US Factory Visits – Insights and Opportunities for Entrepreneurs

I recently returned from the United States, where I visited more manufacturers. During this visit, I wanted to see the factories through the eyes of the owners and ask them questions about their businesses. There's no substitute for boots-on-the-ground research.

This trip took me to Michigan and Pennsylvania. Pittsburgh greeted me with winter temperatures that I'm only used to on the ski slopes! I visited and spoke with an eclectic mix of companies ranging from furnace producers, candy makers, and racecar parts producers to more generic machine shops and 3D printing factories. Most of these companies have <30 employees.

Which key themes arose during your visits?

The companies I spoke with have <30 employees. There are ~240,000 manufacturers in the US, 75% of which have <20 employees and ~93% have <100 employees. I mention this because specific key insights arose repeatedly. So, I think it is fair to say hundreds of thousands of businesses are facing the following issues in varying degrees of severity:

1. Knowledge Loss
2. Labour Shortages
3. Supply Chain Uncertainties
4. Unclear Approach to Retirement
5. Training
6. Lost Relationships
7. Efficiency Gains
8. Spare Parts
9. Data Gathering and Unification
10. Maintenance
11. Quoting

Setting up meetings

How did you arrange meetings?

I set up meetings in two ways: first, by asking my network for referrals, and second, by reaching out to owners. Naturally, referrals tend to work best because someone can vouch that you're not a crazy person.

I didn't record my conversations because I wanted the meetings to be relaxed and for the owners to give honest answers rather than the answers that looked good on camera.

How long were the meetings?

Most of my in-person meetings were two hours long – enough time to build rapport, get an overview of how the business works, and then tour the factory floor or shop while asking questions. I've also done some online interviews with factory owners and employees, but I think you can obtain information with greater context by visiting them in person. Often, when you ask a question, the owner will immediately take you to the spot in the factory where they can show you the answer.

What did you explain in advance to the owners?

I said that I'm an entrepreneur with a software background and that I'm exploring the following trends:

• Labour shortages
• Reshoring of manufacturing
• Opportunities for innovation in their production methods
• Retirement planning

---

Key Themes

Knowledge Loss

Knowledge loss is going to be a huge problem. As senior staff retire, they take undocumented knowledge with them. This creates a massive inefficiency – now your company must figure out how to redo something it previously knew how to do. Often, this involves duplicating past work or hiring a recently retired employee back as a consultant.

I dug further into this and uncovered the following:

• Conveying nuance is hard. Imagine your job – You possess different types of knowledge, varying in complexity. You could produce documents, audio, and videos to explain the general gist, but deciding when to go into greater depth is subjective. If you won't be around to answer questions after you retire, you have to try to pre-empt the questions that could be asked when documenting your knowledge.
• Most of the lost knowledge is tacit knowledge gained from personal experiences at work, which is complicated to explain.
• Engineers know that good documentation takes time and must be maintained.
• Technical literacy could present a barrier to a retiring staff member documenting their knowledge.

Labour Shortages

Manufacturers are facing constrained production due to labour shortages. If you don't have workers and can't fully automate what you do, you have a problem. The factories I visited experienced problems sourcing machinist, maintenance, and production manager skill sets. With more staff, they could increase production levels.

There are several factors contributing towards the labour shortage:

• Record numbers of retirees: In 2024, ~11,000 American baby boomers retire daily.
• Globalisation: Outsourcing production reduced the need for specific skill sets at home. Peace and prosperity enabled complex global supply chains to develop, which we now rely on.
• Conditions and pay: Many career opportunities exist which pay comparable salaries and have better working conditions than manufacturing jobs.
• A lack of encouragement to enter manufacturing: Factories informed me that we need more apprenticeship schemes. Similarly, they suggested that being a tradesperson has been stigmatised in favour of getting a college degree.
• Complacency: Society has amnesia about the importance of manufacturing. If we don't prioritise producing things, we must rely on others to do it for us, which has national security implications.

Supply Chain Uncertainties

Owners reported that they have been extending their delivery times to customers because of supplier delays. Labour shortages, electronic component shortages, war, and shipping delays were common reasons provided by suppliers.

Interestingly, some factories reported having no data or insights into what was causing delays – they simply relied on their supplier's narrative and retold that story to their customers. I talked about supply chain insights and risk management software with factory owners, but they highlighted the recurring issue of not having in-house skills to use and keep on top of such updates.

