Logistics costs often represent large portion of them. To reduce them, depot location and vehicle routing are crucial decisions.

Location is strategic decision problem and routing is either tactical/operations decisions problem. Most of the time two decisions are tackled differently, it results increase in total costs(i.e. sub optimal decisions).

you can develop an algorithm that has to consider the principle is to alternate between a depot location phase and a routing phase, exchanging information on the most promising edges.

It depends if customers' locations are fixed (for a long period of time) or change frequently. If customers are fixed, it makes sense to integrate location and routing problem, altough it is not a simply task. Otherwise an approximated analysis of expexted flows would be sufficient, because it does not worth to solve a specific location-routing problem, if the routing problem is going to change frequently).

To judge upon the strategic value of a location can be done by incorporating prospective costs on the operational level. (This could possibly be done by implementing a policy rather than solving all possible instances on the operational level. But it allows meaningful integration.).

The VRP is usually operational or tactical problem. It also depends on the type of company.

See: Reza Farahani, Shabnam Rezapour, Laleh Kardar, "Logistics Operations and Management: Concepts and Models", 2011.

It is clear that location is long period decision (strategic), while routing can be considered as tactical decision. 

Nevertheless, I consider these problems should be integrated since the distribution of the population among the space (countries, regions, urban areas, etc.) does not vary significantly with the pass of the time.