Which AI customer service startup is growing the fastest?
In our conversational AI market deck, you will find everything you need to understand the market
SUMMARY
Sierra is the fastest-growing AI customer service startup overall today, while Wonderful is growing fastest in pure percentage terms.
The distinction is important because the two companies are winning very different growth races. Wonderful has gone from roughly $1 million to around $70 million in annualized revenue in about a year, while Sierra has built a much larger recurring-revenue base and has now disclosed $200 million ARR in nine quarters.
Decagon is the most credible direct challenger to Sierra. Its estimated annualized revenue has moved toward roughly $100 million from about $44 million at the end of 2025, while the company added more than 100 global enterprise customers during that year.
The strongest companies are no longer being judged mainly on pilots or funding rounds. Sierra says its agents now power billions of customer interactions, Bland handles more than 3.5 million calls a week, and large customers are putting these systems into claims, returns, financial workflows and other production tasks.
Customer expansion is becoming almost as important as new-logo growth. Parloa's 150% net revenue retention and Ada's 146% show that successful deployments are spreading into more workflows after the initial sale.
Wonderful's growth is the most spectacular mathematically, but it is also the least clean comparison. The company is broadening beyond customer service into enterprise AI infrastructure and uses a forward-deployed engineering model with reported gross margins around 52%.
Sierra's advantage is absolute growth at scale. Moving from an estimated roughly $26 million ARR at the end of 2024 to $200 million ARR in nine quarters requires far more new recurring business than a smaller company needs to post a higher percentage increase.
Enterprise-logo momentum points in a slightly different direction. Decagon appears to be signing large companies faster, while Sierra has penetrated deeper into the top of the market and says it works with more than 40% of the Fortune 50.
Voice AI is becoming its own serious subcategory. Bland's weekly call volume proves large-scale adoption, but without a recent revenue denominator it is still hard to compare its commercial growth with Sierra, Decagon or Parloa.
The ranking therefore depends on what we mean by fastest. Wonderful wins on multiplication, Decagon looks strongest as the fast-closing challenger, and Sierra has the best overall combination of revenue added, scale, enterprise penetration and production usage.
This market map, featured in our conversational AI market deck, highlights top companies and startups in the conversational AI market
Which AI customer service startup is growing the fastest?
Why is it suddenly hard to tell which AI customer service startup is growing fastest?
The AI customer service race has become genuinely difficult to rank because several startups are growing at extreme rates, but they are winning on very different numbers.
Sierra has already built a nine-figure recurring-revenue business. Decagon is roughly doubling from a smaller base and keeps signing large enterprises. Wonderful has produced the wildest percentage increase we found, going from about $1 million to roughly $70 million in annualized revenue in around a year. Parloa and Ada are showing unusually strong expansion inside existing customers. Bland has reached millions of voice calls every week.
A year or two ago, most of these companies could be judged mainly on pilots, funding rounds and product claims. Today, enough customers have gone live that we can compare actual revenue, usage and account expansion.
That creates the central problem with the title. If “fastest” means highest percentage growth, Wonderful is ahead. If we mean the largest amount of new recurring business created while already operating at scale, Sierra has the stronger case. Decagon sits between those two extremes.
What does “growing fastest” mean for an AI customer service startup?
For an AI customer service startup, we think the best measure of growth today is how quickly real recurring business is expanding, with customer adoption and production usage used to check whether that revenue is durable.
Percentage growth alone can be misleading. Moving from $1 million to $5 million means growing 400%, while moving from $100 million to $150 million means growing only 50%. Yet the second company added more than ten times as much revenue.
Customer counts have a similar problem. One Fortune 50 rollout can be worth more than dozens of smaller accounts. Usage helps, although millions of conversations do not automatically translate into large revenue.
