"Sociologists conduct polls all day"
In my attempt to answer this question, I conducted a Facebook poll and asked my friends what they think sociology is all about. As a result of this (extremely unrepresentative) poll, I found that a number of "study participants" think sociologists poll people all day.
Actually, this is not the case. Polling is only one possible sociological method among many, many others. Accurately done, polls can be very useful. But the trouble is that it's hard to do them well - especially when the split in public opinion is near the middle, and tiny numbers can make a big difference. Polls are hard because people lie, they're hard to design, and they are often criticised for showing how people answer questions rather than what opinions people have. But most importantly, they are difficult to use as predictors - we want public opinion polls to tell us what president will win an election, but for all the reasons above, this is not always straightforward.
Luckily, sociologists have a number of other tools up their sleeves for studying how society works. We can work qualitatively and analyse interviews, we observe people's behaviors in different environments, we can study social media and analyze memoirs, speeches, newspapers, or dig up documents in archives. We can even work quantitatively without looking at polls, for example by analyzing income data or by studying networks and the links people or groups have between them. These are just a few of the things we can do aside from polling.
"Sociologists try to find jobs in other fields"
Maybe some do. But many sociologists are also passionate about what they do and would't exchange it for anything else! I do a lot of empirical work, which has allowed me to meet some remarkable people and to travel to stimulating, difficult, beautiful places. One of the things I value most about my field is the unique kind of bond I form with people during research - and the way in which that bond holds me accountable before them as I sit down at my desk to analyze my data and to write. This is just one small step on the way towards understanding how society is organized, and then towards imagining how it can be better. To me, sociology is about understanding as much as you can and trying to be as honest as you can, and that is a field I'm not eager to leave.
"Sociologists figure out and predict social trends through statistical analysis"
Some people overestimate the predictive aspirations and abilities of sociologists. Just like with polls, many quantitative studies do not aim to tell us what is to come. Rather, they help us understand a snapshot of the world and the conditions under which people live, think, feel, and interact. Moreover, not all sociology uses statistical analysis. Many studies are based on interviews and observations, which brings me to the next misconception.
"Sociologists make fun of anthropologists"
Actually, sociologists and anthropologists collaborate, read the same books, work together to enrich our understanding of the world. This is especially true for qualitative sociologists, who borrow from anthropological methods. We grapple with many of the same topics, from education to race, to healthcare and gender. There are many ways to articulate the differences between anthropology and sociology, and I would say the main one is sociologists' drive to generalize from one small case to a larger population - as compared with anthropologists' immersion in the unique and the local.
Indeed we do. Sociologists are trained to interview, to poll, to organize focus groups and lead discussions, and many sociological projects involve collaborative group work. But as is the case all across academia, parts of the journey are solitary and even lonely. Most PhD dissertations, for example, involve hours of pouring over hundreds of pages of your own text, and this can be a very isolating yet beautiful experience.
So what do sociologists do?
For one thing, they let people talk, and they quote extensively. So I will leave you with a couple of beautiful interpretations offered by those kind enough to chime on my small Facebook poll:
"Sociologists make the familiar strange. They look deeply at the things most of us take for granted as normal and natural, and try to analyze when/how/why they became normal." (Katherine Felts)
"Sociologists aim not just to interpret the world, but to change it." (Pete Gardner)
"Sociologists observe others(') observing...[sociology is] an essentially recursive discipline - one for which its own study can always be, self-reflexively, the object of sociological study. Any definition of sociology has to include the capacity to study sociologists sociologically." (Jake Fraser)
Regarding predictions in sociology (a point made in the answer given here by Olga Zeveleva), I would say that polls actually work pretty efficiently in many countries in a context of elections predictions from 1936 onwards, when Gallup introduced random sampling and predicted Roosevelt's victory. To my knowledge, forecasts in the USA failed in a couple of cases: in 1948 when Truman's victory came unexpectedly, and in the case of Trump (though it is difficult to predict a small percentage difference in an indirect election system, the reasons for the unexpected elections here are indeed multifaceted and complex). Also, the impact of social desirability (giving the answers the respondent thinks the interviewer will like to hear) differs across cultures.
As for qualitative sociology, it is not always about predictions, but also about associations of different phenomena, classifications of certain groups of people. For example, those who feel incapable of impacting the political situation in a country are more likely also to have negative attitudes towards migrants or other out-groups. In stating that, I am not trying to determine what comes first (I may even suppose that these things come simultaneously).