After I raised my first $10,000 for Métopi, one of the first things I did was hire a freelance industrial designer/mechanical engineer. It was one of the best early decisions I made. He was much better at the technical work than I was, could solve problems faster, and freed me up to focus on the things I was better positioned to do as the founder—talking to customers, applying for funding, building relationships, and figuring out where the company needed to go next.
That experience taught me that you should not try to do everything yourself just because you are an early-stage founder. The right freelancer, consultant, or development partner can dramatically accelerate what you are building. But I also learned the opposite lesson later on: one of the easiest ways to burn money is paying very talented people to solve a problem you have not fully defined yet.
Later, one of our biggest product goals was keeping the inhaler platform as small as possible. I treated that size as a hard product requirement and hired an experienced design firm to engineer around it. We spent months and nearly $40,000 developing a mechanism that would allow us to preserve that form factor. Then I went back to customers and realized they were completely fine with the product being slightly bigger.
The design firm had not done anything wrong. They were solving exactly the problem I had paid them to solve. The mistake was mine: I had spent serious money optimizing around a constraint before validating whether customers actually considered it a constraint.
Those two experiences shaped the way I think about outside help now. The lesson is not “do everything yourself.” It is to do as much as you reasonably can yourself until you understand the problem well enough to know exactly where someone more capable should take over. Whether you are paying someone $200 on Upwork or signing a $50,000 development contract, you should know exactly what you are trying to learn, build, or accomplish before the meter starts running.
Do as Much as You Reasonably Can Yourself First
Early in a company, your money is usually more limited than your time. That does not mean you should teach yourself electrical engineering, regulatory affairs, software development, industrial design, manufacturing, and patent law. There are plenty of areas where specialized expertise is absolutely worth paying for.
But you should understand the problem well enough that you can intelligently hire someone to solve it. Before bringing in outside help, try to get as far as reasonably possible yourself: research the problem, sketch the concept, talk to customers, understand the terminology, look at competing solutions, and write down what you believe the requirements are.
The goal is not to become the expert. The goal is to avoid paying an expert to spend their first several weeks figuring out things you could have clarified before they started. The clearer you can be about where you are, what you already know, what is still uncertain, and what you need from them, the more efficiently they can help you.
Validate the Requirement Before Paying Someone to Build Around It
This was probably the biggest lesson from my $40,000 mistake. Before paying someone to engineer around a difficult product requirement, make sure that requirement actually creates value for the customer.
Founders naturally develop preferences about their products. You want it smaller, faster, prettier, lighter, more automated, or packed with another feature. Eventually those preferences can start feeling like customer requirements even when nobody has actually told you they are.
Before committing serious capital, show customers the tradeoff. Would they rather have a slightly larger product that is significantly easier and cheaper to develop? Would they actually pay more for the feature? Would removing it change their willingness to buy? Does the thing you are about to spend $20,000 developing meaningfully strengthen the value proposition?
If customers do not care, you may be solving an engineering problem that does not need to exist.
A better sequence is:
Understand the customer need
Validate the tradeoff
Define the technical requirement
Hire someone to solve it
Know Exactly What You Are Paying For
Before engaging any freelancer, consultant, or firm, you should be able to explain the expected outcome in one or two sentences. “Help us improve the product” is too vague. Something like “determine whether this mechanism can reliably accomplish X within Y dimensions, and provide CAD files, test results, and recommendations for the next design iteration” is much clearer.
The clearer the deliverable, the easier it is to evaluate the quote, manage the project, and know when the work is complete. Before signing anything, understand what exactly will be delivered, what is excluded, how many revisions are included, what assumptions the quote is based on, who owns the resulting work and IP, what happens if the first approach does not work, whether pricing is fixed or hourly, and what could cause the scope or cost to increase.
This becomes even more important with larger firms because a proposal that looks comprehensive may still contain assumptions that create expensive change orders later.
