How Do You Analyze an AI Competitive Landscape?
A six-step method for mapping competition in any AI market, worked through on Voice with current FalcoScan data: who is in it, who has left, how the field splits, who leads, how much of it is thin wrappers, and what sits next door.
25 September 2026 · 9 min read · Data updated 25 September 2026
How do you analyze an AI competitive landscape?
Based on FalcoScan data updated 25 September 2026, work in six steps: define the market, count who is in it and who has left, segment the field by momentum, pricing and funding, find the leaders by opportunity, check how much of the field is thin wrappers, and compare it with adjacent markets. Worked through on Voice, that gives 298 active products and 28 shut down, 109 labelled hot, 4.6% likely wrappers against 16% catalog-wide, and a saturation score of 41 against 31 in Agents, 39 in Support and 44 in Sales.
The method is shown on one market, Voice, because it contains every state a landscape can hold: 109 products labelled hot and 28 labelled declining, 28 shut-downs and one acquisition, and neighbouring markets, Agents, Support and Sales, where voice products are sold too. Every number below is read from the FalcoScan catalog, and the same steps work for any of the 29 markets.
Step 1: What exactly is the market?
Decide what counts as the market first, because every number after this depends on it. FalcoScan files each product under one primary market, and its Voice market is everything filed there: 326 products tracked, 298 of them active. That is wider than any one product's competition. Sorted by what their taglines say they do, the active products fall into at least seven different kinds of product.
Active products have not shut down. Each product is counted once, under the first segment its tagline matches, so the split is approximate. Source: FalcoScan catalog, data updated 25 September 2026.
A speech recognition API does not compete with a song generator, even though both are Voice. If you are building one of these, the segment is your market, and the market-wide numbers in the next steps are the context around it. The split is by keywords in each tagline, so treat the counts as a guide to scale rather than an exact census.
Step 2: Who is in the market, and who has left?
FalcoScan has tracked 326 products in Voice. Of those, 298 are active, 297 operating and one acquired (Verbio AI Voice), and 28 have shut down: 8.6% of everything tracked, against 7.3% across the catalog.
Count companies as well as products. The 298 active products come from 149 different company websites, and Eleven Labs alone accounts for 15 of them. A field of 298 products is a different fight from one of 149 companies.
The shut-downs are concentrated too. PlayHT (7 entries) and Podcastle (6 entries) account for 13 of the 28. By segment, speech synthesis lost the most: 11 of the 28. Before reading a shut-down rate as a verdict on a market, check whether it is many companies failing or a few retiring several products.
Step 3: How does the field split by momentum, pricing and funding?
Split the active products three ways and compare each share with the catalog, because a share means little without a baseline. In Voice, 37% of active products are labelled hot, against 31% across the catalog. All 28 products labelled declining have shut down, and none of the 109 labelled hot has: in this market, the momentum label and survival point the same way.
| Measure | Group | Voice products | Share in Voice | Share across catalog |
|---|---|---|---|---|
| Momentum | Hot | 109 | 37% | 31% |
| Momentum | Rising | 135 | 46% | 48% |
| Momentum | Stable | 47 | 16% | 20% |
| Momentum | Declining | 0 | 0% | 0.6% |
| Pricing | Free | 34 | 12% | 8% |
| Pricing | Freemium | 178 | 60% | 49% |
| Pricing | Paid only | 83 | 28% | 43% |
| Funding | Bootstrapped | 63 | 21% | 17% |
| Funding | Pre-seed or seed | 43 | 14% | 20% |
| Funding | Series A | 83 | 28% | 26% |
| Funding | Series B or later | 11 | 4% | 6% |
| Funding | Growth | 70 | 23% | 18% |
| Funding | Public or acquired | 28 | 9% | 13% |
Active products only. Shares are of products with a known momentum label, pricing model or funding stage; the catalog column covers every active product in all 29 markets. Funding stage is that of the company behind the product. Source: FalcoScan catalog, data updated 25 September 2026.
