6 tips for increasing delivery reliability with batch size 1

Delivery reliability is a key performance indicator that generally has a direct impact on customer satisfaction. It is therefore not only in companies that want to differentiate themselves from the competition through high delivery reliability that the sales department constantly demands high delivery reliability. They have to meet these demands in production and logistics despite the increased other requirements. At first glance, one would reply that this would require additional capacities and therefore additional costs. This is not an option for most companies. We will show you how you can increase delivery reliability without additional personnel or production capacities.

1. define the term delivery reliability in a standardized way

Which delivery reliability do we want to optimize? Delivery reliability according to customer requirements or delivery reliability on the confirmed date? Do orders that are delivered too early count as deliveries? Are whole orders, order items or pieces measured and are partial deliveries accepted?

Start by answering these questions about the status quo: Like many other companies, you will probably find that every department and even every specialist defines this key figure differently.

You should start by collecting the various definitions and selecting the best option for the customer. You will get to know your company from a different perspective. The selection of the “right” variant varies depending on the industry and company and cannot be answered in general terms. This step alone will increase your visibility on the topic of “delivery reliability”, which can lead to an initial improvement. Management attention is indispensable in this phase. However, explaining the benefits of high delivery reliability will be a challenge for very few people.

In the next step, you should take a closer look at the data basis for determining delivery reliability, true to the motto “Trust is good, control is better”. Established structures – including in the ERP systems – can be to blame for an incorrect conclusion. How is the data collected and at what point in time is the time stamp written? Are the definitions of your analysis and data collection congruent?

2. introduce the correct booking points in your process

You have defined the term “delivery reliability” company-wide. If delivery reliability is not yet measured at all or not measured correctly, you should create the necessary data basis. In my experience, this is where the greatest leverage lies – especially in medium-sized companies. This means that time stamps are recorded during the order fulfillment process when defined interim targets are reached.

This data basis is not only the foundation for improving delivery reliability, but also the basis for many digitalization approaches, such as process mining:

1. Delivery date as communicated to the customer: Can usually be found in the order confirmation (note: the initial date must not have been overwritten!)
2. Actual delivery date: You will receive the following from your CEP service provider or forwarding agent
3. Shipping date/goods issue: The transaction data of your ERP system
4. Release for production
5. Planned and actual shipping date

Once you have created this data basis, it is a good start. You can also go directly to the next step by measuring delivery reliability per department or production line. To do this, you need the planned and actual production dates for each department.

At the end, you should have created a consistent database of transaction data that will form the basis for your further investigations. Draw initial conclusions from the data! Are the main problems in production or even before that?

1. Is the difference between production release and planned completion less than the theoretical production time? The necessary materials may not have been available or the deadline may have been scheduled incorrectly. The problem lies here even before the start of production.
2. Is the difference between production release and planned completion greater than the theoretical production time? How long were the items in the buffers between departments? Were there any disruptions? In this case, you should concentrate on production.

3. check your promises to customers

From a market strategy point of view, there may be good reasons for promising the customer a delivery time of 24 hours, for example. This can be dangerous: Amazon, for example, advertises next-day delivery, but not for all products – there are exceptions here too.

Therefore, get to know your own product mix. What are the lead times for your different product groups? Which characteristics increase the risk of the product not reaching the customer on time? You can proceed in two stages:

1. Ask the sales department how delivery times are calculated and when the customer will receive a confirmed delivery date. Also talk to production and ask them to explain the lead times. It is usually easy to notice if something is wrong.
2. Use big data methods: with the help of Tableau, you can link your transaction data in production with the order data and examine it for anomalies. Discuss your findings with long-serving employees, allowing you to generate additional information. Or use process mining to bring transparency to your processes.

Don’t worry: you can leverage great potential even without big data, the important thing is to just get started!

After you have categorized the product groups according to their lead time, discuss the results with Sales and Production. Can the lead times be reduced? Why are they longer than average? Is a short lead time necessary for the product group or can we extend the delivery time for the customer? Perhaps an early delivery date is less important to the customer than a reliably met later delivery date. Interdepartmental communication is the be-all and end-all when it comes to improving delivery reliability.

4. prioritize the orders correctly

Production capacities are usually tight – especially at peak times – and represent a bottleneck. If this is not the case, we speak of waste. You therefore have the challenge of prioritizing your orders correctly.

