Under the wave of AI technology, the traditional manual customer service industry is facing great changes. AI customer service company Crescendo achieved $100 million in annual recurring revenue (ARR) in just 15 months and was profitable. How does it do it?
The manual customer service has been laid off, although it is cruel, but this is a fact.
Some consulting firms estimate that in the next five years, the Philippines, known as the “world’s outbound call center”, will lose 300,000 personal customer service jobs due to AI.
On the other side of customer service personnel losing their jobs, AI customer service companies are accelerating their growth.
Mike Ryan, senior vice president of AI customer service company Crescendo, recently revealed that Crescendo’s ARR exceeded $100 million and achieved profitability.
Even more surprising is that Crescendo achieved such results in just 15 months.
How does Crescendo do it? The answer isDirectly package AI customer service system + manual team, 90% of routine consultations are automated, and when encountering complex disputes, the manual team can quickly intervene with the help of AI-sorted dialogue context.
In terms of business model, Crescendo has also abandoned the traditional customer service billing model by number or time and adopted result-oriented pricing, which is very attractive to companies that want to embrace AI but lack experience.
With innovative AI customer solutions and a paid model for results pricing, Crescendo has achieved phenomenal growth.
01 15 months, achieve 100 million ARR
The traditional customer service system usually consists of two parts: chatbots and call centers. However, this approach not only leads to a disjointed customer experience, but also prone to an adversarial operating model: chatbots and call centers compete with each other, and businesses and their customers will bear the cost of this misalignment and inefficiency.
To achieve these three challenges, product managers will only continue to appreciate
Good product managers are very scarce, and product managers who understand users, business, and data are still in demand when they go out of the Internet. On the contrary, if you only do simple communication, inefficient execution, and shallow thinking, I am afraid that you will not be able to go through the torrent of the next 3-5 years.
View details >
Not to mention the uneven quality of customer service, customer service work is tiring and stressful, the employee turnover rate is as high as 30-35%, and the service quality is not guaranteed.
In this context, Crescendo, a unicorn in the AI customer service industry, stands out with its human-machine collaboration solution. It packages AI and human customer service solutions into a holistic solution, seamlessly integrating AI and human team service processes.
Specifically, through CX (customer experience) AI chatbots, Crescendo can handle 90% of routine inquiries and respond autonomously based on internal data (such as product catalogs and policy documents).
When encountering complex disputes, real experts can quickly intervene with the help of AI-combed conversation context, and can even retrieve the user’s return and exchange records and shopping history from three months ago to achieve accurate service.
The benefits of this strategy are, first of all:Directly saves millions of labor costs because customers don’t have to look for outsourcing。
All work is done by Crescendo in one place, and the platform supports automatic translation in 50 languages, custom speech and tone settings. The system can be deployed within 2~4 weeks, and 90%+ of regular queries will be automatically processed by Al after 1 month of launch, which greatly reduces the construction time and economic cost of traditional call centers.
Its more prominent advantages are:Be responsible for customer service effectiveness and improve the order rate。
In high-frequency, low-complexity customer service scenarios, AI will indeed perform better than humans. AI deals with 80% of high-frequency and simple problems, while manual customer service only needs to deal with the remaining 20% of low-frequency complex problems that AI cannot handle.
In addition, Crescendo’s quality assessment tool scores the efficiency of each interaction in real time, and if the score is low, the credit line is returned to the business.
For the company, there is no need to worry about the impact of the original personnel turnover and inconsistency on the quality of service.
Based on this business model, Crescendo has moved away from the traditional headcount or duration billing model and adopted results-driven pricing (one is charged per resolution rate and the other is based on AI throughput), which is attractive to businesses looking to embrace AI but lack experience.
Crescendo describes its profit margin as 4 times higher than that of traditional call centers.
In addition to novel solutions, team members’ extensive experience in customer service was key to securing funding. The core members of the Crescendo AI team can be called the “dream team” of the customer service industry:
- CEO Matt Price was the senior vice president of customer service giant Zendesk and is well versed in service system construction;
- CTO Slava Zhakov has the technology precipitation of cloud call center giant Genesys;
- CPO Anand Chandrasekaran is from Indian e-commerce giant Snapdeal and has rich experience in operating consumer scenarios.
Co-founder and chairman of the board of directors Andy Lee is an industry legend, and he founded AIorica, an outbound call center company he founded from scratch, with annual revenue of more than $2 billion.
This time, Andy Lee hopes to use generative AI to reconstruct the service and charging system of the call center industry for 20 years.
Mike Ryan, senior vice president of marketing at Crescendo, recently revealed that relying on the integrated solution of “AI customer service + expert team”, the company’s ARR exceeded $100 million and achieved profitability.
02 “AI + real person” reconstructs the cost of call centers
Crescendo has also been upgrading its business in recent months.
