What are the most common problems with dictation apps?

The most common complaint about dictation apps is not that they mishear you. It is that they do not work when you need them. Reliability is the top problem in critical App Store reviews: 42.9 percent of them report it ¹. Price and subscription comes second at 29.2 percent ², and recognition accuracy third at 21.0 percent ³.

The figures are shares of documents we coded, not shares of users. A review counts toward a theme only when it states that problem, and one review can count toward several themes.

Reliability tops the list in every group of App Store reviews

Among reviews of 1 to 3 stars, reliability is reported more than twice as often as accuracy. In reviews that are not filtered by rating, the order is the same but the levels are lower: 20.1 percent report reliability ⁴, 12.5 percent report price ⁵ and 10.9 percent report accuracy ⁶.

The table lists the problem themes for all App Store reviews we coded.

Showing 8 of 19 rows. See every row in the dataset.

Reliability matters most in professional dictation apps, where 34.1 percent of reviews report it ¹³. In meeting note takers the share is 15.2 percent ¹⁴, and in transcription apps 16.3 percent ¹⁵. People who rely on dictation for daily work notice a failure most.

Cost and support complaints come right after the technical ones

Price and subscription is a bigger complaint than accuracy among critical reviews. Customer support follows at 16.3 percent ¹⁶, and setup and permissions at 10.7 percent ¹⁷. Setup is a larger issue for professional dictation apps, at 17.1 percent ¹⁸.

Speed is a smaller complaint: 9.0 percent of critical reviews mention delay ¹⁹.

Developers and Bluesky users talk about accuracy first

Hacker News commenters and Bluesky posters describe different problems from App Store reviewers. Recognition accuracy is the top problem we measured there: 12.1 percent of coded Hacker News comments ²⁰ and 14.7 percent of coded Bluesky posts ²¹ report it.

These sources are not comparable to reviews. Many commenters discuss tools they build or test, and they write about dictation in general, not about one purchase.

What the numbers do and do not measure

• Not a random sample. App Store pages show the reviews Apple selects, often the most helpful ones. Hacker News comments come from developers and founders.
• Stated problems only. A theme counts only when the document says it. A problem nobody mentions does not appear, even if it is common.
• Shares, not severity. A rare problem can be serious, and a common one can be minor.
• Rating filter. The 1 to 3 star group is critical by design, so its shares are higher than for all reviews.
• Different sources, different questions. Compare themes within a source before comparing across sources.

Documents that used the words in another sense, or said nothing about using speech to text, were set aside and counted in the record. Groups of fewer than 30 documents are left out. The rows and their record are in the full dataset.

What this means when you pick a dictation tool

1. Test reliability first. Dictate for a full working day in the apps you actually use, not one short sample. Reliability is the problem users meet most.
2. Read the pricing before you start. Subscription complaints are common, so check what is free, what renews and what is capped.
3. Check setup and support. Permissions and slow support show up in reviews, so see how an app handles both before you depend on it.
4. Do not choose on accuracy claims alone. Accuracy matters, but it is not the main reason people give a low rating.

For a comparison of products, see the market overview, our overview of dictation software and the guide to the best dictation software.