Since the pandemic, it has been easy to become fixated on forecasting demand and so lose sight of end to end capacity and constraints. It's vital to take a broad view of risks and constraints under different scenarios and take a view on how much it is worth investing and where (e.g. bigger safety stocks, higher unit costs, extra capacity etc.). Without this integrated view, it often happens that one constraint is solved only for there to be another one right behind it.

In addition to the external volatility, there has also been an element of 'self-induced' volatility as a consequence of not properly understanding the knock-on effects of decisions.

Best practice is to formally integrate S&OE (aka Integrated Tactical Planning) with S&OP / IBP processes:

• the starting point is the approved S&OP / IBP plan which needs to be disaggregated into weekly, daily or even hourly horizons by segment or line, depending on your cadence;
• there should be a clear demand control process that is able to distinguish planned versus unplanned demand;
• a formal regular / weekly capacity review serves as the handshake between planning and manufacturing and / or procurement and gives better awareness of constraints and bottle necks that may need to be mitigated;
• it is critical to establish your Time Fence i.e. where the cost to respond to change in demand is significant. This helps prioritisation and allocation when cost to respond to change is higher and clearly understood.
• critically, there should clear policies on acceptable safety stocks and on how to deal with deviations...what qualifies as a legitimate emergency requiring intervention, whose decision is it? This avoids each case being escalated and tying up time and attention;
• at the S&OE level, there will be fewer options and those options are usually more expensive but these learnings about options and costs need to be proactively fed into the S&OP / IBP cycle so that there's an accurate understanding of context at this level and due consideration can be given to whether to invest in generating more / better options. There needs to be a continuity of understanding and a defined handover of long term into short term;
• there should be clear roles, responsibilities, empowerment boundaries, escalation paths, integration measures and feedback loops. Holding firm to roles and responsibilities when under pressure is challenging, especially in relatively low headcount companies where the same people may have roles to play in the short-term and medium-term management processes;
• keep people to their competencies – empowered middle managers managing the short-term according to the rules of engagement and senior leaders managing the strategic horizon. If there is a person with multiple ‘hats’, they must wear their ‘demand hat’ in the Demand Review and their ‘supply hat’ in the Supply Review.
• in the face of uncertainty, try to refine inputs into the demand plan and, where the facts run out, discuss and agree assumptions. Focus on what has changed each month and what you can do to influence demand. Consciously decide which demand plan changes can be accommodated so that we protect the time fence.
• use ‘available to promise’ functionality and/or good communication with suppliers to ensure reliable order promises to customers. Differentiate our handling of ‘normal’ orders which match the demand plan and ‘abnormal’ orders which we were not expecting. This includes regular and honest conversations with suppliers and partners.

What, exactly, are control towers (aka command centres)?

• There are several capabilities that constitute elements or levels of control towers as defined by different organisations, including:
• reporting dashboards which track KPIs for different parts of the business at different levels
• data visibility platforms like data lakes that combine data from different tools
• transport consolidation hubs
• exception reporting systems which provide alerts
• scenario modelling and optimisation
• A useful concept is the CORE framework which presents an hierarchy of capabilities with Configuration at the top, followed by Optimisation, Response and Execution at the lowest level.
• A control tower as a monitoring capability focused on execution sits at the execution level. In practice, this can mean many small control towers, each quite small, focussed on a different aspect of execution like deliver, warehousing etc.
• The concept of a 'command centre' implies an additional analytical capability such as scenario modelling which places it at the optimisation level of the framework.
• A digital twin relates to the top configuration level where it is possible to visualise how all the optimisations relate to each other over time and parameters can be adjusted.

What are the intended and actual benefits?

• The intended benefits are generally related to reducing the latency between exceptions being identified (or anticipated) through to being managed which minimises the impact of disruption to inventory levels and customer service.
• Another common benefit is gaining a better level of visibility and coordination between siloed operating units leading to efficiency opportunities by using combined scale whilst preserving operational autonomy.
• As part of a wider logistics restructure, a central control tower has been the vehicle for taking ownership of physical supply at the point the demand is created for suppliers to manufacture. This has led to more strategic supplier collaboration and earlier visibility on exceptions that would previously have had bigger impacts downstream.

How should the control tower relate to planning?

• Control towers could provide useful visibility on warehouse capacity or container availability or pricing, for example, which could in theory serve supply / capacity-constrained planning and demand management. Referring to the CORE framework mentioned above, this the cusp between execution, response and optimisation.
• It is very important to be clear on this relationship to avoid duplication of effort and, more importantly, the potential for confusion if duplicate systems produce different results.
• One approach is to use information from control towers to update planning parameters but not the demand plan itself.

How to stage / phase implementation and manage stakeholders?

• Of course, it will vary between organisations but common best practices are:

1. consolidate data & create visibility
2. proactively manage exceptions
3. drive scenarios, impact assessments and recommendations
4. integrate internal & external partners 
5. converge & automate processes

• From a change management perspective, a productive approach has been to identify areas that would benefit all stakeholders but with minimal risk.
• Although different operating units may operate in different markets and compete on different aspects, they all 'consume' logistics in the same way which is an effective point of leverage.
• Commonality of planning tools and dashboards are also opportunities to drive convergence for control towers and end-to-end orchestration.

In-house or outsourced?

• Considerations include:
• the risks (as well as potential benefits) posed by making extensive data available to external parties;
• not re-inventing the wheel when it comes to enabling technologies
• strategic data science capabilities to understand and adapt platforms and use cases
• Working with lead logistics providers on control towers offers a balance: the capability is outsourced but as centrally and strategically as possible.

Supplier collaboration

• A common challenge is accessing data from suppliers who can be resistant if they face many requests to provide data to multiple customers, especially if that entails different formats and interfaces.
• This is less likely to be an issue for businesses that are large customers and who already have significant compliance requirements related to cross-border trade for example.
• However, control towers can help by offering a two-way street: greater downstream visibility can feed into longer-range forecasting that, if shared with suppliers, can help them better manage their resources and working capital.