Ask Around Before Hiring Someone
Referrals are extremely valuable. If you are considering a design firm, manufacturer, consultant, developer, CRO, or other specialized provider, ask founders and people in your industry who they have used.
You want to know more than whether the firm has a polished website. Ask whether they were good, stayed on budget, communicated well, understood startups, and whether the person giving you the referral would hire them again.
A firm can be technically excellent and still be a bad fit for an early-stage startup if their process assumes you have a huge budget, a dedicated internal team, or perfectly defined requirements. For smaller, well-contained tasks, platforms like Upwork can also be extremely useful. You may not need a full-service agency every time you need a CAD adjustment, graphic, landing page, spreadsheet model, or piece of research.
Regardless of where you find someone, previous work and references matter. The cheapest provider can become very expensive if you eventually have to hire someone else to redo the work.
Get Multiple Quotes, but Compare Scope, Not Just Price
One thing I learned from working with technical service providers is that two quotes for seemingly the same project can be dramatically different because the firms are actually proposing different amounts of work.
One might include testing while another only includes design. One might include several iterations while another assumes the first design works. One might include project management and documentation while another leaves that entirely to you.
So if one firm quotes $10,000 and another quotes $30,000, that does not automatically mean one is three times more expensive. Ask both firms to explain exactly what is included and, just as importantly, what is not. Sometimes the higher quote is actually cheaper once you understand the complete scope.
Communicate More Than You Think You Need To
Once someone starts working for you, do not disappear for three weeks and wait for the finished product. Stay involved.
You want outside partners working efficiently, but you also want to catch incorrect assumptions before they turn into days or weeks of unnecessary work. A short recurring check-in can save an enormous amount of money.
Ask what they worked on that week, what they are working on next, what assumptions they are making, and whether anything has changed from the original scope. You are not trying to micromanage specialists. You are trying to make sure they do not spend 40 hours moving in a direction you could have corrected after hour four.
This is especially important with hourly work. If someone discovers a new problem, changes direction, or begins work that was not originally anticipated, you want to know before the invoice arrives.
Use Different Types of Outside Help for Different Problems
Not every problem needs a firm. Sometimes you need a freelancer for ten hours. Sometimes you need a consultant for strategic advice. Sometimes you need a specialized firm because the work requires equipment, regulatory expertise, quality systems, testing capabilities, or a multidisciplinary team.
The mistake is assuming the most expensive or established option is automatically the right one. For a small, clearly defined task, a freelancer may be enough. For a question requiring specialized judgment, a few hours with an experienced consultant may save you months. For major product-development work, testing, manufacturing, or regulated activities, an established firm may be necessary.
Match the resource to the problem. Do not hire a $300-an-hour consultant to do work you could reasonably complete yourself, and do not hire a $20-an-hour freelancer to make a decision where being wrong could cost you $100,000 later.
Treat Every Outside Project as an Experiment
One framework I wish I had used more often is asking: What uncertainty will this money remove?
If you are about to spend $25,000, you should know what will be true afterward that you do not know today. Maybe you will know whether the technology works. Maybe you will know whether the product can be manufactured. Maybe you will know which design customers prefer. Maybe you will have the data required to begin a regulatory conversation.
That makes it much easier to decide whether the project is worth doing. You are not just buying engineering hours or consulting time. You should be buying evidence.
The Biggest Takeaway
Outside experts can dramatically accelerate a startup. Some of the most important progress I made with Métopi came from working with people who had capabilities and experience I simply did not have.
The freelance industrial designer I hired after raising my first $10,000 was a great use of money because I knew where his expertise could create leverage. The $40,000 mistake happened later because I had not validated whether the requirement we were designing around actually mattered enough to customers.
That distinction is important. Before you spend serious money, talk to customers, validate the requirement, understand what question you are trying to answer, define the deliverable, ask for referrals, compare multiple scopes, and stay involved once the work begins.
The $40,000 was not wasted because the engineers were bad. It was expensive because I asked good engineers to solve the wrong problem.