So we looked at the measures together. Revenue trajectory receives the most weight. Customer expansion and actual usage tell us whether that revenue reflects deployment rather than experimentation. Funding and valuation are useful mainly as secondary evidence.
| Growth test | Current leader | What the evidence shows |
|---|---|---|
| Fastest percentage revenue growth | Wonderful | About $1M to roughly $70M annualized revenue in around a year |
| Most revenue added at scale | Sierra | From roughly $26M estimated ARR at end-2024 to $200M ARR disclosed in nine quarters |
| Fastest direct challenger | Decagon | About $100M estimated annualized revenue after finishing 2025 near $44M |
| Strongest disclosed customer expansion | Parloa | 150% net revenue retention |
| Fastest voice-AI usage growth | Bland | More than 3.5M calls a week |
As this chart shows, and as featured in our conversational AI market deck, search interest in conversational AI has increased sharply
Is Sierra the fastest-growing AI customer service startup right now?
Yes, Sierra currently has the strongest claim to being the fastest-growing AI customer service startup overall because no close peer combines its revenue size with the same pace of expansion.
Sierra publicly launched in early 2024. Seven quarters later, the company said it had crossed $100 million in annual recurring revenue. In its year-two review, Sierra disclosed that it had recorded its first $50 million quarter and entered its third year above $150 million ARR. It has since disclosed $200 million ARR in nine quarters.
The starting point makes that much more impressive. Sacra estimates Sierra ended 2024 at around $26 million ARR. Even if we use that only as a rough historical anchor, the gap to the company's latest first-party disclosure is enormous.
Very few enterprise software businesses add that much revenue so early in their life. Sierra managed it while moving deeper into large regulated companies, where sales cycles and integrations are usually harder.
That is why Sierra leads our ranking. Wonderful has multiplied faster, but Sierra is creating more new customer-service revenue from an already substantial base.
If you want more recent data on this point, please see our latest conversational AI market report.
Is Wonderful actually growing faster than Sierra?
Wonderful is growing much faster than Sierra in percentage terms, and the difference is enormous.
The Wall Street Journal's latest reporting puts Wonderful at roughly a $70 million revenue run rate, compared with about $1 million a year earlier. Earlier checkpoints make the trajectory easier to believe: reported annualized revenue moved through roughly $7 million and then around $18 million before reaching its current level.
A rise from $1 million to $70 million works out to roughly 70 times the starting revenue. Even allowing for some uncertainty around private-company run-rate figures, nothing else in our core group comes close to that percentage increase.
Sierra's climb is harder in another sense. Adding tens of millions of dollars when revenue is already well into nine figures requires much more absolute demand than multiplying from a tiny base.
So there are two defensible winners. Wonderful has the fastest raw growth rate we found. Sierra is growing fastest once we care about both speed and scale.
This chart, included in our conversational AI market deck, shows annual VC investment in conversational AI startups
Is Wonderful still really an AI customer service startup?
Wonderful still belongs in this comparison, although its newest strategy makes it a less clean customer-service company than Sierra, Decagon or Parloa.
Wonderful initially grew by building multilingual AI agents for customer support, especially in markets outside the United States and Canada. Sacra says the company now serves roughly 100 enterprises across about 50 markets, while recent Wall Street Journal reporting describes operations in 30 countries.
That geographic approach explains part of the growth. Wonderful sends forward-deployed engineers into large enterprises and adapts its systems to local languages, workflows and regulations. It can therefore enter markets that English-first AI companies may address more slowly.
Lately, though, Wonderful has been expanding into a wider enterprise AI platform. Its products increasingly cover autonomous business workflows beyond customer support, with the company presenting itself as infrastructure for enterprise AI deployment rather than only a customer-service agent.
Its economics also look different. The Wall Street Journal reported gross margins of roughly 52%, which is low for a conventional mature software business and fits the company's labor-intensive deployment model.
We still count Wonderful because customer service produced much of its early commercial breakout. But the broader its revenue becomes, the harder it is to compare every Wonderful dollar directly with Sierra or Decagon.
If you want more recent data on this point, please see our latest conversational AI market report.
Is Decagon catching Sierra in AI customer service?
Decagon is the strongest direct threat to Sierra today, but it has not caught Sierra on revenue scale.
Decagon said more than 100 new global enterprise customers joined during 2025, including Avis Budget Group, Block and Deutsche Telekom. More recently, Sacra estimated that Decagon reached roughly $100 million in annualized revenue, up from approximately $44 million at the end of 2025.
