The site is actually quite complex. There is, obviously, your profile. You can add your list of publications ('contributions'), a link to your departmental website, information on previous and current projects and positions, and a list of sometimes rather ill-defined skills that you consider yourself to have.
Where Facebook has 'friends', ResearchGate has 'followers'. You can 'follow' other scientists, the idea presumably being that you get updates on what they are publishing or otherwise doing, but I have to say that I am notified of little of any use (see below). As could be expected, you might feel some implicit pressure to 'follow' somebody back if they 'follow' you even if what they do does not interest you. On Facebook, one can simply ignore a friend request and pretend to have been too distracted to notice it (lalala) but there appears to be no way of keeping somebody from following you here.
The last major aspect of ResearchGate is the Q&A section where people can ask questions, hoping to get helpful advice on some scientific problem or methodological challenge.
The good aspects
ResearchGate is another venue for young and job hunting scientists to advertise themselves, similar to LinkedIn (which I do not have any experience with). A profile on the network generally comes up quite high on Google when searching for a scientist's name, and ResearchGate has a sleek and professional layout that may look quite attractive to those happening upon it from the outside and using it to read up on what you are doing.
Very comfortably, the website itself searches through journals and suggests additions to your publication list if it thinks that you are the author, saving you a lot of bother with keeping your profile updated. Well, I say that as somebody with a globally unique name but I do not know how well it works for the John Smith's of the world. Maybe they get lots of false positives all the time.
When updating the publication list, ResearchGate will often even add the abstract, a link to the article and, if available, a PDF. You can also add a full text copy of the article yourself, e.g. the last manuscript draft if the publisher of your article allows green open access/self-archiving.
If you have not made full text available, the site enables other researchers to send you a message with a PDF request so that you can send it to them even without making it available to everybody (journals always allow the sharing of PDFs between individual researchers for scholarly purposes) and even without knowing their eMail address (some grad students may be hard to google if their university does not put them onto the staff website). And of course you can use the same messaging tool to request PDFs from others.
In your profile, you can see some statistics on how often people have 'viewed' items from your publication list or even downloaded a full text, allowing you to assess how much interest there is in some of your research.
The annoying aspects
There is, however, more than one fly in the soup that is ResearchGate. The site seems to be somewhat over-eager to recruit new members, and its practices have been described as spamming. In particular, it seems to send out eMails that wrongly imply that a colleague has invited you to join when in reality they did nothing of the kind.
Then there are the various silly indicators of your presumed standing in the community. Not only can you list skills in your profile, but other people can 'endorse' you for your skills. That may sound nice in theory. After all, everybody can claim to know how to do phylogenetic analyses but perhaps you are inclined to believe it if they are endorsed for that skill by several colleagues who you already know to be competent.
In practice, this visibly leads to endorsement cartels of everybody endorsing their friends for whatever they have in their profile, and expecting to be endorsed back. Indeed most of the eMails I get from ResearchGate are about me having been endorsed by somebody for something, which looks a bit too much like patting each other on the back instead of doing anything of substance. What is more, many of the skills in my field are rather ill-defined: What does a skill like 'taxonomy' or 'biodiversity' even mean? No idea really.
I also currently have a 'citation confirmation request' in my inbox; a member has asked me to confirm that one of my papers cited one of theirs. Are you kidding me? Just think of what it would mean if we were asked to do that for every paper we have ever cited.
In addition to the endorsements, ResearchGate generates all manner of statistics and metrics itself. The ones for how often your papers have been viewed etc are one thing, but what please is the point of the RG Score?
The RG Score takes all your research and turns it into a source of reputation.Ha! I am always skeptical of something that purports to mix incomparable things into one handy number. In practice, it here appears to mean some mixture of publications, answers and questions contributed to the ResearchGate Q&A stream, and your ResearchGate followers. In other words, three quarters of the score then boil down to your activity on a social networking site. Yeah, as if that would impress anybody's line manager.
To summarize these observations so far, the underlying problem that shows in nearly all aspects of the ResearchGate experience is that it tries to attract and connect serious scientists with gimmicks that any serious scientists would have to consider to be beneath them. Spamming for new members, endorsing each other for some silly, vaguely defined skills, an RG Score that is based on how much of your profile you have filled out and whether you regularly ask questions - all this is indicative of a schoolyard level of social theory, and not even high school at that. It is about as appropriate when directed at a clientele of career scientists as garish adverts for fast food bargains would be when directed at royalty.
Another problem is that I find the Q&A stream rather useless, if not downright embarrassing. Admittedly I have not asked a question myself so far, but that is just the point: I have received scientific training, know how to do literature searches, and am able to read manuals, so I may not need to use it. Others are not so shy. To cite a few examples that show up when I look at the stream now:
Can someone suggest how to prevent encroachment onto lake and wetlands to sustain aquatic ecosystem?
Can measures to conserve biodiversity and reduce poverty reduction go hand in hand?Do these people expect an answer of less than book length to be of any help?
Can anyone tell me the proper way of leaf/root/stem samples collection from the plant for proteomics work with necessary precautions?How about reading the manual of the kit you are using? How about contacting the manufacturer?
Which outgroup and which genes are used for angiosperm phylogeny?
What are the best primers used for Boraginaceae phylogeny?
Does anybody know various theories on the origin and evolution of insects?How about doing a bloody literature search instead of having others do the work for you? There are papers out there that contain this information. This is not pixie magic or anything.
How we can screen the wild plant species for morphotypes, cytotypes and genotypes?I think somebody is confusing the Q&A stream of ResearchGate with the university courses they should have taken before pretending to be a scientist.
Can anyone suggest when reporting of a plant species into a scientific publication?Er, what?
I think you get the picture. Admittedly, I see the point of people asking for specialist help with the identification of plants and animals they collected, or for advice on what software to use for a given problem, but there are way too many questions where the answer has to be either Mu or 'why are you so lazy?'
Wikipedia (accessed 14 Dec 2013) informs me that ResearchGate has investors but I do not understand how they expect to earn money. Signing up is free for scientists and I do not see any advertisements anywhere (which is one of the attractions of the site). Apparently they are looking into earning fees from job advertisements, which surely seems like a win-win because many members would be interested in those, and into 'partnering with companies that manufacture and sell biotech lab equipment', but all in all I wonder how this venture is ever going to be profitable in the long run without doing things that the users would hate.
Another thing I learned from Wikipedia is that the founder of ResearchGate, one Ijad Madisch, 'has stated that he wishes to win a Nobel Prize through the site by disrupting the way in which science is conducted. Madisch envisions a future in which scientists will publish their positive and negative results and data on his site instead of paying to publish it elsewhere.' Lack of self-confidence is definitely not one of his personal issues, but I hope he is prepared for disappointment.
ResearchGate definitely offers some good tools for networking between scientists. It can increase the visibility of individual scientists, and it can facilitate the sharing of scientific results and publications. The poor quality of the Q&A stream is ultimately the fault of the people doing the Q; it has the potential to be better. Likewise, some of the interpersonal dynamics that play out with 'following back' and 'endorsing' each other are due to our personal psychology and not the fault of the provider, or at least not only.
Really the major problem appears to be ResearchGate's weird mixture of pretentiousness and science-unworthy antics. No, they are not going to replace all scientific journals on this planet, and they are certainly not going to do so if they continue to treat their clientele like kindergarten children.
(Updated to correct a wrong verb.)