I believe that any industry can benefit from Big Data. That said, if your data is incorrect or untrustworthy, then the effects on your company can vary from "bad" to "catastrophic." Imagine if a Wells Fargo or a Bank of America doesn't confirm that the interest rate being charged on certain cards is correct per the credit agreement. They can either lose millions (billions?) of dollars in lost fees OR they could overcharge and be subject to litigation. On the other hand, if a smaller company who makes hydraulic valves determines that sampling data (versus a "big data" Hadoop system) is sufficient and they continue to be successful, there is no harm in them maintaining their approach to using sufficient data to make intelligent decisions.
I think that financial companies and those with higher legal liability *could* be more at risk; however, inferential stats allows us to make decisions on populations using samples. Proper application of those techniques - along with data validation/verification - can help mitigate risk.