The “Managed Crowd” Sourcing Model – Another “Made in China” Idea That Works

Innovative sourcing platforms that work While most suppliers would like to mainly attract and deliver high-end, high-margin services, many have to take on low-end, high volume tasks. Among such tasks are testing, transcribing audio files or hand-written documents and even some low-end application development.

Faced with such requests from large clients like Microsoft, suppliers are unlikely to reject the business, but at the same time are not keen on performing high volume, low margin outsourcing projects in-house.

So, what do they do? Some would perform such tasks in-house, doing their best to keep important clients happy at the risk of not meeting margin targets. Others will outsource these tasks to small or specialized suppliers, bearing additional governance costs of managing sub-contractors, again, just to keep their client happy.

Interestingly, Chinese suppliers have discovered another solution – the “managed crowd”. Many of us are familiar with the concept of crowdsourcing where a task is outsourced to an undefined crowd. The crowdsourcing platforms, such as Freelancer, facilitate the space where buyers and undefined suppliers meet, offering the end-to-end lifecycle for a small-scale outsourcing project.

I visited Pactera, the largest Chinese supplier, based in Beijing, to learn what Managed Crowd is.

Pactera’s Managed Crowd platform, called Pactwork, brings together Pactera’s clients and suppliers. The work performed through Pactwork is of high volume and involves mainly linguistic skills, such as the transcribing of audio files into dozens of languages. So, instead of performing the work in-house or subcontracting to another supplier, Pactera manages a crowd of 30,000 suppliers that does this work for them.

How does this work? Suppose a work-package of 1 hour of audio files arrives from a client. Pactera will create a new job on their crowdsourcing platform and invite suppliers to take part in the processing of the work. The invitation will be advertised in various social media sites as well as universities, to ensure good response from multiple sources. At this point the “management” of the crowd starts.

People responding to the invitation will be invited to go through on-line training that will prepare them to perform the work at the highest quality possible. The on-line training will show the crowd the unique tags and symbols they need to use during the transcribing of the audio files and will challenge them through a set of multiple-choice questions to validate their knowledge. On average, there are 6 sections of training and each individual from the crowd can easily spend a few hours training themselves for the job.

Once this stage has been completed, Pactera will invite that person to take a thorough exam. On average, 25% of those who express interest and take the training fail. They are entitled to try one more time before being rejected from the crowdsourcing platform. Those who pass will be reviewed by Pactera and only after a review of their abilities and score, will they get the approval to start performing work through Pactwork.

The audio files are chopped into 8 minute episodes and each episode is allocated to a Managed Crowd supplier. Once the supplier has completed one assignment, another is assigned to him/her. To ensure quality before delivering to the client, Pactera performs quality assurance on the delivered work as well as through an automated QA process of some audio files that were included in the assignment specifically for the sake of quality assurance.

Other suppliers in China have also introduced similar crowdsourcing platforms with a similar business model. iSoftstone runs a ToneLink managed crowdsourcing platform, where testing, SaaS, ERP and software development services are provided. Tonelink, based in Beijing with 1,000 employees and 3,000 suppliers registered, reported that B1.4RMB (about £160M) transacted through their managed crowdsourcing platform since its inception, only 18 months ago.

Tonelink revenues from these transactions are M180RMB (about £20M). Tonelink’s main driver to pursue this crowdsourcing model is based on their ability to control and deliver quality from sub-contracted work and yet benefit from attractive margins while maintaining a slim governance structure. Pactera cited similar reasons for the pursuit of this sourcing model, though Pactera’s scope of work is limited and currently focused on the transcription of audio files with the intention to expand to other business processes.

Tonelink is replicating their business model outside Beijing and has opened 6 other centres across China, that used the same crowdsourcing platform, attracting local business.

Interestingly, a managed crowdsourcing model that is based on an e-platform is not common with suppliers based in the West. One might see this as a missed opportunity, considering that many large suppliers have developed advanced e-platforms, though they have not yet realized their potential.

The Chinese did it again.

[inlinead]

The “Managed Crowd” Sourcing Model – Another “Made in China” Idea That Works

Upcoming Events

HR Shared Services & Outsourcing Week

E-invoicing in GBS - A Roadmap to Efficiency, Accuracy & Transformation

Clean Core, Clean AP: Modernizing Invoice Processing for S/4HANA

Evolving AP into a Spend Management Engine: How Sourcing, Contracts & Procurement Drive Real Value

Autonomous Finance Virtual Summit

Onboarding AI Employees: How Enterprises Are Building Their Next-Gen Digital Workforce

Follow Us

Latest Webinars

Onboarding AI Employees: How Enterprises Are Building Their Next-Gen Digital Workforce

Evolving AP into a Spend Management Engine: How Sourcing, Contracts & Procurement Drive Real Value

Clean Core, Clean AP: Modernizing Invoice Processing for S/4HANA

Recommended

Finance Transformation Imperatives: A Roadmap for 2026

Breaking the KPI Ceiling: How AI is Redefining AP Productivity

SSON's Essentials Guide To: Supplier-First Automation

A Guide to Process Governance That Actually Works

The ERP Acceleration Paradox: 5 Strategies to Beat the 75% Failure Rate

The Future of Order-to-Cash Market Report 2025

70% of Finance Transformations Fail: Here’s the Blueprint That Works

AI in Finance: The Future of Efficiency, Accuracy & Growth

The Urgency of Autonomous Finance for CFOs Today

Introducing Generative & Agentic AI into the O2C Process

10 Steps to Ensure Your Finance Transformation is the One That Succeeds

How does a Finance Center of Excellence Work?

FIND CONTENT BY TYPE

SSON COMMUNITY

ADVERTISE WITH US

JOIN THE SSON COMMUNITY

The “Managed Crowd” Sourcing Model – Another “Made in China” Idea That Works

Upcoming Events

Follow Us

Subscribe to our Free Newsletter

Latest Webinars

Recommended

FIND CONTENT BY TYPE

SSON COMMUNITY

ADVERTISE WITH US

JOIN THE SSON COMMUNITY