“Let our rigorous testing and reviews be your guidelines to A/V equipment – not marketing slogans”
Facebook Youtube Twitter instagram pinterest

The Product Development Process: An Insider's Prespective Part II

by June 16, 2013


In Part I of our Insider's Perspective on Product Development we took a tour of the inner workings of a typical AV manufacturer's development process of a new product concept.  In Part II, we will take you on a tour of the manufacturing process, discuss dealing with outside vendors, show you how pricing strategy/decisions are made and how to avoid last-minute catastrophes.

Editorial Note: All the examples, people, and occurrences that follow are generalized composites of the many events and individuals I’ve encountered over several decades. None of these stories should be construed by the reader as depicting any specific person or event.

OK, so we’ve somehow gotten through the strategic/Product Definition process and we’ve actually decided to make something. Now we’ve got to get it developed and engineered, we have to manufacture it, and we’ve got to price it and deliver it to the market. The fun is just beginning.

Everyone loves the look that the Industrial Design people came up with. But, the design is going to require some <<Gulp!!>> tooling.


“We have to tool that? ‘Tool’ is a four-letter word.”

Tooling. The world of manufacturing revolves around tooling. “Tooling” in consumer electronics is when a company needs to do an injection-molded plastic or cast aluminum part. Whether it’s a tweeter faceplate or a swoopy satellite cube housing or a grille cloth frame or a dimpled port tube, those are all examples of tooled parts.

Tooled parts are made by injecting molten plastic or pouring molten aluminum into a block of steel that has been hollowed out using a very precise computer-numerically-controlled (CNC) program, derived from a 3-D engineering file. Whenever you see a nicely molded, shaped piece of plastic or metal, chances are it has been formed by some sort of tooling process. Tooling is very expensive, which is why smaller manufacturers jokingly refer to the need to “tool” something as a four-letter word. Tooling is expensive, has a very long lead time, and you can’t easily change it if you decide to re-design the part later on.

Injection Molding

Injection Molding Machine

There are two basic types of tooling: Proprietary tooling and open tooling. 

Proprietary tooling is tooling that that particular manufacturer designs and pays for themselves, and then they own the tooling and retain the rights to it forever. There are many distinctive tweeter faceplates, woofer frames, or plastic satellite cubes which we associate with a particular manufacturer’s look and style, and that’s because they’ve done their own tooling. The manufacturer will often protect their tooling appearance with a design patent, in order to make their design immune to counterfeiting and unauthorized copying.

Good examples of proprietary, patented tooling are Boston Acoustics’ automotive Pro Series cast aluminum woofer baskets and Bose’s famous satellite cubes.

boston-driver     Bose Cube

Boston Acoustics Pro Series Woofer                 Bose Cubes

The other kind of tooling is what’s known as “open tooling.” Open tooling is tooling that a vendor of raw parts undertakes themselves in anticipation that many of its customers will order their parts, thus making their (the vendor’s) speculative investment in the tooling worthwhile in the long run. Examples of open tooling are the cast woofer baskets that many driver vendors make for their line of general distribution woofers, for example, or the round plastic frames for ceiling speakers offered by many Far East vendors to “brand name” speaker companies. You can look closely at the metal woofer baskets or round plastic ceiling mounting frames from many speaker companies and find that they are identical, absolutely identical. Nothing wrong with this; it’s common, accepted practice, and it allows manufacturers to offer quality goods at fair prices.

In addition to being able to make parts in almost any shape or color, tooling has the advantage of a very low per-part cost. Once a manufacturer pays for the tool itself, the parts can be “spit out” very quickly from the tool for a very low cost per part in large quantity. If the manufacturer can’t afford to pay for the tool in its entirety up front (often 10’s of thousands of dollars), part of the tool’s costs can be amortized into the product’s cost, usually over a set number of products. So if a tool is $30,000, the manufacturer might pay the tooling vendor $15,000 up front, then the remaining $15,000 is amortized into the next 5000 units. So, each product (up to 5000) will carry an additional $3.00 cost to pay for the tool.