That implies an increase of more than $50 million in roughly seven months. The percentage growth is therefore faster than Sierra's latest disclosed pace, although Sierra remains substantially larger.
Decagon's progress also looks broader than one giant contract. The company has been expanding geographically, including a London operation that grew beyond 50 employees across sales, engineering and customer-facing work. Its customer roster spans travel, telecom, financial services, commerce and consumer technology.
For now, the distance between the two companies is still meaningful. Decagon looks like the startup most capable of closing it.
| Company | Earlier revenue reference | Latest useful revenue reference | What we can reasonably conclude |
|---|---|---|---|
| Sierra | ~$26M ARR estimated at end-2024 | $200M ARR disclosed in nine quarters | Largest absolute growth among focused startups |
| Decagon | ~$44M at end-2025 | ~$100M annualized revenue estimated | Closing quickly from a smaller base |
| Wonderful | ~$1M run rate | ~$70M run rate | Fastest percentage growth |
| Parloa | ~$24M estimated at end-2024 | >$50M ARR disclosed | Roughly doubled in a year |
This chart, included in our conversational AI market deck, breaks down Cognigy's playbook in conversational AI
Are Parloa and Ada growing fast enough to challenge Sierra?
Parloa and Ada are growing very quickly, but neither currently gives us enough evidence to put it ahead of Sierra or Decagon.
Parloa crossed $50 million in ARR after ending the previous year at roughly $24 million according to Sacra estimates. More interestingly, Parloa disclosed 150% net revenue retention. Existing customers, after expansions, contractions and churn, were therefore worth about 50% more than they had been a year earlier.
Ada reported a similarly strong pattern. The company said revenue grew by more than 100% year over year, agentic AI ARR increased 108%, and net revenue retention reached 146%. Ada also had more than 550 enterprise AI agents deployed globally.
The problem with Ada is simple: it does not publish the underlying ARR figure. Doubling a $15 million business and doubling a $100 million business are completely different achievements.
Parloa gives us a clearer denominator, and its account expansion is excellent. Yet moving from roughly the mid-$20 millions to more than $50 million still represents far less new revenue than Sierra added.
These two companies deserve to stay near the top of the watchlist. The available numbers do not put them at number one today.
Is Bland the fastest-growing voice AI customer service startup?
Bland is one of the clearest leaders in voice AI customer service right now, although the company does not disclose enough recent revenue data for us to rank it above Sierra or Decagon.
Bland said in its latest Series C announcement that it handles more than 3.5 million calls every week. Fortune reported the same figure and said the company processed more than 175 million AI phone calls during the previous year for more than 250 enterprise customers.
The calls are also unusually demanding. Bland says a typical complex conversation can last 30 to 45 minutes, especially in healthcare and financial services. That makes its usage harder to dismiss as short automated confirmations or appointment reminders.
Bland has now raised more than $100 million in under three years, including a recent $50 million Series C. It also builds its own voice models rather than relying entirely on third-party foundation models.
What we still lack is a recent revenue denominator. Huge call volumes prove real adoption, but they cannot tell us whether Bland is commercially expanding faster than a company already adding tens of millions of dollars in ARR.
If you want more recent data on this point, please see our latest conversational AI market report.
This chart, included in our conversational AI market deck, shows annual funding in conversational AI startups
Which AI customer service startup is signing big companies fastest?
Decagon appears to be adding large enterprise customers fastest, while Sierra has gone deeper into the very biggest companies.
Decagon added more than 100 global enterprise customers in one year. According to the company's funding announcement, those wins included Avis Budget Group, Block and Deutsche Telekom alongside customers in finance, travel, commerce and telecom.
The more interesting detail is where those deals came from. Decagon has said roughly one-third of the customers it studied had not previously used AI automation, while more than half were replacing older IVR, ticketing or CRM automation. A smaller group chose Decagon instead of building the technology internally.
That tells us AI customer service is increasingly taking budget from existing software and contact-center systems rather than surviving as a separate experimentation budget.
Sierra has fewer comparable customer-addition figures, but its penetration at the top end is unusually strong. The company says it now serves more than 40% of the Fortune 50, after beginning with only four design partners.