Unclear Approach to Retirement

In the next decade, many small business owners will retire. A lot of small manufacturing firms are family-owned and operated. Imagine now that you've operated this business for most of your career, and you were secretly hoping that your children would take over the company. I spoke with owners in this situation, and they highlighted that their children don't care for the family business and don't have the necessary skills to operate it. In this situation, you could:

• Close the business down
• Get acquired
• Hire an heir
• Not retire

If there's one thing all the owners I spoke with have in common, they're busy! If your factory is understaffed, you may not have the time to sit down and carefully plan out your retirement. Many owners I spoke with are integral to the operation of the business. Therefore, they try a combination of speaking with succession planning companies, local competitors, suppliers, customers, or industry people at tradeshows. A managed, competitive bidding process could yield a much better outcome for the seller.

Some factories also reported having workers of a similar age to the soon-to-retire owner. In that situation, even if there was a suitable in-house heir, the owner might be reluctant to promote them. Talking honestly with workers about your impending retirement can be challenging – some workers I met had been with their employer for 30 years. The owners don't want to create uncertainty; they want the best for their staff.

Training

Most factories I visited reported wanting to improve staff training. Factories can have vast differences in the output of workers doing the same job. I asked different owners what prevented them from providing enhanced training. They reported:

• Time: A recurring theme! They wanted to provide better training but were busy with other tasks. Several factories reported that training involved giving the staff members increasingly more complex tasks and having a more experienced staff member help whenever required. There's also a short-term hit to overall output if you ask your most productive workers to train others. Of course, with better training, output would increase in the long run, but long-term planning is often a luxury when short-staffed.
• Computer literacy: In some cases, workers had never received formal training.
• Computer accessibility during a work shift: I must admit that I hadn't considered this one, but it was instantly apparent when I visited factories. If your job involves wearing protective equipment like gloves, helmets, or ear protection, you will interact with computers differently. I asked one factory about placing training videos on tablets near machines. They highlighted that by the time the worker interacted with the machine, they'd be wearing protective equipment.

Lost Relationships

I visited one factory faced with a problem on the opposite side of the world. For decades, they've worked with a sales rep in Asia. The sales rep has secured them a lot of work, maintains valuable relationships, and is integral to their business in that country. Now, the sales rep wants to retire. He has no heir and plans to close his business down.

Consequently, the factory owner has to take a considerable risk and forge a relationship with a new sales rep before retiring. I say risk because the owner wants to sell his business, and losing the Asian relationships would significantly impact the enterprise value.

I asked the owner how this situation unfolded – he's had a great working relationship with this sales rep for many years. He's asked him to delay his retirement, and the sales rep has obliged. But this time, it's different.

So many small business owners will face this problem. Key partners they depend on will want to retire at the same time they are retiring. Without good succession planning, the relationships those partners have will vanish.

Efficiency Gains

I toured different types of factories; in every case, the owners could show me multiple ways to improve efficiency. The key reasons why they weren't doing that were:

• I'm about to retire, and it's a project for someone younger.
• A lack of in-house skills like software engineering.
• Time and money.
• Staff resistance to change.

One factory showed me decades of order data they kept in an Excel spreadsheet. "Has a data analyst ever looked at this?" I asked. The answer was no. I could only imagine the patterns sitting there, unsurfaced from that data.

Spare Parts

If your business involves producing a machine or building a product with subcomponents, you almost guarantee a longtail business for spare parts.

One factory showed me their order data. They frequently sell spare parts at 2-6x multiples. Numerous owners explained how the same situation arises with their customers: When they sell a product, they also provide a recommended spare parts list that the customer should keep in stock. Invariably, someone uses a part and doesn't replace it (a common theme), or a customer makes a short-term cost-saving decision not to stock all of those recommended parts. Then the machine breaks.

In many cases, the customer doesn't have the internal resources to shop around for a quote. They took a risk by not stocking the recommended parts, and now their business is losing revenue every day that the machine is offline. Price sensitivity is very different in such a situation. In some cases, I heard that a factory can make so much on a spare part that they'll remove one from a machine currently on the production line. That impacted machine still gets delivered on time because they have a reliable spare part supplier and can order another one.

This situation strikes me as a significant opportunity. How many companies get themselves into this situation? At any one moment, someone in the US probably has that spare part, and if there were more suppliers, then the price would naturally be driven down. The customer simply uses the original supplier because that company knows them, and they can quickly outsource their problem to them and get the machine back online.

Data Gathering and Unification

Even in small factories, there are many sources of data. I spoke with various owners who fully understood the importance of data, but several barriers existed between where they were today and where they wanted their business to be:

• What data does this machine gather?
• How does the data across my machines differ?
• Is that data accessible?
• If it is, is it worth accessing this data?
• How can we set up a data pipeline?
• How can we clean the data and then manipulate it?
• What are we optimising for?
• How can we interpret the data?
• How can we automatically surface the most relevant data insights?

