In Greek mythology, Prometheus stole the secret of fire from the gods and gave it to man. In retribution, Zeus chained him to the side of a mountain where a sacred bird pecked out his liver on a daily basis for eternity.

That might sound just like daily life, circa August 2011, for many of you in the solar business. Solar has made tremendous strides in providing an economic alternative to fossil fuels. Solar modules have dropped from $21 a watt to manufacture in the early '80s to 75 cents a watt today. Solar thermal developers say they will be able to deliver solar power 24 hours a day via salt storage and generate power for 7.5 cents a watt by 2016, less than the cost of coal and new nuclear plants. Double-digit growth in unit shipments and revenue has become the norm for many companies.

Yet it sometimes feels like an involuntary organ harvest. SunPower grew revenue to $592.3 million in the second quarter, a 30 percent hike, and reported a net loss of $147.9 million. Trina Solar had record sales, along with slower demand and high industry inventories. Evergreen filed for bankruptcy, Solon closed a plant, and BrightSource Energy seems to face an endless onslaught of NIMBY suits.

Name your cause: high capital costs, low barriers to entry, commoditized product designs, changing local policies. To top it off, solar has to compete against incumbents like coal, nuclear, and natural gas that foist some of their costs -- health hazards, environmental cleanup, cheap royalties on public lands -- onto taxpayers. 

And with a 25-year lifetime on most products, the replacement cycle in solar does not exist.

Wouldn't tying balloon animals or betting on football cards be a more stable way to make a living?

The PC market isn’t a bad analogy for what's happening in solar. That's the good news. Lots of peple made money in PCs. Yes, once-mighty companies like Packard Bell, AST, Commodore, Be and Zenith PC shipments have swooned and price cuts regularly eviscerate profits. But the industry has has also spawned titans like Dell and Apple. Contract manufacturers like Flextronics can live on 3 percent operating margins, typically worse than solar, yet generate hundreds of millions in net income.

Weirdly, the PC industry has experienced only two years--1985 and 2001--where unit shipments actually shrunk. Will tablets kill them this time, or are tablets just the latest reincarnation?

Quit whining. Death could be worse.

So what can a solar company do to thrive? I covered IT for years. Here are some ideas from that industry.  

1. Service. revived enterprise software with the as-a-service concept. Power plants do the same for alternative energy. Once the power plant is built, a manufacturer/developer is insulated from further price declines and can count on a revenue stream that will last for decades. (Alternatively, they can sell the power plant to someone for a present value figure and that third person can collect the monthly bill.) Even better, many customers -- large utilities -- won’t go out of business and can’t repudiate a contract without incurring the wrath of state agencies and public interest groups.

Developer/manufacturers can also hone a specialization. Recurrent Energy, now part of Sharp Solar, concentrates on urban plots and disdains desert solar and commercial rooftops.

Utility contracts don’t guarantee success. Optisolar raised over $330 million and signed two mega-deals with utilities before it cratered. Suntech tried to develop projects through its Gemini joint venture, but floundered. Think of it: have you ever met a real estate developer who has not been bankrupt during portions of his or her adult life?

Nonetheless, selling power and not panels seems like a good bet.

2. The Technological Niche. The challenge here is finding one that will flourish for a few years before technical know-how becomes widespread and commodity pricing takes over. LCD TVs were incredibly profitable until Vizio, Westinghouse and other companies with lower overhead figured out the secret sauce.

Your original, ground-breaking technology also has to work with conventional products. Evergreen Solar messed up with a kooky-sized solar cell.

Electronics seems to be one of the more promising areas for technological differentiation. Enphase Energy has succeeded with microinverters. Expect to see AC panels with SolarBridge inverters and more arrays equipped with DC maximizers. Most solar deployments still revolve around centralized inverters, so the market opportunity remains vast and no one really knows yet what combination of next-gen chips and software will work best. China is a bit behind here too, although it could catch up quickly.

Thin wafers show promise, as well. Crystal Solar says it can reduce the thickness of solar wafers from around 160 microns to 20 to 50 microns. Cells made using this process could produce a watt with a gram of silicon. Others in the wafer market include 1366 Technologies, Twin Creeks Technologies, Alta Devices and Astrowatt. Added bonus: most of the epitaxial expertise remains in Europe, Japan and North America. It may not get replicated quickly.

