Skip to content

General, Product Selection

REV: 2020-12-13

Situation: You are going to be making a lot of important and costly purchasing decisions.

Intent: You want to learn-by-preparing the general research and decision-making process of committing to specific parts, tools, materials, and systems.

Trusted Company

A trusted company is one in which you have confidence that their products and practices will meet your expectations for good quality.

Your trust in a company grows through research and experience. For example...

  • You are impressed by the product's advertised features.
  • You like what you see on the product datasheet.
  • You validate that their product has been tested and certified for relevant standards at a national lab.
  • You hear a recommendation from someone you trust.
  • You read a review from someone who seems to know what they're writing about.
  • You see successful companies using the product.
  • You try the product and you like it.
  • You've been using the product for the last 10 years and you've loved it.

Trust is variable. The less you trust a company, the more research you'll feel you need to do in order to reach your confidence threshold. The more you trust a company, the more quickly you can commit to their products because you're confident they've done the due diligence for you.

Trust can get deep. With normal trust in company A, their products are expected to be good in a balanced way. With deep trust in company B, their products are expected to be good in a way that is optimized to what you need and are really looking for. For instance, suppose you are big on durability and portability; you'll like company X who makes their products with hard-anodized aluminum in the pursuit of similar design priorities.

Trust has objective limits. Sometimes a company disappoints you but they might actually have done the best that was possible with current technology. For example, all enterprise hard drives (even those with 1.5M+ MTBF) have an annualized failure rate of 1% or more because that is the practical reliability-yield limit of current hard-drive-QA technology. It's important to know these limits so that you can prepare better (ex. use redundant drives to lower data-risk far below 1%) and protect your trust with good companies that still deserve it.

Information Accuracy

Datasheets can contain errors. For example, we noticed a compliance declaration mistake on a fuse datasheet by cross-referencing with their substance declaration (PCS); we emailed the company and they fixed the data within a week (the PCS was correct, as expected).

Datasheets can contain misleading labels. For example, the RoHS3 directive (EU) normally requires a lead content of less than 0.1% by mass, but application-specific exemptions allow compliant parts to have up to 4% or even 85%. Similarly, "Pb-free" in datasheets for high-power electronic parts usually applies only to the external finish.

Standards can be misleading. For example, MIL-STD-810 is a flexible standard that allows companies to pick-and-choose which test methods they wish to apply to their product; a company advertising in bad-faith might claim that their product has a MIL-STD-810 rating while having conducted no testing at all!

New Products

Generally, you want new products. They build on the lessons of previous generations, leverage key technology upgrades, and overall can represent the latest and best of what a company has to offer.

That said, new products can be overpriced and marginally-better (or actually worse). A new product from company X might be better than their previous model but way behind the industry frontier. A new product might have problems; the more changes, the more unproven risks. The risk of any new product is especially severe for its long-life reliability because the product simply cannot have been tested for long enough in both real and test conditions (new-product lifetime specs are usually based on projections from accelerated test data).

Sometimes, new products have started moving in the wrong direction in a misguided pursuit of improvement. For example, some websites have become bloated where they are hardly better while loading much more slowly and losing key workflows that existing users have relied on for years. We know of a distributor that did not solve the slow-loading and key-workflow problems of their new website even after a year of A/B testing- and forced customers over- and had to immediately restore the old website due to complaints!

Sometimes, new products improve in low-priority areas instead of retaining or resolving more valuable opportunities and problems. For example, mobile phones already reached a practical thickness (10 to 15mm) that offers a good balance of carry-compactness and bend-durability, but some companies continued to focus their efforts on reducing thickness with diminishing returns, instead of power efficiency, thermal performance, and known failure points. Similarly, the smart-device pattern over the last decade has plummeted reliability where the focus on barely-used advanced-features has overshadowed the importance of long-life, stability, and easy-UI.

Specific Availability

It's important to make sure that you can:

  1. Get the product from a dependable source (low-friction ordering and fast lead-time/shipping).
  2. Get the exact product you are expecting.

Checking availability starts by understanding, customizing, and establishing the full ordering ID of the part that you want.

Generally, you can source from a distributor, direct from the manufacturer (the primary-company managing the product), or a third-party.

Sourcing from an official distributor is usually lowest-friction for ordering with balanced RMA/warranty coverage (ex. quick DOA replacement), but beware volatile mark-ups. Distributors with large revolving catalogs occasionally misprice and miscategorize new products.

Sourcing directly from the MFG usually provides the best bulk pricing and the most recent batches, though communication and lead-time can be slow. This is particularly useful for products that expire or involve non-internet updates. For example, we've had to order an industrial tool directly from the MFG while specifying that we want the latest firmware, in order to avoid a known issue that users were experiencing with old stock.

Sourcing from a third-party is volatile. Sometimes you can get a used product with plenty of lifetime at a great price and fast shipping, plus valuable advice/support. Maybe even new-condition stock that they want to liquidate sooner than later. Sometimes you can get the wrong product or a broken product (partially or fully) at a terrible price with shipping problems and zero return/warranty coverage.

Product Evaluation

When you research or experience a product, there is a certain logic that your mind applies to intuit a level of quality, whether high or low. Here we aim to think more consciously and critically about that appraisal process.


First, you must understand the product and its potential. Ideally, the company provides all relevant specifications in the datasheet. In practice, you either take extra effort to fill in the gaps or accept imperfect information (often the case with complex products or hard-to-test details). Ideally, you are able to use or simulate using the product in the real-world; consider the possibilities, limits, and risks of its lifecycle.

Second, you must map your understanding-of-the-product to the discrete-breakdown-of-good. Ideally, you are able to consider all relevant goals and form a list of qualitative/quantitative ratings for each goal against the product specifications/experiences. For example, consider the noise rating, in decibels, at normal user proximity; then judge that on the spectrum of good and bad.

Third and finally, you must determine the absolute and comparative quality of the combined map then decide the future. How good is the product independently; how close does it get to the ideal experience? How much better or worse is the product compared to alternatives? Proceed with the product or wait/choose an upgrade/variation/alternative?


Some goals are pretty universal. You want safety, high-performance, reliability, ease-of-use, ease-of-cleaning, energy-efficiency, quietness, compactness, light-weight, and more.

Some goals are more application-specific with various overlaps. For example, delicious-taste is specific to food. Bad-smell is specific to both expired-food-detection and gas-leak-detection. Moisture-sensitivity-level-1 is specific to pre-assembly electronic parts under J‑STD‑020.