Decagon currently looks stronger at rapidly accumulating enterprise logos. Sierra has already reached a level of Fortune 50 penetration that none of the other focused startups publicly matches.
Is Sierra's AI customer service growth coming from real production use?
Sierra's latest usage data strongly suggests that its revenue comes from production deployments rather than a pile of expensive AI pilots.
Sierra says agents built on its platform now power billions of customer interactions. Those interactions include refinancing homes, processing insurance claims, handling returns and completing other tasks that require access to company systems.
Its customer list includes ADT, Cigna, DIRECTV, Discord, Ramp, Rivian, SoFi, Sonos and Wayfair. The company also said one major healthcare organization went live seven weeks after project kickoff.
The type of work is important here. Answering a basic FAQ is easy to test and easy to cancel. Processing a claim or changing a financial product requires integrations, permissions and confidence that the agent will behave correctly.
Sierra has also expanded its product lately toward longer-running work. Its Horizon agents can manage inbound and outbound interactions over days or weeks, while the company's outcomes-based model charges customers for results rather than simply tokens consumed.
The combination of billions of interactions and deeper transactional work gives Sierra's revenue trajectory more substance than revenue generated mainly through short-lived proofs of concept.
This chart, included in our conversational AI market deck, compares the main business model options for conversational AI enterprise platforms
Are AI customer service customers actually spending more after they deploy these products?
Yes, the strongest AI customer service companies are increasingly expanding inside existing accounts, and Parloa and Ada give us the clearest numbers.
Parloa's 150% net revenue retention means an existing cohort becomes roughly 1.5 times as valuable over a year after accounting for churn and contractions. Ada's 146% figure tells almost the same story.
Those rates would be exceptional for many mature enterprise software businesses. Here they suggest companies start with one agent, channel or workflow and then give AI more of the customer-service workload once the deployment works.
Sierra does not publish directly comparable net retention. Its revenue progression points in the same direction, but we already use those figures to establish Sierra's overall growth, so they work better here as supporting evidence than as a separate new proof.
This account expansion may eventually matter more than customer counts. Winning one bank and spreading across dozens of workflows can create a much bigger business than signing many companies for narrow chatbot deployments.
Are investors already treating Sierra as the winner?
Investors currently value Sierra far above the other focused AI customer service startups, although the gap says more about future expectations than today's revenue.
Sierra's latest financing valued the company above $15 billion. Decagon was valued at $4.5 billion after its valuation tripled in under six months. Parloa reached $3 billion. Wonderful has now jumped from a reported $2 billion valuation to $5 billion after raising another $550 million.
Wonderful's repricing is particularly striking. The company's valuation increased by roughly $3 billion in around half a year as its revenue run rate surged and its geographic footprint expanded.
Sierra remains in a different tier. Its valuation is about three times Wonderful's and more than three times Decagon's.
Those numbers should not decide our ranking. Private-market valuations depend heavily on expected future market share, fundraising conditions and investor competition. They do show that Sierra's unusually fast revenue growth is being treated as more than an isolated early spike.
| Company | Latest reported valuation | What investors are pricing in |
|---|---|---|
| Sierra | >$15B | Category-scale enterprise AI platform |
| Wonderful | ~$5B | Continued hypergrowth and broader enterprise AI expansion |
| Decagon | ~$4.5B | Rapid enterprise customer-service share gains |
| Parloa | ~$3B | Strong expansion in enterprise customer experience |
If you want more recent data on this point, please see our latest conversational AI market report.
This chart, featured in our conversational AI market deck, illustrates revenue distribution by customer segment in the conversational AI market
How much can we trust the revenue numbers from these private AI startups?
We can be confident about the broad ranking, while exact comparisons between some companies still require caution because the quality of disclosure varies sharply.
Sierra itself has disclosed the $100 million, $150 million-plus and now $200 million ARR milestones. Parloa publicly announced more than $50 million ARR. Ada gives percentage growth and retention but withholds the underlying revenue base.
Decagon's latest roughly $100 million annualized revenue figure comes from Sacra rather than a company financial statement. Wonderful's roughly $70 million run rate appears in both Sacra's current estimate and recent Wall Street Journal reporting.
