Often, when engineering a design, you may be faced with having to decide between using commercial off-the-shelf (COTS) parts versus designing new, custom parts for the design. Do you “make” or do you “buy”?
While this question appears straight forward, there are actually quite a few factors that need to be considered before making this determination. Part cost vs. purchase volume analysis is a good place to start, but part performance, quality, assembly issues and supply chain reliability may also come into play.
At Simplexity, clients often engage us for cost reduction and product simplification work. They need us to evaluate existing products or prototypes that have either more complexity than desired, or that seem to have excessive cost. Supply chain concerns are still very real and evaluating how to land on the most cost-effective, reliably manufacturable, best performing solution is a frequent challenge our engineers are asked to address.
How do you determine the best path forward? How can you maximize the value of your design? What factors come into play and are there basic rules-of-thumb to help guide these types of critical design decisions? In this blog, we’ll examine detailed working rationales, cost considerations, and use cases supporting make vs. buy options.
Defining “Make vs. Buy”
First, let’s clarify what is included in the Make vs. Buy discussion.
When we say “Make” that includes:
- Items Simplexity, or a design partner, designs
- Items created by the client
- Using custom designed product components for production tools
Example: Using the product PCA as part of the test system for the product.
When we say “Buy” that includes:
- COTS solutions.
- Custom/semi-custom solutions from specialist suppliers that do one niche thing very well.
Example: Electrical boxes or cable harnesses
Low vs. High Volume Production Cost Analysis
A useful starting point is to consider the cost of developing a solution versus purchasing a COTS component. When production volumes are low, it generally does not make sense to develop a custom component due to the non-recurring engineering (NRE) costs. At some production volume and part cost combination, the bill of materials (BOM) cost savings of a custom component could outweigh the cost of the design effort. In this case, from a costing perspective, it would make sense to build a custom component.
What is the cost situation?
- What is the COTS solution cost? This would be the part cost times the number of parts needed.
- What is the estimated development cost (hours x rate) to develop a custom solution?
- The development cost is the cost to design, build, test and verify that the custom solution meets all of the relevant requirements of the COTS solution. That typically requires a lot of hours.
- If a component is available and costs $1,000 and you need 1x of these components for a single machine: can you design, build, test and verify a solution in less than a day of engineering effort? (Spoiler: rarely)
- If a component costs $10, and you’re building 100k units in production, it is likely that the design and qualification effort will cost < $1M. In this example, a custom design could be warranted.
When to use a custom gear design
Factors Beyond Cost When Deciding on Make vs. Buy
Beyond the basic cost analysis, there are factors that should be considered when deciding whether or not to create a custom component. The part/component performance, risk to overall project cost and schedule, long term assurance of supply and product support all need careful consideration.
In some cases, there may be more nuanced factors that could shift the decision one way or another. Every situation is different. These factors should be reviewed and assessed for each specific situation to determine the proper approach for the situation.
The following are some questions to help inform and guide the make vs. buy process.
Does a COTS solution exist that meets the requirements?
- There may be situations where specific performance needs cannot be met by COTS solutions so a custom solution is the only option. This situation warrants a discussion with your design partner about what the cost/performance tradeoffs are for the system. If you can get close to the desired performance for a significantly lower cost, a reduced performance design might be a more market-viable option.
What is the development risk situation?
- This is very entwined with the cost, but the development of anything new entails some level of risk that things won’t work right. Getting to the root cause of a test failure, updating the design, rebuilding and retesting will likely be needed. This can easily increase the cost and delay the schedule significantly. If the decision hinges on getting it right the first time, you should seriously consider using a COTS solution.
What is the assurance of supply (AoS) of the component?
- If the component is to be used in a production environment (manufacturing, production test, etc.), there are long-term considerations that must be taken into account.
- What is the long-term availability of additional parts to scale production or replace failed components?
- Are there sub-components that may not be widely available in the future (ex: specific microprocessor)?
- For a COTS solution, what is the supplier’s committed product production life (often available upon request).
- For a custom solution, what is the plan to provide manufacturing data to replicate the design?
- Manufacturing documentation, processes, quality/test procedures
- Qualified suppliers
- For a custom solution, full design documentation must be created. If changes are needed, has enough information been provided so that someone skilled in the art can pick it up and make the required updates?
How is the component serviced?
- Is there sufficient documentation and are there tools to allow troubleshooting? Is there enough budget to create that for a custom solution?
- Is there technical support available to support troubleshooting efforts? How long would your design partner want to support a custom design?
- Is a warranty needed? Note: If a warranty is needed for a COTS solution, it is likely best for the company producing the product, rather than the design partner, to purchase the parts so that their name is on the contract for any warranty claims.
How will a custom design be verified?
- The component will need to be verified that it meets the defined requirements. Does the design partner have the ability to perform the testing required for verification?
- If not, the costs of verification tend to go up significantly, as failure report data quality tends to go down, and fix/test cycle time tends to increase. The location of initial testing needs to be factored into the verification testing cost and schedule.
- What happens when the first verification test fails? It is very common for some portion of a verification test to fail. If there is a failure, how easy is it to determine which part of the design failed?
Working with Informed Design Partners
Ultimately, the determination to make or buy parts can be greatly influenced by the design partner you engage to engineer your design. Partnering with a design engineering firm that can support the performance validation, quality management and ultimate manufacturability of the design well is critical. Leveraging cross-functional engineering teams or engaging a design partner that employs a full system development approach to design can often render design solutions that leverage COTS parts in innovative ways that a more siloed engineering approach may miss. Chances are, the more integrated the engineering team, the more options they will be able to present to help create the best performing, most cost-effective solutions. And, whether you ultimately make or buy, the right design engineering partner can give you the confidence that your design is the result of the most informed, smartest use of component selection possible.
If you are working to minimize the need for custom parts or just want to have a conversation to determine if a design evaluation might help you reduce cost or simplify your design, contact us. We’d love to hear from you.