Okara
Start Free
Back to Blog
By · Published October 17, 2026 · 18 min read

Amplitude Revenue: How a Failed Voice App Grew to $410M ARR

Amplitude reached $410M ARR in Q2 2026 and $343.2M revenue in 2025. See how a failed voice app became a public analytics company, with timeline and numbers.

Amplitude reported $410 million in annual recurring revenue (ARR) at the end of June 2026, up 22% from a year earlier. In the same quarter it booked $100.9 million in revenue, and for full-year 2025 it reported $343.2 million.

The company did not start as an analytics business. In 2012, its founders went through Y Combinator pitching Sonalight, an Android app that let people send text messages by voice. The app went nowhere. The tool they built to watch how people used it became Amplitude.

What followed was a slow start, a fast middle, and a hard stretch after going public. Amplitude pitched more than 70 investors for its first round and took two years to close it. Then it grew revenue from $68.4 million in 2019 to $238.1 million in 2022, listed on Nasdaq at a fully diluted value of about $7.1 billion, and watched growth slow to about 8% in 2024.

In 2025 and 2026 growth picked up again, helped by AI products, bigger enterprise contracts, and an unusual deal to take over the customers of Statsig, an experimentation company OpenAI had bought.

Amplitude by the numbers

MetricFigureDateSource
Annual recurring revenue$410 millionJune 30, 2026Amplitude Q2 2026 results
Quarterly revenue$100.9 millionQ2 2026Amplitude Q2 2026 results
Annual revenue$343.2 millionFiscal 2025Amplitude FY2025 results
2026 revenue guidance$407.2 million to $411.2 millionAugust 5, 2026Amplitude Q2 2026 results
Paying customersMore than 5,200June 30, 2026Amplitude Q2 2026 10-Q
Customers above $100,000 ARR824June 30, 2026Amplitude Q2 2026 results
Customers above $1 million ARR56December 31, 2025Amplitude FY2025 results
Net revenue retention (trailing 12 months)105%June 30, 2026Amplitude Q2 2026 results
Employees780December 31, 2025Amplitude 2025 10-K
Last private roundSeries F, $200 million at $32.02 a shareMay to August 2021Amplitude S-1
Value at Nasdaq debutAbout $7.1 billion, fully dilutedSeptember 28, 2021CNBC
Market valueAbout $1.66 billion (our calculation)September 25, 2026Yahoo Finance, Amplitude 10-Q

The market value is our own estimate: the $13.30 closing price on September 25, 2026, multiplied by the 124.9 million Class A and Class B shares Amplitude reported outstanding on August 3, 2026.

Amplitude's growth timeline

YearWhat happened
2011Company incorporated in Delaware as Sonalight, Inc.
2012Spenser Skates and Curtis Liu join Y Combinator (Winter 2012) with a voice-texting app
2014Founders go through YC again; Amplitude launches out of stealth as a real-time mobile analytics product
2015$9 million Series A led by Benchmark
2016 or 2017Series B led by Battery Ventures (see note below)
2018Series D led by Sequoia ($79.8 million gross); Startup Scholarship program launches
2019Revenue of $68.4 million
2020Revenue of $102.5 million; $49.8 million Series E
2021$200 million Series F; direct listing on Nasdaq on September 28; revenue of $167.3 million
2022Revenue of $238.1 million; ARR of $255 million
2023Revenue of $276.3 million; ARR of $281 million
2024Revenue of $299.3 million; ARR of $312 million; buys Command AI
2025Revenue of $343.2 million; ARR of $366 million; buys Kraftful and two AI teams
May 2026Takes over Statsig's brand and customer contracts
June 2026ARR reaches $410 million

A note on the Series B: Amplitude's own company timeline lists a $15 million Series B in 2016. Business Insider reported it as $15.9 million in 2017. We have shown both.

The analytics tool that was a side project

Sonalight was a Siri-style app for Android. CNBC described the 2012 YC pitch as an app "that would let users send text messages by voice."

To see how people used the app, the team built its own tracking software. According to CNBC, other startups became interested in that software. The founders went back through Y Combinator in 2014, this time with Amplitude, and came out with a check from Benchmark's Eric Vishria.

The Slack comparison is obvious (a game company whose internal chat tool became the real product), and CNBC made it. The lesson is narrower than "pivot when you find something." Amplitude's founders were their own first customers. They knew what a product team needed from behavioral data because they had been that team, watching a consumer app fail.

