidelines you during the evaluation, treat that as a red flag.
Eye gaze technology lets a child control a device by tracking where they look. It gets assessed for any child who lacks reliable hand or head control, and it requires good visual acuity and consistent access to their visual field. Because eye gaze systems cost more than touch-access devices and take longer to calibrate, they're usually recommended only when other access methods genuinely aren't workable.
A standard speech evaluation looks at articulation, language skills, fluency, and voice. Feature matching for AAC goes further, into motor access, sensory profile, symbol understanding, environmental factors, and device trials, and it produces a different kind of output: a device recommendation with funding rationale, not just a therapy plan. Some SLPs handle both in one process; others refer out for the AAC piece.
As for vocabulary, the main frameworks are core vocabulary (high-frequency words that work across contexts), fringe vocabulary (topic or person-specific words), activity-based vocabulary, and whole-word or phrase-based systems. Feature matching weighs which organization fits the child's language level, cognitive profile, and daily communication needs, and most well-built systems end up combining core vocabulary with customizable fringe pages.
Sources
- ASHA, Augmentative and Alternative Communication: Overview: ASHA defines feature matching as 'the process of comparing an individual's abilities and needs to the characteristics or features of AAC systems and devices.'
- Augmentative and Alternative Communication journal (Taylor & Francis): Research in Augmentative and Alternative Communication documents AAC device abandonment rates, with some studies reporting abandonment in roughly a third of device users.
- ASHA, Scope of Practice in Speech-Language Pathology: ASHA's scope of practice includes AAC evaluation and states that individuals of any age, type, or severity of disability may benefit from AAC with no prerequisite skills required.
- ASHA, State Assistive Technology Programs: ASHA notes state assistive technology programs are federally funded under the Assistive Technology Act and offer device lending at no cost.
- Journal of Autism and Developmental Disorders, Schlosser & Wendt 2008 (no-prerequisites research): Research in JADD found no evidence that symbolic understanding, cognitive ability, or oral language level should be treated as prerequisites for AAC.
- Medicaid.gov, Early and Periodic Screening, Diagnostic and Treatment (EPSDT): Medicaid coverage for children under EPSDT requires coverage of medically necessary assistive technology in most states regardless of cost.
- ASHA, Aided Language Input / Modeling: Aided language input, where partners model language using the AAC system, is one of the most evidence-backed strategies for building AAC use and does not inhibit speech development.
- Communication Matrix, Oregon Health & Science University: The Communication Matrix tracks communication functions across seven levels from pre-intentional to language use and is free online.
- U.S. Department of Education, IDEA Individuals with Disabilities Education Act: Under IDEA, school districts must complete evaluations within 60 days of written parental request, and IDEA mandates assistive technology evaluations for eligible students at no cost to families.
- ASHA, AAC Evidence Maps: ASHA's evidence maps document research support for AAC interventions across populations including autism and childhood motor speech disorders.
- American Academy of Pediatrics, AAC and Communication Supports Policy: AAP supports early identification and referral for communication supports including AAC for children with developmental and communication disorders.