The results of the regression analysis originate some considerations on the relationships between the productivity
of a LSP and both operational and non-operational variables even if two of them have not confirmed the expected
behavior. In particular the STEM TIME presents a positive influence on the productivity. This is probably due to the
fact that a driver knows that he has to make a long run to perform the first delivery, and he will organize his
activities in order to complete deliveries and pickups faster, so his productivity could in turn increase. Relating to
the ratio between the stops for deliveries and the stops to pick up the model shows a negative relationship with the
number of stops and this could be explained because the system of payment of drivers is based on the successful
deliveries that they perform. Therefore a driver is likely to spend much time of its workday to successfully delivery
a parcel and trying to avoid failures, especially for B2C services, and this negatively affects productivity. The
negative impact of TIME WORK shows that if the driver has a less time to complete is activities he will likely to
rush more so that the productivity will be enhanced. Similarly, huge MASS SATURATION on the vehicle reduces
the potential number of stops, because the number of parcels that the driver could effectively load is lower. This is
especially true in the case of B2B deliveries where the volume and the weight of each single parcel are usually high.
For this reason the company should always pay attention to the vehicle loading strategy in order to enhance its
productivity. Then referring to KM TOT, outcomes have confirmed that the driver is productive if he makes more
km and this is due to the fact that there is a higher opportunity to meet more customers. Coherently, both the total
services completed by a driver and the area of the warehouse positively influence the productivity
Seven out of eleven variables show a significant impact on the number of stops. This result shows the level of
complexity of the system under analysis. This complexity has been also highlighted by Tamagawa et al., (2010) that
consider challenging the modeling of urban freight transport. However, in this environment two main managerial
levers can be identified for the improvement of the system. The first one is associated with the design of the network
and encompasses the STEM TIME, the TIME WORK, the kilometers covered by a driver, the number of services
that are completed, and the trade-off between the number of pick up and the number of deliveries. In particular a
more efficient location of the warehouses, an extension of the area covered by each driver and a more efficient route
structure can significantly improve the level of productivity. The second lever refers to the vehicle loading strategy
and to the dimension of the warehouse. In fact vehicles should not be excessively loaded, especially with big
parcels, so that the business can be performed more efficiently.