Amplitude's 2025 annual report still frames the origin this way: the founders "had seen first hand how important it was to make data-driven decisions in product development," and how hard that was to do well.

Seventy pitches and a better deck

Fundraising was slow. Skates told Business Insider he pitched more than 70 firms for a seed round starting in 2012. It took two years to raise $2 million from SV Angel, Quest Venture Partners, and Data Collective.

His own diagnosis was blunt: "I didn't do a good job talking about customers and the pain they had."

The Series A deck changed that. Skates said it leaned on hard data, numbers, and customer feedback. That deck raised a $9 million Series A led by Benchmark in 2015.

This matters because the same thing happened to the product. Early Amplitude was a tool for analysts. Its later growth came from telling a business story: this data tells you which product changes make money. By the 2021 S-1, the pitch was that Amplitude "transforms what has traditionally been viewed as a cost center into a revenue center."

A free tier that did the marketing

Amplitude has always had a free plan, and it has been open about what that plan is for.

In the 2021 S-1, Amplitude described its free Starter plan as a self-serve option that included core analytics and up to 10 million events a month, with unlimited user seats. It said the plan "serves as effective brand marketing and a low friction model for getting a single team started."

Two details in that description explain the design.

The first is unlimited seats. Amplitude wrote that Starter users "are encouraged to add additional team members across functions." Every extra person who logged in was another potential champion inside the company.

The second is the upgrade trigger. When a team outgrew the event limit or wanted predictive analytics and behavioral reports, it had to move to the paid Growth plan.

Amplitude added a second free program for startups. The Startup Scholarship, which the company dates to 2018, gave young companies a free year of the Growth plan. The current version, on Amplitude's pricing page, offers one free year of Growth to startups with under $10 million in funding and fewer than 20 employees.

The free plan has shrunk since 2021. Amplitude's pricing page now offers 2 million events a month on the Free plan, and says every plan includes unlimited seats.

The self-serve side also got a middle tier. Amplitude's 2024 annual report describes a self-serve Plus plan "built for growing startups and small teams." Paying customers rose from 1,994 at the end of 2022 to 2,723 in 2023 and 3,875 in 2024. The number of customers paying over $100,000 grew much more slowly in those years, from 480 to 511 and then 591. The likely explanation is that a large share of new paying customers were small self-serve accounts.

Land with one team, expand across the company

The free tier brought teams in. Sales turned teams into accounts.

Amplitude calls it land-and-expand. In its annual reports, the company says customers "typically start out using Amplitude for a specific business use case" and expand to more teams, more products, and more features.

The S-1 showed how well this worked before the listing. Dollar-based net retention across paying customers was 116% at the end of 2019 and 119% at the end of 2020. In plain terms, the average paying customer spent 19% more in 2020 than a year earlier, after accounting for customers who left or cut back.

The count of large accounts climbed at the same time. Amplitude had 208 customers paying more than $100,000 a year at the end of 2019, 262 at the end of 2020, and 311 by June 2021. Customers above $1 million went from 11 to 22 over the same stretch.

The go-to-market mix reflected that. The S-1 described "direct sales, including field and inside sales, product-led growth initiatives" and marketing. In 2019 Amplitude hired Matt Heinz, a sales leader from Marketo, as chief revenue officer. The customer list in the S-1 read like a product-led growth hall of fame: Shopify, Instacart, Peloton. Amplitude described these as companies that are "among the 1% of companies who lead with a product-led growth" model.

Naming the category, then renaming it

Amplitude has changed how it describes itself several times, and each change tracked a bigger buyer.

In the 2021 S-1, Amplitude claimed to be "pioneering a new category of software called Digital Optimization," and called its product the "Digital Optimization System." The pitch was aimed at executives: connect digital products to business outcomes.

By its 2022 and 2023 earnings releases, Amplitude called itself "a leading digital analytics platform." In 2025 and 2026 it switched again, to "the leading AI analytics platform," and its CEO started talking about "self-improving products."

The underlying product did not change as much as the labels. Each rename widened the audience from product analysts to marketers, executives, and now AI teams. It also let Amplitude sell more products under one umbrella: Experiment, Session Replay, Guides and Surveys, and AI agents.

Amplitude also taught its method, not only its software. It publishes a "North Star Playbook," a guide to picking the single metric that best reflects the value a product delivers. Content like this puts Amplitude's framing in front of product managers long before they shop for a tool.