3. Ads and Word of Mouth. Some scoff at Yingli’s sponsorship of the World Cup or SunPower’s radio ads. But, again, look at electronics. Do you really think Apple sold the iPhone on price and performance? Did cops and insurance agents really have a latent need to play Fruit Ninja and look up restaurants before the iPhone? The relentless barrage of ads made the device into a cultural signpost.

HTC has become a top brand in phones. Less than five years ago, it was a faceless contract manufacturer. Like Acer before it, HTC exploited marketing and technology alliances to make a name for itself.

Advertising and branding will never be as important in solar as it is in other industries. Solar panels sit on your roof for decades and chances are you won’t go up there to visit them. Still, ads embed concepts and name brands into your head. They work. The industry has eschewed ads for far too long to its detriment.

4. The Dell Bear Hug. Many attribute the stunning rise of Dell to lower cost. Dell cut costs by selling direct, the theory goes. But if you talk to Michael Dell, direct sales were really about establishing trust with buyers. Customers felt they were being listened to. As part of that ethos, Dell himself used to (and still might) pull anonymous shifts on the customer support desk to get direct feedback from buyers.

Dell, in fact, often had higher average selling prices than competitors during its boom years in the late '90s and early '00s. Go figure.

Sungevity, the residential installer, is putting quite a bit of emphasis on customer comfort and relationships. I went through their sales process once to test it out: it was like dealing with a hotel chain or Starbucks. Yes, the Sungevity representative was trying to sell me $20,000 worth of equipment and so had a vested interest in being nice to be, but the process felt good. Everyone likes being catered to.

Grape Solar, meanwhile, is working with retailers like Home Depot, Costco and Amazon. Will these stores' brand names and consumer trust rub off on Grape? It’s hard to say, but you have to salute the effort. Specifications and economics remain crucial to a solar sale, but warm and fuzzy can seal the deal.

5. Quit Being Embarrassed About Government Sales. The U.S. Navy wants to convert half of its bases into net-zero-energy enterprises by 2020. The U.S. Army will spend $7.1 billion on renewable energy in ten years. School districts have received grants for solar. The federal government accounts for two percent of all energy consumption in the U.S.

Tea Partiers can jump up and down and cite excerpts from The Fountainhead all they want, but government support for industry goes back to the Erie Canal. The Department of Defense was the primary customer in the semiconductor industry for several years, and today, companies like IBM and Oracle/Sun still generate significant profits from governments.

6. Partner With Someone From Asia. Samsung has laid plans to become a top solar provider by 2015. Panasonic became a big player in solar when it bought Sanyo and has said it wants to be number one in solar electronics by 2018. TSMC invested in Stion and TFG Radiant rescued CIGS maker Ascent. Innovalight, now part of DuPont, sells virtually all of its technology to China, India and South Korea. AUO, the spawn of Acer, has deals with SolarBridge and SunPower.

Power consumption and air pollution are both on the rise in mainland Asia and mass manufacturing has been honed to an art. Asia will become both the largest market and producer in solar. You might as well partner now.

7. Target Marketing. PCs started as big boxes. Then came workstations, notebooks, blade servers, smart phones and tablets. Variety rules.

Solar manufacturers started to move away from monolithic manufacturing and product design a few years ago. Suntech has panels (and frames) geared toward utility projects that differ from their home panels. SunPower has the 1-megawatt power plant in a box.

Concentrators and hybrid PV/solar hot water systems like those from ZenithSolar could prod variety further along. Concentrator systems also vary wildly.

8. The Applied Way. It may sound strange to hold Applied Materials up as an example to follow in solar. Last year, the company embarrassingly bowed out of making equipment for amorphous silicon. But the company remains the leader in the semiconductor equipment market.

Many thin film companies like MiaSolé are actually equipment companies disguised as panel makers. Most of the intellectual property is locked up in the equipment, not the panel itself. By repositioning oneself as an equipment vendor, a panel maker could sell itself off to an existing giant and sell to its former competitors.

9. Think Logistically. In the 80s and most of the 90s, companies didn't manage their own computing needs. Instead, they hired integrators and resellers to do it for them. The once might reseller channel has shrunk since then, but it was a strong business for years. Solar customers will always want third parties: to install panels, fill out the asinine paperwork, repair and replace panels. Racking, delivery and maintenance will always be needed. Power plant developers might sell these services, but there's a large chance that many of them will outsource these tasks for at least a number of years.

Good luck.