When it comes to quality planning, hardware startups tend to spend most of their time working on ensuring that a product will work as promised however many teams do not spend significant time addressing the subject of how the product will not work as promised. Returns, return logistics, and possible recalls can be financially devastating especially in the early life of a product. What may appear as a small change in warranty rates can have significant impact on a company’s bottomline and financial viability. For example, if a company assumes a $250.00 total cost, $50.00 margin and sales of 100,000 units per year, a change from a 5% to a 7.5% warranty rate can decrease profits by 17% and increase working capital by $50,000. A recall or major quality failure can easily bankrupt a company.
Avoiding increases in warranties and instances of recalls involves identifying and assessing hundreds of “what-if” scenarios including:
- How the user might use/misuse the product
- What range of environmental conditions the product will be exposed to
- What changes in assembly, material or process control could have a negative impact
- How parts may wear, fatigue or degrade over time
All of these factors can and should be included in the design process as early on as possible. Teams often depend too heavily on the DVT and PVT steps late in pilot production to “shake and bake” out problems or to find issues with usage. Finding reliability and robustness issues late in production can involve expensive and time-consuming redesign. It is far better to understand and build the requirements into the early design concepts. There are many tools teams can use to capture and prioritize requirements at an earlier stage including:
- Failure Modes and Effects Analysis
- Risk Management
- Specification Documents
- Quality Test Documents
- Fault Tree Analysis
- User Testing
Each of these tools can help teams to identify likely sources of failure. In addition, teams can gain great insights into issues by looking at publicly available data about how other products fail. If you don’t use this information, your competitor will.
When evaluating the likely quality issues with the purpose of developing a quality and life testing plan for a new product, there are three primary sources of data we refer to in order to gain an understanding of the typical failures and/or things customers are sensitive to:
I. Prior generation of the product
If your team is creating a second generation or updated version of an existing product, reading through the prior generation warranty claims can provide great insights for the new team. We recommend reading the raw data files rather than simply looking at the summaries and trends. There is significant insight to be generated by reading the customer comments and seeing the root cause analysis reports.
II. Amazon / Online reviews
There is a plethora of social media and reviews that can provide insights to what the customer cares about. Reading all of the one and two star reviews as well as the top rated can identify issues that are likely to be hot button issues for customers. In the space of hardware startups, often there is no direct comparison for a product but evaluating proxies can still provide insights. For example, if a device has a camera, looking at the typical complaints can highlight what the customer cares about. Looking at dedicated cameras, the complaints focused on battery power, image processing time, expectations of features similar to an iPhone and ergonomics. If a device has a tablet form factor, those issues included the unit feeling cheap (note some companies add weight to their products to give it the impression of quality), speed, LCD quality, light leakage, application compatibility, and wi-fi connectivity.
III. CPSC (and other government recall databases)
The Consumer Protection Safety Commission has a comprehensive database of all US recalls going back several decades. In addition, there are similar databases for the NTSB and FDA. Reading them can provide insights into the types of failures that can lead to a recall. Recalls are made for products for a number of reason that can include:
- Pinch or severed digits or limbs
- Chance of falling or inadvertent stopping on bikes or vehicle
- Choking or laceration from swallowing products
- Strangulation and entrapment.
- Objects falling on people
- Shock, burns or fire
- Falling or crushing objects
- Failure of a critical safety device
- Furniture giving way underneath
- Not following lead or other contaminant testing
For example, of the 300+ recalls that were issued in the last 12 months the following were some of the major root causes:
- Loosening of joints/connection
- Small parts or magnets swallowed by children
- Not following or adhering to federal safety standards.
- Pinch, cut or severing risks for fingers
- Breaking/cracking or other failure
- Battery failures
- Excess material or insufficient material
- Small pieces or magnets falling off
There are many sources of available historical data that can be used to brainstorm the likely failures. Learning from the mistakes of others rather than relying only on your own understanding of the product may just save you millions.