Some companies will go to great lengths to avoid tooling expense. One well-known speaker company did not have the financial resources to do plastic-tooled baffle boards for its new line of flush-mount in-wall speakers. So they made the baffle boards out of MDF, painted them with several coats of black satin semi-gloss paint, and the baffles looked like a beautifully-finished heavy resin or some similar material. It was obvious they weren’t plastic because they were much heavier than plastic baffles would have been.

In a classic case of making lemonade out of lemons, the company touted its “Resonant-free MDF baffles, made from the same material as stand-alone audiophile box speakers, unlike conventional in-wall speakers from everyone else that use flimsy, resonant-prone, ‘ringy’ plastic baffles.”

This was great marketing touting a real acoustic benefit, but the “benefit” was brought about by the financial weakness of the company, not because of some brilliant pre-planned strategy.

Again, “outsiders” have no idea about how or why things like this come to be.

Deciding on Specifications

Ever wonder how a company arrives at its published specifications? There is so much gray area in specsmanship that it’s difficult to always do the right thing.

Let’s take just one specification for speakers: the crossover (x-o) frequency. This is a strange spec anyway, because it is of no actual use to the end user. It’s not like knowing the x-o will help them use the product in any way, unlike an amplifier’s power output or a speaker’s sensitivity or power handling capability. If those specs are accurate and representative of actual performance, then they can help the consumer choose and use the product.

But enthusiasts like to know the crossover frequencies anyway. However, x-o’s are not cut-and-dried. The electrical crossover circuit has a very specific, definitive characteristic and when you “run the voltage curves,” you’ll see exactly where the LP and HP curves “cross over” each other.

Drivers, on the other hand, are not so well behaved and predictable. They have mass, stored energy, inertia, resonances, etc. The electrical crossover circuitry may “tell” the woofer to cross over to the midrange at 500 Hz, but when you run the acoustic curves of the individual drivers through the crossover filter, the acoustic crossover between the woofer and the midrange might very well occur at a little over 600 Hz. And to top it all off, good crossover parts typically have a +/- 5% variance and even excellent drivers themselves are doing very well to have T-S parameters within +/- 10% of each other, so crossover frequency specs are just not that exact.


So here’s what the Project Engineer typically says to the Product Manager:

“Where do you want to say it crosses over?”

“If I say 550 Hz, can you live with that?”

“Yeah, that’ll work for me.”

I remember seeing the crossover spec for a popular speaker from one of the biggest speaker companies in the world specified at “2634 Hz.” I just laughed.

The U.S Manufacturer-Overseas Vendor Relationship: An Insider's Prespective Part II pg2

US-China GearsBoy, is this an important aspect of the whole product development process! Nowadays, there are virtually no all-in-one US-based manufacturers. There used to be, back in the 1950’s-60’s-70’s: Companies like Klipsch, AR, KLH, Advent, etc. actually used to manufacture almost all the parts that went into their finished product. They used to make their own drivers, wind their own coils, and some companies even had woodshops where they made their own cabinets.

Not anymore.

These days, companies may design things at their US corporate headquarters, but virtually everything is made overseas by a foreign vendor. And the foreign vendor often supplies a lot of the actual engineering as well. The US-based company will say to the vendor, “We’d like a such-and-such, with performance specs of so-and-so, what’ve you got hanging around? Can you send us a few for us to test? And while you’re at it, modify two of them, one to have specs of + X and one that’s – X. After we test them, we’ll let you know. Oh, tell us how much they cost, how long it’ll take to deliver, and what the minimum order is.”

The degree of domestic design/engineering vs. how much is done by the vendor varies from company to company, sometimes from product to product. For example, a US-based speaker company may do all its own speaker design in this country. But, if that company decides to add a line of electronics, they may very well decide to have the vendor do all the design work and source the product. The US company may add their own cosmetic touches and request a few special features, but the actual design/integration work is all done by the vendor.

The vendor quotes the price of the sub-assembly or finished product based on the quantity that the manufacturer estimates they’ll buy within a certain timeframe.

Overseas vendors vary from one extreme to the other in terms of their capabilities and size. Some of them are huge, with multi-acre “campuses” that have every operation imaginable: complete R&D labs, woodworking facilities with complete painting and finishing capabilities, tooling, electronics /PCB fabrication, final assembly, packaging, warehousing, etc. Soup to nuts. The workers live on the campus.

