As demand for the AltMill continues to grow, long lead times have been a sticking point for many customers either waiting for their machine to ship or considering a purchase. This article is designed to outline our philosophy and approach to company and production growth so that we can provide clarity on our goals and intentions for the longer term. This article will also show that long lead times are a temporary and transient part of our production process and intentional to allow us to provide the best value CNC machines on the market.
Disclaimer, I did my best to interpret the data as accurately and as representative as possible. However, it shouldn’t be used to estimate and predict future lead times, as things may change. Some outlier data may have been removed to better represent the average customer.
Historical Data
To outline the actual change in lead times customers have been experiencing, I’ve graphed data looking at the average wait time per customer, based on the month they ordered a machine versus the actual order completion date. This graph outlines the historical changes over time in lead time, showing a general downward trend for lead times. Please note that the dotted lines are estimates and may vary in real life. For more up to date and the best real world estimates on specific orders, please see our Order Status page.
AltMill (Launched March 2024)
LongMill (MK1, MK2, and MK2.5)
Why is this important to look at? First of all, lead times are initially long for machines when they are first launched. You can see large spikes at key points such as March/April 2024 for the AltMill, Oct 2019 for LongMill MK1, May/June 2022 for LongMill MK2, and LongMill MK2.5 in May 2024. Why do these large spikes exist?
- Version changes and new machine launches are typically announced before the full production of the machines is complete. This means that machines aren’t ready to ship right away.
- Lead times may be higher when demand is higher than expected, so initial batch sizes are smaller than demand, which makes parts availability a bottleneck.
- Delays due to quality issues, shipping in of materials, and development of assembly processes typically happen at the beginning of each production run.
But of course, as production continues, matures, and scales up, quality and process issues are resolved, and our batch sizes grow, lead times trend down, as we can see for all machines.
Cost, quality, speed
Many of us are familiar with the “cost, quality, and speed triangle – choose two – idea.” To put things simply, we have always chosen cost and quality, because speed comes from consistent growth – as you’ll see as you read through this article.
Our company’s mission is to make CNC more accessible, and cost plays a major role in accessibility. Yes, in theory, we could build machines faster if we spend more and consequently passed those costs onto the consumer, but we chose not to because, in the long run, we can have all three: cost, quality and speed.
Here’s a graph showing the number of customers segmented by the number of weeks they waited for their LongMill:
You can see that the vast majority of customers wait less than 4 weeks. So yes, when we look at the lead time graph over time, we have some large variability, the number of customers actually impacted by long lead times are actually minimized by a certain number of factors:
- While there is a large spike in orders after a product launch, overall demand is slowest at the start of each launch. I believe this is typical because not a lot of people outside of our community know about the product yet, or customers are waiting for reviews and other information to come out before making a purchase.
- Due to the long initial wait times, we push away customers who either don’t want to tie up too much money for a long period of time; are impatient to get a CNC and go with another company; or prefer to wait when lead times are shorter.
We also see demand grow as:
- Customers receive their machines and they share their positive experiences with the community.
- We can serve customers who are not tolerant of waiting a long time for their machine to ship.
At the time of writing, the AltMill is still in the early stages, before we’re able to see a trend in having the majority of customers experiencing a shorter lead time. But, we’re slowly getting there, and we plan to get there as soon as we possibly can.
Here’s a current distribution for AltMill:
Focusing on long-term production over short-term gains
Growing production takes significant time and investment to scale. Here are some of the things that we work on over time.
Inventory: As a hardware company, we need materials to build and ship products. Not only do we need to purchase and stock the optimal quantity of parts, but we also need to forecast sales and production rates for individual components so that we can reduce the amount of inventory we sit on while ensuring we have parts on hand so that inventory shortages don’t cause delays in production. Scaling up inventory at the start of a new product can be difficult since we can’t entirely predict what the demand will be, plus it can take some time to accumulate new cash to purchase more materials. While under-purchasing can lead to longer lead times, over-purchasing could lead us to running out of money on hand to make payroll and cover other expenses, depending on changes in demand.
Production space: As we scale the number of machines we produce per month, we also need to allocate space for inventory and space for assembly. Scaling space can have several risks, with one of the main ones being making a long-term commitment to signing a lease (for us, we’re currently on a 5-year lease on our building). Regardless of whether we use the space or not, we’d be on the hook for paying for the space we sign up for. Typically, we also need to invest capital to build out the space to fit our needs, which could include electrical and purchasing racking and benches. Space also doesn’t come often or exactly as we need, and there may be a wait period for another tenant to move out, plus a negotiation period for rates and terms to be agreed upon.
Human resources: We require skilled labor in building and assembling machines, which takes time to hire and train. This not only takes time to interview and hire new candidates, but our senior staff also need to dedicate time outside of their production time to work with new members to get them up to speed. Automation and strategic outsourcing also play a role in the amount of human resources we need. Time is required to identify tasks that are done repeatedly which are best suited for automation or outsourcing, which means that production at the start of a new machine tend to require more hands on work.
Technical support and resource development: Building and shipping a product is only half the battle. We want to ensure every customer is able to succeed with our products as well, which is why we dedicate so much time to videos, tutorials, and resources to help users every step of the way. Additionally, our support staff need continual professional development to be familiar with our products. You can read more about why our resources are so important to our business model here.
Design changes and improvements: To improve yield, quality, ease of production, and ease of assembly, our engineering team continually looks for ways to improve the design of our products and work with manufacturers to make the best components possible. Small design changes that shave of 10 minutes of assembly time could lead to thousands of dollars saved over the long run in lower production costs and shorter lead times.
So, breaking things down, here’s a list of costs related to scaling production that could negatively impact the cost of products and prevent production from being as cost-efficient as possible:
- Unused space means rent is being paid for unproductive space and utilities.
- Overstocking of inventory leads to additional costs for storage, plus money being tied up in parts rather than being used for other productive purposes.
- Being overstaffed means there may not be enough work to be done while overhead costs stay the same.
- The quantity of products being sold may not be enough to cover the cost of engineering and development costs.
However, scaling too slowly may come with other risks and costs, including:
- We give time for competitors to gain market share.
- Customers have to purchase other options if the lead times don’t align with what they can accept.
But the costs and risks above only apply when demand is higher than what we can supply, which means that they can, eventually, be overcome. Hardware, as they say, is hard. Mistakes in the production process and scaling too quickly can rapidly get out of hand. For example, messing up on just one of approximately 140 unique parts that go into an AltMill could result in production stopping completely. And when some parts can take 3 or more months to produce and arrive, some mistakes could seriously jeopardize our lead times.
Ultimately, we could charge more, which would allow us to have production at a smaller scale and more easily cover the costs of mistakes, plus lower overall demand and the volume of machines we need to build, but I believe that this would lead us to divert our focus on actually building affordable and accessible CNC machines. At the end of the day, there are many options for CNC machines that have shorter lead times, but it will come at a higher cost or a compromise to performance. It’s up to the consumer to decide what they value most at the current moment. However, if we continue to follow our trend of slowly reducing our lead times down to our general goal of 2 weeks or less across all products, we can protect our position in the market by not only providing the best value CNC machines on the market, but also at the shortest lead times as well.
Feedback?
While these are the processes we’ve been following to this day, we’re always open to hearing from our community on what they think we can do to continue supporting our ecosystem. Does it make sense to raise our prices if we can deliver faster in the short term? Should we implement dynamic pricing? Feel free to let us know your thoughts!