Where category language gets decided now

The words buyers use for a category used to be decided by analysts and search results. More and more, they come from AI assistants. When a product manager asks ChatGPT for "the best product analytics tool," the answer decides who makes the shortlist. Okara's GEO Agent tracks up to 15 of those buyer prompts in ChatGPT, Google AI Overviews and Google AI Mode and shows which competitors get cited instead of you.

Going public without an IPO

Amplitude went public through a direct listing on September 28, 2021, instead of a traditional IPO.

It did not need the cash. Earlier that year it had raised a $200 million Series F in several closings, at $32.02 a share. Bloomberg reported the round as $150 million at a $4 billion valuation in June 2021; the S-1 shows later closings brought the total to $200 million.

Skates had strong views on IPOs. "IPOs traditionally underprice companies, and not by a little bit, by a lot," he told CNBC. Speaking as a fiduciary to existing shareholders, he added: "it's just totally unacceptable to give them a bad deal."

The shares opened at $50 and closed the first day at $54.80. CNBC put the fully diluted market value at about $7.1 billion. Reuters' headline valued the company at $5 billion. The gap likely comes from counting outstanding shares versus the fully diluted share count, which includes options and restricted stock. We have shown both.

Benchmark, Amplitude's largest investor, owned 15% of the company, a stake CNBC valued at more than $835 million at the first-day close. Not bad for a firm that backed a second trip through Y Combinator after a failed voice app.

The hard years after the listing

The timing was close to the top of the software market.

Revenue growth went from 63% in 2021 to 42% in 2022, then 16% in 2023 and about 8% in 2024, when revenue reached $299.3 million. ARR grew 10% in 2023 and 11% in 2024.

Retention showed the strain. Net revenue retention across paying customers, which had been 119% in 2020, was 97% on a trailing twelve-month basis at the end of 2024. That meant existing customers, as a group, were spending slightly less than a year earlier.

The stock followed. From a first-day close of $54.80, it traded at $13.30 on September 25, 2026.

Amplitude responded by tightening costs and pushing further upmarket. Its 2025 10-K says: "As we continue to focus on, and may become more dependent on, sales to larger organizations, we expect our sales cycle to lengthen." By the end of 2025, the company said enterprise and multi-product customers made up 74% of total ARR. Free cash flow reached $23.5 million for 2025, a record for the company.

Buying the platform piece by piece

Much of Amplitude's product growth after 2020 came through acquisitions:

  • ClearBrain, a predictive analytics startup, in 2020
  • Iteratively, a data pipeline company, in May 2021
  • Command AI, an in-app user assistance startup, in October 2024. TechCrunch reported the price at "north of $45 million."
  • Inari, an applied AI team, in June 2025
  • June.io and Kraftful in July 2025; Kraftful brought voice-of-customer technology
  • InfiniGrow, a marketing analytics company, in January 2026

The pattern was to buy a team, then ship a product quickly. Amplitude's 10-K says it launched Amplitude Guides and Surveys in February 2025, "just four months after our acquisition of Command AI."

Each new product gives the sales team another line item to add to an existing account. That ties back to land-and-expand: it is easier to sell a second product to a team already sending you data than to find a new customer.

The Statsig deal

The biggest move of 2026 was also the strangest.

Statsig was a fast-growing experimentation and feature-flag company. Amplitude's 2025 10-K listed it as a competitor. In 2025, OpenAI bought Statsig, and MarTech reported the price at $1.1 billion.

On May 1, 2026, Amplitude signed an agreement to take over Statsig's customer contracts, order backlog, trade name, and a perpetual license to its technology. The Statsig team stayed at OpenAI. Amplitude said it "will maintain and develop the current Statsig platform across the cloud and data warehouse."

The 10-Q shows the deal was booked as an asset acquisition. Amplitude recognized $8.7 million of intangible assets and took on $17.9 million of deferred revenue, meaning prepaid customer contracts it now has to deliver.

The numbers moved fast. In Q2 2026, Statsig customers contributed about $6.5 million in revenue. CFO Andrew Casey said Statsig added $17 million in ARR in the quarter and core Amplitude added $19 million.

Not everyone was convinced. MarTech quoted an industry critic who called it "a race car without a driver," because Amplitude got the code but not the people who wrote it. The same article noted that Amplitude now owned two overlapping experimentation products.

For a growth story, the deal shows something simple. Buying a book of customers with signed contracts is one of the fastest ways to add ARR, if you can keep them.

Growth is picking up again

After the 2024 low, Amplitude's numbers turned.