Those differences affect how precisely we should write the rankings. We cannot treat a third-party annualized estimate as though it were an audited annual figure.
The gaps, however, are large enough that small measurement errors would not change the main result. Wonderful is clearly multiplying revenue faster than the rest. Sierra clearly operates at much greater scale than Parloa. Decagon has clearly moved into the top group.
Trying to rank these companies to the nearest million dollars would create fake precision. The broader order is much harder to dispute.
So which AI customer service startup is growing the fastest today?
Sierra is currently the fastest-growing AI customer service startup overall, while Wonderful wins the narrower contest for the fastest percentage revenue growth.
Wonderful has produced the most spectacular number in the group: roughly $1 million to around $70 million of annualized revenue in about a year. On pure multiplication, nobody else is close.
Its widening focus complicates the comparison, though. Wonderful now sells a broader enterprise AI proposition, uses a labor-heavy forward-deployment model and reportedly operates at gross margins around 52%. Some of its newest growth therefore extends beyond what we would normally call a pure AI customer-service software business.
Decagon is the cleaner threat to Sierra. Its estimated annualized revenue has moved to roughly $100 million after finishing 2025 around $44 million, and it added more than 100 enterprise customers in that year. If Decagon keeps compounding anywhere near that pace, today's gap could close surprisingly quickly.
Sierra still has the best combination of evidence. The company reached nine-figure ARR seven quarters after launch, has since disclosed $200 million ARR in nine quarters, penetrated more than 40% of the Fortune 50 and now powers billions of customer interactions.
Our answer is therefore Sierra. Wonderful is growing faster mathematically, but Sierra is adding an extraordinary amount of revenue while already operating at a scale its direct startup competitors have not yet reached. Decagon is the company most likely to make that answer uncomfortable next.
If you want more recent data on this point, please see our latest conversational AI market report.
This chart, included in our conversational AI market deck, shows how AI chatbot platform technology has evolved over time
OUR METHODOLOGY
There is no single clean metric for deciding which AI customer service startup is growing fastest. Private companies disclose different numbers at different moments, so we compared recent revenue trajectories, the scale from which that growth was happening, enterprise customer momentum, expansion inside existing accounts, production usage and, more selectively, valuation.
Revenue trajectory receives the most weight. We look at both percentage growth and absolute revenue added so that an exceptional jump from a tiny base does not automatically outrank a company adding much more recurring business at scale. Customer expansion and production usage are used to test whether that growth reflects durable deployments rather than pilots.
We prioritized first-hand company disclosures whenever they were available. Where a company did not publish a comparable figure, we used high-quality external reporting or clearly labeled estimates, and kept ARR, annualized revenue, run rate, retention, usage and valuation in separate categories instead of treating them as interchangeable.
We did not turn the evidence into an artificial numerical score. The disclosures are not standardized enough to support that level of precision. Instead, we compared the strongest recent signals across the different dimensions and gave more weight to evidence that directly demonstrated commercial growth.
Sierra's latest first-party disclosure is especially important for freshness: the company has now reported $200 million ARR in nine quarters, which supersedes the older $150 million-plus milestone as its newest revenue reference. Its August 31 update also reiterates that Sierra works with more than 40% of the Fortune 50.
Key Sierra sources include Sierra's $100M ARR milestone, its year-two review, the newer $200M ARR disclosure, its scale, Fortune 50 and customer-interaction update, and its Horizon product update.
For Decagon, we relied on the company's Series D announcement, its Deutsche Telekom deployment, and its London and EMEA expansion update. For Parloa and Ada, the key sources were Parloa's $50M-plus ARR and 150% NRR disclosure, Parloa's $3B valuation announcement, and Ada's revenue growth, NRR and deployment update.
For voice AI and Wonderful, key sources include Bland's Series C announcement, Fortune's reporting on Bland's call volume and enterprise customers, Wonderful's Series B announcement, its Series C and broader enterprise-AI strategy, its forward-deployed engineering model, and The Wall Street Journal's reporting on Wonderful's revenue run rate, gross margin and international footprint.
In our conversational AI market deck, we identify pain points entrepreneurs should prioritize
Related blog posts
- Who are the top conversational AI startups by revenue today?
