I've seen in many research projects that they give people who fill the questionnaires or participate in the experiment (more often the latter) some amount of money to acknowledge their help. But I can't pay that much because it's a small project. But also I want to attract many people to fill my online questionnaire. Is that right (according to research ethics or anything else) to say to them that I will give only random gifts, for example only 10 gifts when more than 500 people participate? My concern is that the participants subjective perception that they will receive the gain for their participation is not true (in terms of probability) and maybe I'm cheating on them, in a sense ?
Questionnaire design is a science.
Within that science, there is a body of literature on the ethics of incentives, and on maximising response rates. Behavioural economists and other social scientists have put a lot of work into this: you just need to make best use of their findings.
Just because questionnaire design looks easy, and lots of people do it shoddily, doesn't mean you should. Apply to questionnaire- and incentive-design the same academic rigour that you will apply to the analysis of the results.
In other words: the answer to your question is in the literature, so consult it: and then you've got some proper peer-reviewed literature to cite in your methodology section. And discuss the ethics with your university's committee and with your more experienced colleagues.