Some factories I visited required an overarching strategy for both data and software. They couldn't deliver this in-house. The owners installed software at those factories in a piecemeal fashion. Without a CTO to guide the strategy, several factories reported installing software that did not meet expectations.

I visited one machine shop that was doing things differently. They were building software, a factory operating system. This was an exciting business with a visionary owner exploring deeper integration between those on the factory floor and software engineers.

Maintenance

Every factory that I visited had a different way of handling maintenance. They were at various stages of maintenance evolution – some were still doing reactive maintenance, whereby assets were allowed to run to failure. Others performed preventative maintenance by scheduling weekend downtime when someone would service the machine. Some had third-party maintenance contracts, and if anything went wrong, a technician would arrive the next day to repair the machine.

All of the factories I visited had yet to achieve predictive maintenance levels. I think that a sound data strategy is necessary before a factory can achieve a predictive maintenance strategy – you need to be confident that if you install industrial IoT devices, you have the skills and resources to interpret the alerts and data coming off those devices.

Quoting

The leading problem factories reported to me was taking skilled labour away from production and involving them in the quoting process. As a software guy, I naturally thought, why can't we have people outside of the factory contributing to these quotes, or better yet, could AI do this? Based on what the factories told me, I found it difficult to answer this question. I discovered a much-loved quoting software solution called Paperless Parts during my research process. I visited two different factories that gave me two opposing opinions. One said it couldn't capture the complexity of what they do, and another said it was perfect and had revolutionised their quoting. Both could be true because the factories produced different products of varying complexity – I'm simply highlighting that one solution doesn't work for all.

There are significant opportunities for entrepreneurs...

I believe that there are significant opportunities for technology entrepreneurs in the following spaces:

Technology.

• Wearables designed for harsher manufacturing environments. Subject matter experts use these devices to document core knowledge. Others use them to see that knowledge when operating machines.
• Automated data collection systems that gather data, group it, and surface insights from it with a particular focus on knowledge learned on the job.

Education and Training.

• Upskilling and reskilling workers into manufacturing.
• Creating apprenticeships and schools for trades.

Artificial Intelligence and Automation.

• Redesign how machines function to lower the technical competency required to operate them.
• Make AI accessible to small and medium-sized factory owners with limited subject knowledge.
• Some companies are developing humanoid robots. Often, in manufacturing, it's possible to automate something, but it's not cost-effective to do so. Similarly, there will be tasks that make economic sense to pass off to a humanoid robot. Most small factories I visited would struggle with a) determining if humanoid robots are worth implementing, b) implementation, c) monitoring, d) maintenance and e) data insights to drive ROI.

Marketplaces.

• Create industry-specific marketplaces that match retirees with individuals or cohorts interested in buying their business.
• Create a marketplace for spare parts which requires minimal time and technical interaction to match buyer and seller.

Software Services.

• The trend here is to tailor software services to be very simple to use and to have low time investment.
• Fractional CTO services to small factories to give them the skills required to have robust software and data strategies.

Factory owner retirement planning as a software service.

• Real-time factory order tracking updates software for the customer.
• Supply chain risk management software that provides tailored alerts for that factory. It could automatically check in with suppliers about specific risks and suggest alternative suppliers to the factory if the probability of disruption is high.

When I think about reshoring manufacturing, I often think about population differences between China and the US. Recent US policy has focused on derisking supply chains by bringing critical technology manufacturing back home or to closely allied nations. The differences between China and the US are profound – China has benefitted from decades of rural-to-urban migration, which has enabled a constant stream of new workers who can be trained to produce goods. In comparison, the US deprioritised manufacturing, and workers consequently shifted to other sectors. Reshoring manufacturing on the scale that the US wants to achieve will involve serious upskilling and reskilling of labour and, in my opinion, a lot more automation than we are seeing today.

I believe that the US has a wonderful opportunity right now with manufacturing and automation. The economic, political, and demographic climate is different today. In recent decades, outsourcing for lower labour costs and reliance on other nations has been viewed positively. Today, at least in some critical technology categories, that is no longer true. Consequently, opportunities exist for entrepreneurs to figure out how to manufacture more efficiently in the US. We can not expect to manufacture products in the same way as nations with populations 4x as large and wildly different living standards. This is further complicated by the record number of daily retirees.

We must get creative, redesign how we produce things, and optimise for a future of increased supply chain resilience. The renaissance of manufacturing is underway.