As regular readers are aware, much of my consulting practice centers on working with software-based companies such as SaaS and infrastructure software. But I have substantial hardware market experience as well, so I also work with hardware companies. A debate around hardware vs software business models is something that you don’t often see a lot. But sometimes there is a choice between these two distinct business models (or a combination of both). So let’s take a look at the differences and relative advantages of each.
Hardware vs Software Capital Requirements
One of the larger differences is that software companies generally require much lower capital to reach profitability and continued growth. This is primarily because they don’t need to invest in expensive equipment. That means semiconductor development tools and semiconductor masks for chip companies. For hardware device OEMs, that is manufacturing plants/equipment, manufacturing engineering personnel, unfinished and finished goods inventory, etc. So except for startups aiming at huge, established markets and backed by substantial institutional capital, it’s much easier for typical startup software companies to adequately financially capitalize their businesses. Software startups can raise much smaller funding rounds, relative to their hardware and semiconductor counterparts.
Another important area where software companies generally have an advantage is in margins. Both in the area of typical gross margins as well as the potential for higher net margins. This is primarily due to the small or negligible cost-of-goods-sold enjoyed by most software companies. As a result, it is easier for software companies to reach profitability sooner. If a large enough market is found, sustain that profitability. Remember, throughout this article, I am speaking in “on average” terms. There are hardware businesses with excellent gross margins (dominant semiconductor companies come to mind) as well. But in general, this is an area where the advantage goes to software.
Software vs Hardware Pricing
The big difference between software vs hardware business models here also is related to costs. The major difference comes down to direct product cost. In the long run, this creates a pricing floor for anyone who would actually like to make a profit. Optimal pricing of both hardware and software should be based upon a value-based approach, with market segmentation as the key to success. However, in my consulting practice, I rarely find this pricing approach to be used rigorously. That’s true whether the company markets a software or hardware product.
In the hardware business, you tend to see a lot of simple pricing models that are cost-based. For software businesses, the negligible direct product cost often acts as the other end of the proverbial double-edged sword on pricing. In a competitive software market, you often see aggressive competitors who literally “give away” the initial product. Freeware or freemium models are used to build an installed base. Companies then rely on an upgrade revenue stream or advertising revenues to make a profit downstream. This can strain the profitability of the entire market segment and in severe circumstances can suck all the profit from the market.
You see this scenario most often started by weaker competitors or in market segments where switching costs for customers are high. While hardware pricing can be just as competitive in general, it is less likely for a weaker competitor in a hardware market to introduce a “zero-margin” program. This is because it is often tougher to hang onto a customer in the second generation, especially if the market has been commoditized. Plus, the market leader often has a gross margin advantage, unless the upstart has innovated on cost. That usually makes this tactic an ill-advised maneuver, other than an attention-getting, short-term promotion.
The Internet has created a major difference in distribution between software and hardware companies, where there was very little difference in the past. It has made direct distribution much more practical for small software companies. In many market segments, a simple download or a SaaS-based model is practical. Of course, hardware companies sell directly over the Internet as well. But the cost advantages aren’t the same due to shipping, inventory, etc. For those companies which aren’t direct-only, distribution is similar for hardware vs software companies. Traditional distribution through third parties tends to be similar. However, higher inventory costs are a burden that hardware companies need to manage closely. This is true both for in-process inventory as well as finished goods held by the channel.
Defensible Strategic Advantage
This is an area in which the hardware vs software debate includes both similarities and large differences. Both hardware and software companies value patents as a form of providing a sustainable competitive advantage. But in my opinion, the inherent malleability of software makes patent protection less useful in software than in hardware. It is easier to “find another way” of accomplishing the same end result when you are dealing strictly with software code. It’s also easier to segment markets in software. Creating a targeted, niche version of a software product for a specific segment allows a smaller competitor to nip at a market leader’s share without drawing their fire.
Although not impossible, it’s much harder for a small hardware company to differentiate itself this way. On the other hand, a hardware market leader that establishes itself and creates a large volume business gains important competitive advantages. Volume cost efficiencies and brand recognition are huge, defensible advantages. So I believe this point comes down to scale. In software markets, it’s easier for a small competitor to overcome the scale of larger competitors and develop a niche strategic advantage. While in hardware, the largest competitors can use scale to create the ultimate competitive advantage.
Hardware vs Software Localization Requirements
This is an area in which hardware companies normally have an advantage. They usually have simpler user interfaces, and sometimes utilize symbols extensively in their interfaces, greatly reducing translation requirements into local languages. The qualifier in this area is if a hardware company “gets it wrong” initially, it’s harder to change! Hardware companies do have to deal with some physical differences in industry or geographical standards, such as electrical. But these tend to stabilize over time and are often handled easily in one or just a few standard product version(s). This allows the selling of a product worldwide, minimizing inventory/distribution issues.
Conversely, software user interfaces are usually language intensive and more complex., Much more extensive user documentation is required. This requires many software companies to live with higher localization costs and longer lead times if they want to market worldwide. The exception to this is complex software sold to highly technical/educated users, where English is often used as the standard language on a worldwide basis.