In October 2024, Crescendo acquired PartnerHero, an outsourcing service provider focused on brand customization, adding 200 clients and 3,000 CX professionals. In the same month, Crescendo received $50 million in venture capital led by General Catalyst, raising the company’s valuation to $500 million.
Simple technical output is no longer enough to meet the demand, customers need solutions that can directly produce business results. Crescendo’s shift to the “AI + expert” packaging model indicates that the maturity of AI technology, data accumulation, and scenario understanding have reached a critical point, and the integrated delivery of “system + service” is possible.
In the crowded AI customer service track, Crescendo is not the first customer service company to adopt the “AI + real person” model, Decagon, Sierra AI, and Shulex, three AI customer service companies have adopted the same strategy, but have also made their own characteristics.
(1) Decagon data transparency and build customer service agents based on usage scenarios
Founded in 2023, Decagon has grown rapidly since its inception. Its AI agents are based on state-of-the-art models from companies like OpenAI, Anthropic, and Cohere, and are trained on internal data such as how-to guides, manuals, and past customer service conversations. Employees score and review AI-generated responses to improve the system. Decagon also allows businesses to have a clear view of the decision-making process as well as data usage.
In February, Decagon partnered with AI audio generation leader ElevenLabs to create voice agents to engage in more natural, human-like conversations with customers.
Decagon has helped several businesses save on customer service outsourcing costs.
Credit card provider Bilt has reduced its customer support team from hundreds to 65 with Decagons. Fitness giant ClassPass used Decagon’s AI agents to conduct 2.5 million conversations with customers, reducing customer support costs by 95%.
To date, Decagon has signed contracts worth more than $10 million. The company is negotiating a new $100 million funding round, led by a16z and Accel, with a post-investment valuation of $1.5 billion.
(2) Sierra AI, not “help you wire”, is “directly solve the problem”
Sierra is positioned to create an automated customer support platform with human interaction. Its platform simulates the model of traditional outbound call centers, with a multi-model coordination system behind it, which makes it extremely stable and fault-tolerant in real customer service scenarios.
Sierra’s breakthrough in AI customer service is to upgrade from an “answering robot” to a “closed-loop interaction system”. In other words, it’s not “helping you with labor”, but making “you don’t need labor at all”. Things it can do include:
Continuous dialogue of emotion recognition + contextual memory Multi-channel (email, web page, chat window) unified customer service response is integrated with the enterprise back-end system, and closed-loop actions such as refunds, changes, and complaints are automatically pulled from the knowledge base and historical records, and dynamically optimized speech.
According to Sierra pilot data, a SaaS company has increased customer service response speed by 400%, customer satisfaction by 37%, and the size of the manual customer service team has been reduced by 60% after connecting to Sierra.
Sierra AI completed a $175 million Series B financing more than a year after its establishment, with a valuation of $4.5 billion and an ARR of more than $20 million.
(3) Shulex focuses on specific scenarios that customers are concerned about and customizes AI/manual services.
This domestic overseas company adopts the AI digital human model, which can switch between manual and machine customizations smoothly, and can deliver quantifiable work results to enterprises, while providing consumer insight reports.
Shulex’s AI Agent can reach the intermediate and advanced level of human customer service, and can quickly replicate during peak demand periods (such as Black Friday Network One), providing “100+ intermediate and advanced AI digital customer service on standby”. More importantly, Shulex provides not only technology, but also a “companion” service that includes consulting, training, implementation and even continuous optimization.
Speaking about the advantages of Shulex, founder Guo Chenlu said:
AI is not the core barrier, the most important thing is the experience of application scenarios, especially the very vertical one-stop scenario-based experience is more difficult to replace.
Therefore, Shulex focuses more on specific scenarios that customers pay attention to, such as logistics consulting for large furniture, knowledge self-learning capabilities of 3C digital, and small language services.
The company has just completed a new round of 100 million yuan financing, led by Shanda Capital. Its customer service AI Agent has served 100+ global leading brand customers in China, Japan, and the United States.
03 The logic of customer service has changed
Although the customer service market may seem crowded, early players have been slow to deploy generative AI in practice, and the quality is not high.
These solutions rely more on AI technology itself, automating outbound call center tasks by creating playbooks, and there are no products built from scratch based on generative models. This has brought opportunities to many AI customer service companies.
More importantly, AI is fundamentally changing the customer service business, that is, transforming from a cost center to a value center.
In the past, it was difficult for customer service to provide differentiated services, and general services were not directly linked to profits, so enterprises were doing everything possible to reduce customer service costs. But now it’s different, when customer service becomes highly customized, it has the potential to become an important part of product services and even a driver of revenue growth.
As Decagon CEO Jesse Zhang said:
In the past, businesses simply saw customer support or call centers as cost centers, which may not be at the core of your vision. But with proxies, they are very customizable, and it starts to become a competitive advantage for a business or product, becoming a revenue driver.