Then repeat the split inside the market, segment by segment, because the market-wide figures average together businesses that behave very differently.
| Segment | Active products | Avg opportunity | Hot | Paid only | Bootstrapped | Wrapper risk 60+ |
|---|---|---|---|---|---|---|
| Music and sound generation | 67 | 62 | 31% | 15% | 31% | 0 |
| Speech synthesis, cloning and dubbing | 65 | 66 | 34% | 28% | 26% | 1 |
| Speech recognition and transcription | 46 | 69 | 53% | 28% | 17% | 2 |
| Audio editing, recording and clean-up | 29 | 60 | 14% | 17% | 28% | 0 |
| Voice agents and phone calls | 28 | 72 | 40% | 64% | 0% | 1 |
| Meeting recorders and note-takers | 19 | 65 | 39% | 22% | 26% | 8 |
| Real-time and conversational voice | 18 | 79 | 72% | 28% | 0% | 0 |
| Other voice and audio products | 26 | 67 | 33% | 42% | 15% | 1 |
Segments by tagline keywords, as in step 1. Avg opportunity is the mean product score, 0 to 100. Hot, paid only and bootstrapped are shares of active products in the segment with a known value. The last column counts active products scoring 60 or more on wrapper risk. Source: FalcoScan catalog, data updated 25 September 2026.
The real-time voice segment has the most momentum, with 72% of its products labelled hot, and the audio editing segment the least, at 14%. The voice agents segment is the most paid-first, with 64% of its products paid only, and none of them is bootstrapped. At the other end, 31% of products in the music and sound segment are bootstrapped, the highest share of any segment.
Step 4: Who leads on opportunity?
Opportunity is scored product by product, from 0 to 100, so the leaders are the products with the highest scores. These are the six highest-scoring products still operating in Voice, counting one per company website so that a company with several entries appears once. The two figures on each card are the product's opportunity and saturation scores.
AI medical documentation from clinical conversations
Emotionally intelligent voice AI for empathic applications
AI speech recognition API with best-in-class accuracy and features
Fastest AI speech recognition API for real-time applications
Real-time accent neutralization for customer service teams
Of the top ten, every one is labelled hot, eight come from companies at Series A or later, and eight are freemium. The real-time voice segment averages the highest opportunity, 79, and the audio editing segment the lowest, 60. Every product in the market, with its scores, is on the Voice market page.
Step 5: How much of the field is thin wrappers?
A wrapper is a product that is mostly a layer over someone else's model, and it can be copied as quickly as it was built. In Voice, 4.6% of tracked products score 60 or more on wrapper risk, against 16% across the catalog and 2.5% in the median market. Of the 29 markets, 15 have a lower share, eight of them none at all. The few there are sit together: of the 13 active likely wrappers, 8 are in the meeting notes segment, which has only 19 products. They include Fathom AI Notetaker, Fireflies AI Notetaker and Granola Notetaker. None of the ten leaders from step 4 scores above 17 on wrapper risk.
Read the share in both directions. A low share means incumbents own something of their own, a model, data or infrastructure, so a new entrant has to match that. A pocket of wrappers inside a segment is where a thin product would land, next to others that can be copied just as easily.
Step 6: How does Voice compare with adjacent markets?
Buyers do not shop by FalcoScan's market labels, so look next door. In Agents, 11 active products mention voice, speech or calls in their tagline, among them Hume EVI and PolyAI; in Support, 9, among them Parloa Voice AI and Elyos AI; in Sales, 18, among them Attention and Siro. A voice product for customer service or sales calls competes in those markets as much as in Voice.
| Market | Active products | Saturation | Opportunity | Wrapper share | Shut-down rate | Hot | Paid only | Bootstrapped |
|---|---|---|---|---|---|---|---|---|
| Voice | 298 | 41 | 66 | 4.6% | 8.6% | 37% | 28% | 21% |
| Agents | 381 | 31 | 73 | 19% | 7.1% | 57% | 24% | 14% |
| Support | 253 | 39 | 68 | 39% | 6.3% | 34% | 65% | 7% |
| Sales | 301 | 44 | 65 | 18% | 5.9% | 31% | 55% | 10% |
Saturation and opportunity are market averages, 0 to 100. Wrapper share: tracked products scoring 60 or more on wrapper risk. Shut-down rate: shut-down products as a share of everything tracked. Hot, paid only and bootstrapped: shares of active products with a known value. Source: FalcoScan catalog, data updated 25 September 2026.