In most of the companies I have advised, there were several bottlenecks within production. This greatly increases complexity and prioritization without simulation tools can be difficult. The aim should therefore be to set up production in such a way that a bottleneck is created. This allows you topull from this bottleneck(pull principle) and reduce the set-up times in this process. You therefore focus on prioritizing this bottleneck first.

If the bottleneck is not immediately apparent, check your data: Where is the largest buffer in terms of time? Do the buffers differ for the various product groups?

Further options for eliminating bottlenecks

If bottlenecks cannot be eliminated using these organizational methods, you have two options:

1. Increase in capacity through additional personnel or new systems; in this case, you must compare the rising manufacturing costs with the increased delivery reliability
2. Creation of defined buffers before each bottleneck and optimization of bottlenecks (set-up time, planned maintenance, etc.)

In most cases, priorities are set at the beginning and work is carried out according to the FiFo principle from this point onwards. In many cases, the production employee or chance decides which product is processed next. Find out how the departments prioritize! Do all employees within a department or production step act according to the same criteria? You will find that each specialist optimizes their own step, but across the entire supply chain this may not be an optimal solution.

Create transparency about the current situation and use your GMV (common sense) to consider what future prioritization should look like. The most important thing is to involve the entire workforce, ask questions and take their suggestions into account. Only then will a good solution emerge, which will then be accepted.

One solution: prioritization through remaining buffer time. Estimate the expected production time for an order – the minute is not important here, as we are talking about batch size 1 – depending on the characteristics of the order and calculate backwards from the time of dispatch:

Remaining buffer time = [Cut-off time of the shipping service provider] – [Remaining production time of the subsequent processes incl. planned buffers]

This allows you to introduce a dynamic pull principle that takes into account the special features of small batch sizes. Please note, however, that the right software support is a prerequisite.

5. set up a project to drive forward the implementation of measures

In the course of the data and process analysis and the definition of key figures and prioritization rules, you will come up with some measures or they will be suggested by the employees. The biggest mistake you can make is to stop at this point, because then you will not have achieved any sustainable improvement and staff motivation will drop. Imagine your suggestion for improvement simply being ignored! How do you feel about that?

To ensure that the optimization is sustainable, the implementation of measures must be organized. To this end, a project is set up and structured according to the company divisions:

• Distribution
• Purchasing and disposition
• Work preparation and production
• Logistics and shipping

In our experience, it is helpful to implement no more than two to three measures per area at the same time. One sub-project manager per area is responsible for implementing these measures and progress should be discussed every one to two weeks. Each area assesses from its own perspective and can contribute new ideas. Set achievable goals and create commitment by setting and checking deadlines!

6. set yourself achievable goals

The foundation has been laid and you have noticed the first “low hanging fruits” during data preparation, which you were able to optimize directly. Now you should set yourself targets so that you can justify any investments. Remember that it is easier to increase delivery reliability from 60% to 80% than from 92% to 95%!

The goals should be SMART: Specific, Measurable, Attractive, Realistic, Timed.

• Specifically: Which delivery reliability do you want to increase? See point (1)
• Measurable: You have created the data basis and can measure delivery reliability at any time
• Attractive: The benefits of improving delivery reliability are easy to explain: Customers complain less and are happy – what more could you want?
• Realistic: 100% delivery reliability is not really achievable in practice, what do you dare to do?
• Scheduled: When should, for example, 90% delivery reliability be achieved? This is the only way to create a project plan!

These targets should be broken down to individual departments; measure the departments not only by productivity, but also by delivery reliability:

• Could a product not be produced because materials were not available? This order is allocated to purchasing or materials planning
• Was a deadline scheduled unrealistically in order to build up pressure from the sales department? The responsibility lies with the sales department
• Based on the movement data in production and dispatch, the reasons for the delay can be broken down to individual production areas.

In the next step, those responsible in the departments can get to the bottom of these delayed orders. The prerequisite is the correct provision of information!

If you visualize the goals and key figures everywhere, you increase awareness and improvements become immediately visible. There is no easier motivation!

Summary

With my 6 practical tips, you should have taken a good first step towards improving delivery reliability. The 6 tips were recently implemented as part of a comprehensive catalog of measures at a tool manufacturer and have led to an improvement in delivery reliability from 92% to 96%.