Oftentimes, a representative from the US company will visit the vendor to oversee the initial production and ensure everything’s going correctly. When they’re there, they may see their competitor’s products running down the assembly line right next to theirs and see their competitor’s goods stacked up in the warehouse! (They may even see someone they know from a rival US company: “Hey Al! How’s that new speaker of yours?”)

So much for the theory that a given line of goods is produced by an “exclusive, boutique” factory!

Unfortunately, just as often, the overseas vendor is a small, family-run shoestring operation. They farm out their tooling and PCB fabrication to 3rd-party suppliers, leaving the US manufacturer that much further removed from the process. Sometimes, you end up “getting what you get,” despite what you designed way back when. That’s just how it goes. It sure keeps the US marketing people busy, as they frantically re-write the literature to accommodate last minute changes in specs, finish, dimensions, etc.!

There are all sorts of dynamics and unpredictable interactions between the manufacturer and the overseas vendor that can affect the final pricing and all of these interactions play a huge role in the timing and final form/performance/features of the finished product.

Here are some of the things that really happen and again, “outsiders” have no idea about the amazing degree to which these occurrences influence the way the final product takes shape:

  • The vendor comes back in the final hours updating the pricing—always higher than the price the manufacturer has understood it to be all along. “Sorry—it’s an unavoidable materials availability issue. You’re lucky you’re getting the part at all.”
  • The manufacturer calls the vendor to tell them, “The project’s a go! We’re writing the PO now.” But it has been a year and a half since the initial price quote and the vendor assumed the project was dead!
  • Engineering says to the vendor, “You know that sample you sent us three years ago? It’s going to be perfect for this new project we’re doing. We’ll need 50 pieces for our prototype work and pre-production samples in 2 weeks. You’ve still got all the specs and can build these immediately, right?”
  • Mass production is scheduled for next week, but at the last minute, the head US-based engineer announces there needs to be a “minor” change that they forgot to tell the vendor about before the first parts were shipped.
  • Marketing “promised” the vendor the project would be selling about 1,000/month, but the opening order (including “pipeline fill,” which is a larger quantity than normal follow-up orders) will be 500, with another 250 needed in 6 months. “But the price and shipping schedule remains the same, right?”
  • The head US-based engineer gets fired, and the new guy needs to “make his mark” to show his authority and project ownership, so he changes all the spec requirements. However, he insists to the vendor that neither the schedule nor price will be affected, right?
  • The US company announces to their vendor, “We’re going to delay the release of the product until after the next show” [six months from now], after pushing the heck out of all their suppliers to start running production at full speed by the end of this month. Naturally, the US company expects their vendors to be completely unfazed, to hold the line on pricing and shipping, and to take all their future project requests just as seriously as they originally took this one.

If the US company-to-overseas vendor relationship gets really out of whack, to the point where neither side wants to deal with the other anymore and the entire project is in jeopardy (an all-too-frequent situation), then Dear Leader will get involved and, as they say, “Make the call” to the vendor’s US representatives. With any luck (and that’s all it is—blind luck—because Dear Leader usually knows absolutely nothing about the details and particulars of the project’s developmental history) he’ll be able to talk the vendor off the ledge and get them to agree to resume work on the project.  This is usually accomplished by virtue of his long personal relationship with the vendor’s US representative (swapping a few tired old war stories of times past when they worked together at another company) and with some vague promise of additional future work he’ll toss their way.

If this works, then it’s just another reason that Dear Leader will take credit for anything good that happens.


Dear Leader


The ‘Cost of Goods’ to ‘Selling Price’ Relationship

This is a good time to discuss this, because it’s probably the least understood aspect of professional product design and development that there is.

Companies have a formula for determining the sell price of a product based on its raw cost of goods. Usually it starts the other way around: Marketing/Sales comes up with their desired sell price and applying the formula in reverse, this tells a company how much the raw cost can be.

