Of course, the Sourcing Interests Group and its members have traditionally
been concerned with a broad range of sourcing issues. So what has
Crowdsourcing got to do with “sourcing”?
In its earliest manifestation crowdsourcing was a cuckoo in the nest.
You might remember crowdsourcing as the ideas platform your company championed
to generate a new innovation pipeline, or a design contest like Peugeot’s
Concept Car challenge or big events like the X-Prize competitions.
To begin with, crowdsourcing was a cheap way to get plentiful ideas. But it
has now moved well beyond that. Under the radar, in fact, it has moved right
into the middle of mainstream sourcing.
How come? Crowdsourcing is now a fully-fledged “platform” business. Just like
Facebook aggregates millions of people to interact socially, platforms such as
Passbrains aggregates tens of
thousands, and even hundreds of thousands, of workers for specific enterprise
tasks. And crowdsourcing as a platform is being implemented by large, small and
Massolution, a specialist crowdsourcing
research and advisory firm,
reports that the enterprise crowdsourcing market is growing in excess of 75%
year-on-year and the work-force is currently doubling each year with over 6
million crowdsourcing workers available for work in 2012.
In the large enterprise space, crowdsourcing is used by giants such as
Microsoft, to market test products before launch. There is also public-sector
applications – increasingly state tax returns in the USA are being checked by
members of large private crowds, as an alternative to mass temp hiring.
The advantage of the crowd in both these cases is obvious, when you think
about it. They represent tasks for which a large number of people are needed in
a very short time. They allow clients to scale up quickly and then to scale
down. That makes them true on-demand platforms where clients only pay for the
work they need and are not billed on an hourly or per diem contract.
The previous solution to “scaling-up” in these situations was to hire in
temporary staff and reallocate permanent staff. But that carries a number of
burdens, like shifting resources from other tasks that need completing, hiring
costs, training costs, etc..
In a crowd solution, in place of the hire and training, expert task
designers, redesign the tasks at hand so they can more easily be done with a
very light training load. That redesign also makes the tasks highly scalable.
Thousands of people on a crowd platform can simply self-select their
participation and see how they perform. Of course, it also introduces privacy
and confidentiality issues. Task designers, however, are also expert at breaking
tasks down in such a way that a client’s confidentiality and non-disclosure
concerns are fully addressed. Take an instance of high confidentiality in the
financial services sector or indeed taxation. No worker will ever receive the
full social security details or bank account of any one end-customer, nor
personal details that link to financial information.
We introduced earlier the idea of the “private crowd”. It’s an important
concept. Clients can always opt to use a wide open public crowd. But people who
work in crowdsourcing tend to work hard to optimize and continually develop
their own skill sets. They become adept at changing course and taking on new
skills, however small that skill set. That flexibility becomes an asset. Workers
are further motivated by their typical commitments to Non-Disclosure Agreements
and that their income may be conditional on their respect for privacy.
Private crowds are like any work force in that respect. The platforms that
hire them take on the responsibility of relationship building, like an employer
does. And the worker tends to perform well because that ensures continuity.
We also said that the crowd can be applied to small and medium sized
enterprises. An area that the crowd platform Passbrains is highly active in is
market and product testing and these are perfect applications for all sizes of
enterprise from small to large.
Traditionally smaller companies have not been resourced to do extensive
market testing. But we are living in the age of lean entrepreneurship. Customers
have very limited attention and it is imperative for any company to test its
products, to iron out any interface or usability issues, and see what levels of
passion it can generate before they launch. Large, Fortune 100
enterprises, such as eBay, have become heavy users of exactly this type of
The crowd, then, is not just about scalable “muscle”. It can also be used for
very high value tasks.
Smaller companies can bring customers, vendors and employees together on the
same platform to create a more collaborative atmosphere that will help them
recognize and exploit new opportunities. Passbrains’ Enterprise Testforce
Management solution is used by multiple SMBs in software testing area, exactly
for this purpose. The Enterprise platform allows a Testing organization to
develop its private crowd comprising employees, vendors and other groups. One of
its key features, called “Knowledge Centre”, targets problem solving and micro
consulting through the crowd. It goes one step further and facilitates knowledge
networking, with an inbuilt intelligence to identify and connect with top
experts in the crowd, associated to a given topic or problem.
The evidence shows that this kind of collective can accelerate complex
problem solving by around 30% and a recent study conducted by McKinsey that such
real-time knowledge-networking, possible on platforms can increase
organizational productivity by 20-25%.
So what kind of tasks are now crowdsourced? We already mentioned testing, and
that can also involve product development and interface feedback or expert
networking and micro consulting.
But the roster of crowdsourced tasks is much bigger. Document processing on a
mass scale, such as tax returns is one such task. Others include data
management. In the area of data management data quality is often a huge issue
because the cost of keeping data up to date and accurate often exceeds the
value. That’s becoming a much bigger issue with the advent of Big Data. Crowd
workforces are being used to clean data so that it is more usable and to free
analyst time for high value data interpretation tasks.
In social media there is a problem with sentiment analysis – that is the
analysis of what the public is saying about a brand or company on blogs, Twitter
and Facebook. The accuracy of sentiment analysis is 80% at best but in reality
it falls far short of this. Crowds are being used to check automated sentiment
analysis and to course-correct for the client.
The crowd is even being used to further medical research, not just in
clinical trials but by using the crowd to identify particular types of experts
and to build expertise directories.
Crowds are being used for translation especially for online content where
high cost professional translation would not be justifiable on cost grounds. And
they are being used to content creation.
The reality is that crowd labor is allowing tasks to be fulfilled that
previously were unaffordable. In that respect, it does not replace outsourcing
and is often not a competitor to it. What it does is introduce new opportunities
for labor to find paid work and new opportunities for companies to improve their
products, their customer engagement and their viability. That’s what we used to
call a win-win. Perhaps we still do?