It is definitely not the sexiest job!
[How sexy can it be if it is sandwiched between commercial and IT function?]
Expect to use most of your time cleaning the data
Expect to use your time to try to align different stakeholders as they have different viewpoints on how the data should be clean or categorised or use. Different people have different assumptions in cleaning your dataset
Most data science jobs, especially in the non - tech firms, do not use machine learning algorithms. Simplicity and interpretability are keys
There is no ready data science tools in the organization and expect to wait for months in order for your instances in e.g AWS to be made available to you X amount of time.
Rather than using all your interactive d3 or plotly charts, ppt still the number one choice. D3 impress but when it comes to practicality, ppt still win
Organizations have polarised view on data science: Some thinks that Data Scientists are magicians and can make all things happen while some thinks that data scientist are not relevant to their business at all
Lastly, data scientist might not be earning as much as you think it is!
This is a question that I answered from Quora!