Lots of factors go into the formula:

  • Transportation/freight from overseas (either raw materials or finished goods)
  • Company overhead once those materials or goods come into the building in the US. Remember, you have to pay the rent (or mortgage), electricity, heat, taxes, phone, internet, buy computers and coffee and clean the carpets, pay salaries/insurance/benefits, petty expenses , etc. Every product or material brought in is assigned a percentage “overage” to cover this. Let’s say 20%. So a tweeter that has a cost of $20 is costed at $24 to cover its share of the U.S. operation’s overhead.
  • Dealer margin requirements (“home” speakers are typically 50% on retail “list price”; custom in-wall/in-ceiling speakers are typically 60% on retail.)
  • Sales commissions (reps, distributors, etc.)
  • Freight out (if the company “pre-pays” the freight on orders to the dealer, this has to be accounted for)
  • Fast-pay discounts (if a company’s payment terms to its dealers are 3% 30, net 31, then most dealers will end up taking the discount)
  • Bad debt allowance (some dealers go out of business and never pay)
  • Advertising allowance
  • Warranty repair allowance (this actually is required by law to be kept in a separate escrow account, unless the company can prove its warranty expenditures are less than x% annually)

And so on.

The ID [Internet Direct] cost model isn't much lower than the B&M [Brick & Mortar] one as some people may think.

All of this is loaded into a formula and a raw cost of goods of say, $37 will yield a projected sell price for a stand-alone box speaker of $225 ea. That same $37 raw cost would be a sell price of perhaps $260 for an in-wall product.

So you can see that the “ratio” is easily 6:1 or 7:1 retail to COGS. This is not a “rip off.” It’s simply what a company needs to stay in business, pay the electric bill, pay the receptionist, and replace a blown tweeter under warranty.

But when an outsider says, “They should have used the Seas tweeter instead of the Peerless, or Solen caps instead of common electrolytics,” if those raw materials cost differences were just $15, then your $225 ea. speaker is now $300 ea.—and very likely, unsellable and uncompetitive. If the company is forced to “eat” the cost increase and they sell 10,000 speakers that year x the $15 upcharge each, well, there goes $150,000 out the window. Maybe you can’t advertise 8 times a year in, hypothetically, Stereophile at $6000/page. Maybe you can’t give your prized engineer his well-deserved (and expected) raise. Maybe you can’t pay your best supplier on time and they cut—or suspend— your credit line.

The very sharp mind of a well-known company president once said that the best product decisions were made by answering one simple question: Will you sell one more or one less product because of using the more expensive part? Fancy internal wiring, expensive “boutique” capacitors, cast vs. stamped baskets, etc., all these should be evaluated on that basis, in his view. Resist giving away “invisible gifts,” as he called them.

Sometimes you do need the better stuff, and Marketing needs to make the appropriate big deal out of it in their advertising and literature. But very often, 95% of the performance is achieved with less expensive parts that allow the company to stay in business, pay its people and pay its bills, all while delivering a product that sounds really good at a price people can afford. It all gets back to Knowing Who Your Customer Is, like we said up top.

If you need fancy caps to sell the product into its market (including getting good reviews, if that’s critical to the product’s success), then by all means, do it. But far too often, the fancy-shamcy part isn’t needed, and when the company makes a foolish decision and includes it anyway, they hurt themselves for no reason and perhaps even go out of business—which helps no one.

Internet Direct Companies

But wait, you say. Now we have all these “Internet Direct” audio companies that cut out the so-called middle man and avoid all those costs like rep commissions, dealer mark-up requirements and the like. Surely, those companies are not working on a 6:1 Sell Price to COGS ratio.

Well, yes they are. Pretty darned close.

First, these Internet Direct companies are usually smaller than the big national brands, so they’re buying their parts and finished goods in smaller quantities, which means higher prices. Often, much higher.

Secondly, they aren’t shipping 2000 speakers or AV processors in an economically efficient manner to a retailer’s central warehouse. Instead, they’re shipping one set at a time to every individual customer. It’s not just the higher shipping costs either. Instead of one sales order, one invoice, one warehouse pick, and one trucking appointment being generated for a 2000-unit order, now you’ve got 2000 sales orders, 2000 invoices, 2000 warehouse picks and 2000 trucking appointments. Internet Direct companies usually have much more liberal return policies also, as an incentive to get people to take a chance on their products. Goods sold through conventional retailers don’t come back as often because customers can see, touch, feel, and hear the product before purchasing it (less surprises) and they can ask the store questions if they’re puzzled a bit either at the store or when they get it home. The operational and administrative costs for Internet Direct can be far, far higher, and those expenses must be built into a product’s cost structure if the company plans to stay in business, service its customers, and—hopefully—make a profit.