Average opportunity and saturation across each market's products. Tool counts are active products, which have not shut down. Source: FalcoScan catalog, data updated 25 September 2026.
Agents is less crowded (saturation 31 against 41), has four times Voice's wrapper share (19% against 4.6%) and sells 24% of its products on paid plans only, against 28%.
Support is about as crowded as Voice (saturation 39 against 41), has nine times Voice's wrapper share (39% against 4.6%) and sells 65% of its products on paid plans only, against 28%.
Sales is more crowded (saturation 44 against 41), has four times Voice's wrapper share (18% against 4.6%) and sells 55% of its products on paid plans only, against 28%.
What does the worked example tell a founder?
Voice as a whole ranks 11th of 29 on saturation and 16th by active products. That average is not the market any single product enters. The segment with the highest average opportunity, real-time voice at 79, also has the most momentum (72% hot) and no bootstrapped products. The opening there suits a funded team more than a bootstrapped one. Wrapper risk is concentrated in the meeting notes segment, where 8 of 19 products are likely wrappers. And the speech synthesis segment accounts for the largest share of the shut-downs, 11 of 28.
The adjacent markets change the picture for anyone selling voice to a business: the same product can be filed, and compared, in a market with a very different wrapper share and pricing norm. Run the six steps on your own segment and on each market your buyers might file you under, and the landscape you end up with is the one you will actually compete in.
How does FalcoScan measure these?
Every product carries an opportunity score, a saturation score and a wrapper risk score, each from 0 to 100, plus a momentum label, a pricing model and the funding stage of the company behind it. Market figures are recomputed from these every day, and FalcoScan checks whether each product is still running, which is what separates active products from shut-down ones. The segments in steps 1 and 3 are this guide's own grouping by tagline, not a FalcoScan field. This page is rebuilt from the latest refresh, and the date at the top is the date of the data it used.
Where can you take this further?
Start with the Voice market page and its neighbours, Agents, Support and Sales, or compare every market at once in the Market Terminal. For the voice agents inside these markets, read how many AI voice agents are on the market, and for the checklist version of this method, what a startup should research before entering an AI market. The reports go deeper on single steps: what died, and where it went on shut-downs, the wrapper risk index on thin products, how AI prices itself on pricing and who funds AI, and at what stage on funding. For what moved this week, see the Pulse.
Questions people ask
How do you analyze an AI competitive landscape?
In six steps: define the market, count who is in it and who has left, segment the field by momentum, pricing and funding, find the leaders by opportunity, check the share of thin model wrappers, and compare it with adjacent markets. FalcoScan publishes each of these measures for all 29 AI markets it tracks and updates them daily; this guide works through them on Voice with data updated 25 September 2026.
How crowded is the AI voice market?
It sits in the middle of the 29 AI markets FalcoScan tracks: Voice scores 41 out of 100 on saturation, 11th of 29, and 4.6% of its tracked products are likely wrappers against 16% across the catalog, based on FalcoScan data updated 25 September 2026. It has 298 active products from 149 different company websites, and 28 have shut down.
Which AI voice products lead on opportunity?
Counting one product per company website and only products still operating, Eleven Labs (90), Abridge Medical (88) and Hume EVI Voice (88) have the highest opportunity scores in FalcoScan's Voice market, based on data updated 25 September 2026. All of the top ten are labelled hot.
Related articles
Photo: Tima Miroshnichenko / Pexels. Colour-graded for Falcoscan.
Citing these numbers
Every figure here is read from the FalcoScan catalog, and this page refreshes daily. The numbers above are from the data updated on 25 September 2026. When you quote one, quote that date with it.
FalcoScan, “How Do You Analyze an AI Competitive Landscape?”, data updated 25 September 2026. https://falcoscan.com/articles/how-to-analyze-an-ai-competitive-landscape