Network optimization


How do you balance between accuracy and complexity when assessing and optimizing your warehouse footprint?


All companies operating a physical supply chain periodically need to assess the logistical network to make sure it is cost efficient and delivers on current as well as future customer service requirements. A total optimization of warehouse footprint usually requires significant effort and resources to provide strategic guidance for needed changes. Given the potential magnitude of investments and impact on running costs, decision makers often push for a high level of details and accuracy. This tends to lead to high complexity in data gathering and analyses and surprisingly often counteracts the ambition by undermining findings when data integrity or analysis methods are questioned. A typical response to that is to involve one of the logistics network optimization tools available on the market, in order to allow complex analyses and add credibility to the final results. This risk however is that the optimization tool only acts as a “black box” that hides the analyses (or make them too complex for the human brain to interpret) instead of adding the desired level of accuracy.


All in all, the key question for any footprint optimization project is to balance between complexity and accuracy in order to get results that deliver on ambitions defined. Our experience is that there are a couple of guiding principles that can help you avoid these pitfalls.

  1. Always start from a future customer perspective and secure that all analyses are based on forecasted volumes and customer requirements on a given time horizon (typically 5-8 years). Make sure to involve sales, marketing, R&D or anyone else in the company that have input on this and can validate assumptions. A common mistake is to do a logistical network optimization as an internal exercise in the supply chain function, which adds significant risks for the end result.
  2. Define a simple and crystal-clear analysis model that captures all evaluation criteria (cost and service) and spend time on communicating this to all stakeholders. Surprisingly often re-work is required in later phases due to misunderstanding around simple cost definitions.
  3. Be smart when collecting data on current situation. It is very tempting to ask for “everything” to keep all options open for how you want to simulate the network – but it will just lead to a lot of wasted efforts and increased complexity in modelling. Instead, focus on defining early on how you want to model the current network/baseline and focus your data collection on the most critical data points. Don’t be afraid to simplify as long as the overlying principles of the model is kept intact.
  4. Allow sufficient time for data gathering, and when your feel your done allow as much time for data washing and validation with key stakeholders. Do not move into analyses before everyone have confidence in the data and the granularity level it has.
  5. Avoid politics based on the current footprint. This is extremely hard but usually is key to securing a robust end results. Make it clear to everyone that the analysis need to be based on a “white paper” approach in order to identify the most optimal setup, then in later phases it will be aligned with current footprint to determine a final recommendation on footprint that also include costs for realization. A common mistake is to very early define “no touch” warehouses and exclude them from the analysis, it might be convenient from a political perspective or reduce complexity but it will also deteriorate the total outcome of the optimization.
  6. Reduce complexity in optimization by focusing on a limited number of scenario exploiting different supply chain strategies (e.g. continental hub-and-spoke, dedicated geography, dedicated product range, dedicated by sales channel). Running iterative simulations within each strategy quickly produces results that help you narrow down the options further.
  7. Use tools to support complex modelling – but do not use a more sophisticated tool than you need to solve the optimization. Quite often an Excel based simulation will be sufficient. If you venture for one of the more advanced specialized software on the market – make sure to understand business drivers and the logic of optimization. The risk otherwise is that you end up with a “black box result”, which you will have a hard time explaining the rationale for.
  8. Allow time for discussions and communication. The outcome will not be worth much if key stakeholders don’t buy in on results and recommendations – so make sure everyone understands the analyses and have had an opportunity to speak their mind. Referencing back to #1, this will be especially important for stakeholders outside the supply chain function.


If you successfully follow these guidelines you’ll have a much higher chance of both arriving at a solid recommendation for future footprint and making sure you bring key stakeholders in the organization along on the change journey.



Digital readiness for supply chain

January 15, 2018

New publication on transformation



Kungstensgatan 62

112 39 Stockholm, Sweden

+ 46 76 191 61 15



Management consulting with true flexibility to adapt to your specific needs in order to create sustainable results.

Copyright @ All Rights Reserved