And last, the advertising costs are far higher too, because without a store to advertise and display the product, the only way people will ever know about you is if you advertise, advertise, advertise.

Bottom line: the cost structure for Internet Direct product is nowhere near as much lower as people might think.

We have several articles discussing ID vs Brick&Mortar listed below if you wish to read more on this topic:

ID vs Brick&Mortar Speaker Companies

ID Speaker Company Comparison

The 11th-hour Disaster

You’ve made it through the Product Def process, you’ve settled on an ID that everyone agrees to, the Project Engineer has finally gotten drivers he can work with and he’s voiced the speaker (along with the Product Manager) to wipe the floor with its chief rival in the marketplace.

The Sales Department is chomping at the bit and has already promised its biggest customers delivery in time for their next big catalogs.

 VP Sales

Mr. VP Sales

The only thing that’s left is for Dear Leader to give it one last listen and sign off on the crossover so Purchasing can order the PCBs and crossover components. Those are fairly short lead-time items, so if we have approval today, those pieces will arrive in about three or four weeks. Cabinets are already on the way. Drivers are done and ordered. Packaging is designed, drop-tested, and ordered. The owner’s manuals have been written (by the Product Manager) and those have a one-week turnaround from the printer, so they’re not the gating items, as the saying goes.

All systems go. Just listen one last time and sign off on the crossover.

Dear Leader comes into the listening lab. VP Sales, VP Marketing, VP Eng are all too scared to be there. Only the Product Manager and Project Engineer are there.

Dear Leader has very definite sound preferences and very definite ideas about design and crossovers.

He listens for a short time. He never says he likes anything, instead remaining poker-faced at all times.

Finally, after what seems like an eternity he says, “How big is the choke that you’re using on that woofer? 1.6 mH? That’ll cost a fortune.”

“But you’ve already listened before, you liked the sound of the speaker, and you’ve seen the schematic.”

“I never saw a 1.6 mH choke. Those are two bucks each. I never would have approved that.”

It’s pointless to argue. The Product Manager sees his life flashing before his eyes as his entire Jenga-esque project threatens to come tumbling down around him. Sales will hold him responsible for missing the Christmas season. Marketing will blame him for allowing a product to be developed that he should have known would never be approved. Mr. VP Eng knew everything all along, but he’s hiding in his office, trying to figure out how to be unaccountable.

Dear Leader bellows, “Try this .7 mH choke. I bet it sounds just as good.”


Dear Leader

It doesn’t, of course. The speaker—which was a real killer before and would have set the market back on its heels with its knockout sound—now sounds like a honky, over-midrangy mess.

All three (Dear Leader, the Product Manager and the Project Engineer) know that the .7 mH choke ruins the speaker. Ruins it. But now, the goal is to find a way to salvage the speaker’s sound, preserve the schedule and give Dear Leader some face-saving way out.

The Product Manager—using those Kissinger-like negotiating skills—says, “Hey, why don’t we try a 1.1 mH choke? That’s still about a dollar less than the 1.6 and the speaker will still sound fine."

We try it. It sounds good, not fabulous, but good. The Product Manager says, “Dear Leader, that was a good idea of yours to pull some cost out of that crossover. If you can just initial the change on the schematic, I’ll walk this down to Purchasing and we’ll get these things selling by the 4th quarter.”

“Fine.” He signs it.

No one ever tells him that the tweeter was only down 12 dB at resonance instead of the 18 dB at resonance that he usually insisted on.

A bullet dodged, a breath let out, and a product completed.

A million near-disasters, compromises, ego crises, vendor emergencies, credit crunches/cash shortages, and changes of heart along the way. Yet somehow, the product—whether it’s a speaker, a line of electronics or a niche accessory— is completed and brought to market.

That’s how it works.


About the author:
author portrait

Steve Feinstein is a long-time consumer electronics professional, with extended tenures at Panasonic, Boston Acoustics and Atlantic Technology. He has authored historical and educational articles for us as well as occasional loudspeaker reviews.

View full profile