Last year, I set up an account on ResearchGate, a social network site that describes itself as "built by scientists, for scientists." One motivation for doing so was simple enough: some of my work had already been archived there, and it seemed reasonable enough to go ahead and claim that work, and have another place to display other research that I have published. My experience so far has been somewhat positive, but with a few concerns.
On the positive side:
1. Setting up an account is simple enough. I merely had to provide an academic email address, create a password, and within minutes I was creating a profile.
2. It works like a social network, as its developers intended. In many ways I find it somewhat more social than Linkedin or Academia.edu. There appear to be more users, and there does appear to be some meaningful interaction. There is a space for participation in discussions, but it is less intrusive than LinkedIn and I don't have to wonder how much my inbox will be flooded by conversations that have no interest to me. Many of my colleagues from a variety of institutions are on ResearchGate and I enjoy having a convenient means of keeping up with their recent work.
3. I have one more gauge of how well my research is received. Google Scholar is fine for giving a rough and fairly comprehensive idea of how often my work is cited. Although ResearchGate is less comprehensive when it comes to citations, it does offer gauges that measure downloads (the site now merely refers to them as "reads") and profile views. These have been helpful to me as I assess how newer publications are faring. I can also see how well my work fares compared to others within my institution (or at least those who have a ResearchGate account).
4. Their estimate of impact factor is fairly up to date, which I find helpful as I continue to move forward with the research side of my career.
5. Although a bit sluggish in response time, ResearchGate has been good at adding peer review journals to its database, as long as the appropriate information is made available to them - e.g., ISSN number, journal website, etc. That is helpful for those of us who have published in very obscure journals that although legitimate, do tend to slip through the cracks.
On the negative side:
1. One disadvantage is that the staff at ResearchGate tend to take their own sweet time to respond to questions and concerns. Last year, it took several days. Apparently things have deteriorated this year, and I have at least one question that has been ignored for a couple weeks as of this writing.
2. My department did not have its own profile within my university. I requested a department profile, and provided all sorts of information from my university's website to back up the legitimacy of the department. Over a year later, and still no response.
3. Although in theory an academic account is supposed to be required to set up a ResearchGate profile, it strikes me that RG is prone to the same problem with fake profiles that plagues other social network sites. For example, in my department, according to ResearchGate, we have an individual who has over two decades-worth of research in physics, even though our department is strictly devoted to the behavioral sciences. That individual does not exist anywhere among our institution's staff. I did come to realize that one of our students matched the name on the profile, and I asked that student about this apparent late career change. Suffice it to say, the student was rather taken aback that there was even a profile in that student's name. I have alerted RG to the problem to no avail so far. Facebook works faster at dealing with bots and fake profiles, and I have rarely been impressed with Facebook's response to such situations.
4. There is a problem of individuals claiming work that is not their own. Although that has not happened to me personally, I am aware that others have had that happen to them. ResearchGate desperately needs to find a way of dealing with situations like that, other than to simply ignore their existence.
5. I am skeptical of the so-called RG score's usefulness, especially given how secretive RG is regarding how that score is formulated. In the scheme of things, this is a minor problem.
Overall, I think the advantages still outweigh the disadvantages. It's fairly easy to use, easy to interact with fellow researchers, and has some features that can allow one to gauge how their work is being received. However, the disadvantages noted above are ones that will affect ResearchGate's credibility in the long term if not handled soon. If the company is understaffed, they need to work on that in order to make sure that member concerns are handled in a timely manner. If there are bots abusing the system for whatever reason, that needs to be dealt with immediately while it is still a fairly rare occurrence. My concerns aside, I will continue to use ResearchGate for the time being.
Update (3 Nov 2015): Regarding point 3 on the disadvantages of RG (i.e., the creation of fake profiles), I did some detective work regarding our potential false profile. The student whose name was usurped is younger than some of the publications listed in the profile in question. We'd have to believe some very unscientific things for such a profile to be even remotely plausible. RG's staff apparently cannot be bothered to care. That is not a good harbinger for what could be a promising social networking site for those of us in the sciences.
Final update (10 Nov 2015): Eventually RG took care of the apparent false profile, a few days after my last email in which I made the case that it would have been quite implausible for the profile holder to have the publication record that was purported to exist. I was actually surprised as my last email received no acknowledgement. That notwithstanding, I was relieved. Hopefully the affected student had been corresponding with the RG staff as well. Although I think RG has somewhat better safeguards against false profiles than other social media sites, there are some apparent holes in their vetting system that will likely need addressing. For now, I will continue to use RG as its benefits continue to outweigh its potential drawbacks.