Techcrunch has an article on how uTest, a crowdsourcing test startup, just got another round of funding. Given that funding is something hard to get these days, the fact that the company got another round is quite interesting.

uTest offers something like Amazon Mechanical Turk, but specialised in testing. You specify some requirements and restrictions and let the “community” test your webapp.

The main problem I can see in the offering is how it is going to be faced by the market. As an example, see this misinformed piece at the first article published by Techcrunch on it:

The uTest offering makes a lot of sense to me and I expect it will be warmly embraced for several reasons. First, consider for a moment that for many SMBs, QA management solutions such as those offered by Mercury or IBM-Rational are beyond their reach. Speaking from personal experience, the majority of the testing cycles I’ve seen have been “managed” on Word or Excel. The “on-demand” model which has been proven time and time again on the Web from CRM to ERP, should work just as well for QA. The fact that the platform will be offered free of charge, pushes the offering to “no-brainer land”.

Second, beyond the manageability aspects of uTest’s service, companies will obviously enjoy the cost saving aspects of paying testers by the bug. It’s this aspect that in my opinion will make uTest’s offering relevant not only to SMBs but to large organizations as well. Logic would dictate that if they embraced off-shoring and near-shoring, crowd-sourced QA shouldn’t be too jagged of a pill to swallow.

Third, recruiting a userbase of testers should not be difficult. There are droves of potential testers in countries such as India, China, Russia, Bulgaria, Estonia, etc. Also, getting hired through sites like oDesk, Elance and RentACoder is becoming increasingly difficult due to the growing number of service providers. These same individuals can theoretically provide testing services instead of programming.

Free, inexpensive or open source tools offer most useful features present on expensive products, so there’s really nothing special about uTest in this aspect.

Paying per bug found is something common in some companies and often end up in some funny and inefficient environments.

Recruiting a good tester is difficult (just as recruiting a good developer or project manager) and uTest won’t help here. Outsourcing tests is something really hard to get right and outsourcing those to random folks around the world won’t make it any easier. Actually I think that using uTest for the full QA process will get results as bad as the worst waterfall implementation, it’s like going back to the classical waterfall approach.

What I see that uTest can help with is to test applications targeting the general public. In such projects there’s never enough time or resources to test it in all desirable configurations or geographic areas. Also as testers participate on the requirements gathering and acceptance criteria definition it is quite useful to have an outsider to look for odd behaviour.

Companies spend a lot of money in creating “test departments” that do this kind of job. It seems like uTest can do in a very cost-effective way.

Will it work? I have no idea. What I know is that it is definitely one more option to consider.