https://www.sciencedaily.com/releases/2020/03/200325110842.htm

Mobile geolocation data from Baidu Inc, combined with a rich epidemiological dataset from the Open COVID-19 Data Working Group, showed that local person-to-person transmission happened extensively early on in the coronavirus outbreak and was mitigated by drastic control measures. However, with an average incubation period of 5 days, and up to 14 days in some cases, these mobility restrictions did not begin to positively impact the data on new cases for over a week -- with things appearing to get worse in the 5-7 days immediately after the lockdown as local transmission was well under way.

Among cases reported outside Hubei, 515 cases had a known travel history to Wuhan and a symptom onset date before 31st January 2020, compared with only 39 after 31st January, illustrating the effect of travel restrictions in decreasing the spread to other Chinese provinces.