At Home Depot recently I saw that Cree dropped its price for a 60W bulb to $4.97. But Philips had dropped its price to $3.97! It got me to thinking.
In the early 70s, when many digital IC companies were emerging and a price war erupted (as the commercial general-purpose computer industry blossomed), TI bombed the price on its new 7400 TTL series for several years. It sold it at a tremendous loss, at first, and severely wounded its competition. But TI, with volume and economies of scale, actually ended up dropping its prices dramatically and went on to create a giant, profitable TTL-logic business, with TTL perhaps being the most popular digital IC component technology ever. TI did a similar thing on a smaller scale around that time with its TO-220 bipolar power transistors.
Lightbulb manufacturing efficiencies and cost reduction do not at all scale the way a) wafer manufacturing and b) chip packaging do (where in cases a and b small-volume production is astronomically more costly than high-volume production). So one will not see LED bulbs sold profitably at 30 cents for decades, if ever.
Could it be that Philips will ultimately crush Cree's lightbulb efforts? Let's dissect the question. We'll first note that the prices I saw reflect the rebates given in Massachusetts. But things are relative. If in another state the un-rebated Home Depot prices are 30% to 50% higher, no matter, as long as the Cree/Philips ratio is the same.
It is generally said in real estate circles that value is about "location, location, location." Well, in retailing, it is all about "shelf space, shelf space, shelf space" in a store, but more importantly, as an overall number nationally. Proctor and Gamble, GE, Philips, Gillette, Nabisco, and the like, can snap their fingers and know they can have a new product on the shelf in thousands of locations, without the necessity of negotiation, price-cutting, or the ring-kissing of buyers.
They capitalize on decades-long, mutually profitable relationships with those food or non-food commodity retailers. It defies common sense that Cree, with only a handful of single-technology bulbs, can compete successfully with the likes of GE, Sylvania, Philips, and Feit, all of which offer hundreds of bulb types of several technologies and sell them in every nook and cranny of the US, from the corner hardware store to the big box stores. Each of them has a hundred or a thousand times more shelf space than Cree, which really only has a retail presence at Home Depot, where newly hired sales managers hear the siren and go to make their sales quota bonus. (Is LSG the poster boy?)
For decades GE, for example, has sold hundreds of types of bulbs at prices far above what you can get at Walmart or today at HD or online. Brand recognition, shelf ownership, and ubiquitous distribution are powerful forces.
Cree cannot remotely compete with Philips if it decides to own this market. Philips, like TI before it, can sell its bulbs for a buck for three years and not miss a beat, and when it comes back and raises the price to $3 (or whatever), Cree will not be in the business.
In 2006, when LED chips were 90% of the bill of materials for a China-made bulb of the necessary lumens, it might have made sense for chipmaker to make a complete bulb. But with increased efficacy and a 10:1 drop in dollars per Watt since then, it's a totally different game.
All LED A-style or PAR-type bulbs have now gravitated to an almost identical buck regulator driver topology (or a minimally different transformer isolated version). PFC? EMI filtering? Ripple filtering? A couple of cheap components to facilitate proper dimming tricks? This is now old hat … all now Driver Design 101 and not a nickel's relevant difference from one to another. All the driver ICs, power transistors, magnetics, filter caps, etc. are made in China. (Don't think for a minute that any one of these guys gets his parts cheap from some high-quality vendor the other guys don't know about. That's like suggesting that Gulf's gas perhaps comes from "better" oil wells than Exxon's.)
The enclosure, shell, heat sink, etc. … these are cheap molded or cast parts made in China for diddly. The same goes for LED PCB or MCB: nary a difference among the crowd. Assembly cost? Everybody is so similar that such cost differences are not worth talking about. They are all made in the same types of shops in China.
What's left? LED chips. So maybe Cree makes its own and can have a 3% lower total BOM cost. But what if one of the other guys can get his plastic or metal made cheaper and get his driver made cheaper because he has more clout? What does Cree really bring to the party in terms of IP or manufacturing know-how if it farms out virtually everything? What's its endgame here?
It takes years and years and years -- decades -- to build up powerful national distribution channels that can guarantee shelf space. Sure, you can "break into," or buy your way into, the shelf space club if you offer exceptional differentiation, choice, value, etc. to the point where distributors and store managers just can't say no. It's pretty obvious that Cree has not been able to do that anywhere but Home Depot. It has never had a consumer sales infrastructure. Procter and Gamble, GE, Philips, and Sylvania are the ivy in that league.
GE can decide tomorrow to get into the business big-time -- not just at Walmart. It can have the driver PCB and housing made just as cheaply as Cree, and have it assembled at the same cost or less. It has the clout to buy LED chips right out of China (now that certain threats have been addressed and Asian chipmakers can sell into US-market products). But unlike Cree, GE or even Sylvania does not have to buy its way into Home Depot.
Sylvania has been technologically dysfunctional for many years, but I give it credit for being the master at using its distribution and brand recognition. I have had occasion to meet with its headquarters marketing people from time to time. Their charter was to "find stuff" out there they can import or private-label in some way. They can slap the Sylvania label on it because they know they have longtime loyal customers, like Lowe's, who will stock it if they offer it, just because it has the Sylvania "trusted" moniker. Cree hardly has a long-loyal following in the light business.
The 50,000 non-big-box retail outlets and supermarkets are all capable of selling a few GE, Philips, or Sylvania products at rather high prices, because of the selection, convenience, and name. Feit is the rascal-relative in the bunch. It does have very wide distribution and a broad product line. Surprisingly, it is now doing some smart things with its LED bulb designs.
I think Cree is in a race it cannot win. If it had a far better mousetrap, maybe, but that is hardly the case, and at this stage maybe not even possible. It is interesting that around 2006, rumors were being constantly published about GE acquiring Cree, or Cree seeking to be acquired by GE. Needless to say, there would be no need today for GE to do that to meet its objectives. It recently acquired an LED driver company and an LED luminaire company. It is just getting started again.
As I said in my recent post and in the follow-on conversation, Cree is simply not a "lightbulb company." Aside from Home Depot, it seems to have almost no retail presence in the US. Philips, GE, and Feit lightbulb products of every conceivable type now probably collectively occupy 1,000 times the retail shelf space of Cree. It's only a matter of time before that reality foretells what label marking is likely to be on the LED bulbs being sold in 2020, after the dust has settled.