ARR growth went from 11% at the end of 2024 to 17% at the end of 2025 and 22% at the end of June 2026. Take out the $17 million Statsig added and ARR would have been about $393 million, roughly 17% above the $335 million of a year earlier (our calculation).

Retention recovered too. Trailing net revenue retention rose from 97% at the end of 2024 to 105% by mid-2026. Customers paying over $100,000 grew 30% year over year to 824.

Amplitude now expects 2026 revenue of $407.2 million to $411.2 million. It is also spending cash on buybacks: its cash flow statement shows $89.5 million of stock repurchases in the first half of 2026.

AI features are central to the pitch. Amplitude's recent releases list a Global Agent that answers questions about customer behavior, an MCP server that brings Amplitude data into tools like Claude, Cursor, Slack, and Figma, and an Agent Analytics product for measuring AI agents. How much ARR each one adds is not disclosed.

What founders can take from Amplitude

The first lesson is about the first customer. Amplitude's founders built analytics for their own failing app, so they understood the need firsthand. Many good B2B products start as internal tools; Amplitude's early pitch problems came from describing the software rather than the pain.

The second is that a free plan can be a marketing budget, if you design it that way. Unlimited seats meant a single free account could spread through a company. The event cap meant success forced an upgrade. Amplitude said in writing that the free plan was brand marketing.

The third is about pricing power and retention. Amplitude's best years had net retention near 119%. When retention dropped below 100% in 2024, growth nearly stalled, even as the paying customer count rose 42%. Adding many small customers does not make up for existing ones spending less.

The fourth is that positioning follows the buyer you want next. "Product analytics" sold to analysts. "Digital optimization" sold to executives. "AI analytics" sells to teams building with AI. Each rename came with new products to justify it.

The last is less comfortable. Going public at the peak gave Amplitude cash and credibility, and a stock that later fell about 76% from its first-day close. Growth and retention matter more to a public company's value than the story told on listing day.

More startup growth stories

Frequently asked questions

What is Amplitude's revenue?

Amplitude reported revenue of $343.2 million for fiscal 2025, up 15% from $299.3 million in 2024. In the second quarter of 2026, revenue was $100.9 million, up 21% year over year. In August 2026, the company guided to full-year 2026 revenue of $407.2 million to $411.2 million.

What is Amplitude's ARR?

Amplitude's annual recurring revenue was $410 million as of June 30, 2026, up 22% from $335 million a year earlier. About $17 million of the increase in that quarter came from former Statsig customers, according to Amplitude's CFO.

What is Amplitude's market cap?

Using the $13.30 closing price on September 25, 2026 and the 124.9 million shares Amplitude reported outstanding on August 3, 2026, Amplitude's market value was about $1.66 billion. That is our calculation. At its Nasdaq debut on September 28, 2021, CNBC put its fully diluted value at about $7.1 billion.

How many customers does Amplitude have?

Amplitude had more than 5,200 paying customers as of June 30, 2026, according to its Q2 2026 10-Q. It had 4,797 paying customers at the end of 2025, including 27 of the Fortune 100. As of June 30, 2026, 824 customers paid more than $100,000 a year.

Who founded Amplitude and when?

Amplitude was founded by Spenser Skates (CEO) and Curtis Liu (CTO); CNBC also names Jeffrey Wang as a co-founder. The company was incorporated in November 2011 as Sonalight, went through Y Combinator in Winter 2012, and started the Amplitude business in 2012. It was renamed Amplitude, Inc. in December 2014.

How much funding did Amplitude raise before going public?

Amplitude's disclosed rounds include a $2 million seed, a $9 million Series A led by Benchmark in 2015, a Series B led by Battery Ventures, a $79.8 million Series D led by Sequoia starting in November 2018, a $49.8 million Series E in April 2020, and a $200 million Series F in 2021. Bloomberg reported a $4 billion valuation for the Series F in June 2021.

How did Amplitude go public?

Amplitude went public through a direct listing on Nasdaq on September 28, 2021, under the ticker AMPL, rather than a traditional IPO. Shares opened at $50 and closed the first day at $54.80, giving it a fully diluted value of about $7.1 billion according to CNBC.

Why did Amplitude buy Statsig?

Amplitude did not buy the whole company. On May 1, 2026, it acquired Statsig's customer contracts, backlog, trade name, and a perpetual license to its technology, while the Statsig team stayed at OpenAI. Amplitude said the deal expands its experimentation and feature management products. Statsig customers added $17 million in ARR in Q2 2026.

Sources

Tags

Growth stories