Potential for Dominance
I’m using mostly empirical, anecdotal evidence here. It seems that there have been a lot more hardware companies that have dominated their respective businesses for long periods of time than software. For every Microsoft, it seems there are many more examples like Intel, Cisco, IBM, HP, Dell, etc. Hardware markets tend to commoditize more easily. But this occurs with standardization on a few leading brands in the long term. It’s harder to make money in the long run in hardware unless you are one of the top two or three players.
Hardware markets are also relatively larger in top-line revenue than the corresponding “large” software markets. This allows market leaders to more fully utilize their profit and cost advantages over competitors, by spreading marketing costs over larger revenue bases. So if your desire is to build a truly dominant company from scratch, the odds appear greater in hardware. Although you probably still get better odds by heading to Las Vegas and putting money on roulette red!
There are many more ways to contrast and compare hardware vs software business models for tech companies. But I will end it here. What other points would you add? As usual, post a comment or send me an email message to further the discussion.
Follow Phil Morettini and Morettini on Management via Twitter, Facebook, LinkedIn, RSS, or Subscribe to the Morettini on Management Newsletter hosted by LinkedIn. Contact Phil directly at email@example.com
David Locke says
Giving stuff away for free works only if it has switching costs, and you are giving it to geeks, or technical enthusiasts. If you are in the early adopter (economic buyer), custom application market; early mainstream market; late mainstream market; or phobic markets. Free doesn’t work.
Giving stuff away for free does not work for consumer goods, only technology goods.
There is probably a boundary on B2B vs B2C as well. I don’t know.
And, you only give stuff away if you can become the market dominator. If market share doesn’t matter, or results from last week’s promo spend, then giving it away, won’t work.
Carolyn McClain says
Great article. As you point out, generally software companies are more nimble, but they are more vulnerable as a consequence.
Another activity in the hardware world that consumes time and energy is supplier management. For every component in the product, you’ll want to have two suppliers at a minimum, and usually another in reserve. You need to qualify and requalify the suppliers and deal with their product cycles, which pretty much never correspond with your own.
Joe Schmelzer says
I have many comments in mind but will make just two:
1) When you talk about business in this way you can’t leave out Profitability. You touch on it with “Margins” – I would probably substitute “Profit” for “Margins” and spend more time in that area, in your analysis. Gross margin isn’t a great indicator in a software business, for reasons you mention. Profitability is. You may see two companies with essentially zero COGS but with radically different cost of sales or G&A, leading to very different profits.
2) The companies you mention, like IBM, HP, and CISCO are not “hardware” companies. In 1980, IBM was a hardware company. If you like at their revenue statements today you will see a blend of hardware, software, and services revenues. And substantial differences in gross margin and profits in each area.
Very very few international “product” companies that can exist purely on hardware.
Apple. Ultimate hardware (success) company, right? There are currently 820 jobs posted in software engineering and 559 in hardware engineering…
Phil Morettini says
Joe, A couple of comments to your two comments:
1) of course it’s true, this is obvious. I was contrasting software and hardware companies directly, focusing specifically on their differences. The difference from a CATEGORY point of view is the gross margins. Both hardware and software companies can have high or low fixed costs–it’s not dependent on a software or hardware business model.
2) Again, somewhat true but obvious. All hardware companies are heavily based upon software technologies. That’s not what this article was about. It was about the differences in profitability potential between the two models. The fact that there are a lot of software engineers in hardware companies isn’t relevant to the discussion. And yes, all of those companies are still hardware companies from a business model perspective.
Jennifer Kyte says
I agree with Phil in that HP, IBM, CISCO, etc. sell hardware (something physical) and focus on the business of selling that hardware. All hardware needs software in order to do something (embedded, applications, a company’s website, etc.), but they are definitely focused on selling the hardware and not the software. Software companies really only focus on selling non-physical items and not the hardware may be used on.
Microsoft is interesting though since they usually sold software, but now they have Xbox. I wonder if they use different business models on the different parts of their companies.
Joe Schmelzer says
1) Gross Margin % is not a metric to use when evaluating software business. Not applicable. It’s useless and misleading. The point you make about Capital Requirements was sufficient.
2) You don’t understand my point. “All hardware companies” are not just BASED ON software technologies, in the sense of needing some OS to operate properly, or something similar. Their VALUE is software. The iPod value proposition, for example, is as much about software as hardware. IBM’s value proposition is (1) server (2) software on the server and (3) services to put it all together for your enterprise. They’re not just selling hardware.
If you want to really draw an accurate comparison, you should be looking at companies like FOXCONN as your hardware company. They’re maybe the best example. Their value add is literally mass producing hardware. Gross margins probably below 5%, certainly below 10%. They don’t sell anything else.
Or McDonalds. They’re a hardware company.
These tech companies you mention are not hardware companies.
The reason that’s an important point is you’re trying to offer some guidance/analysis, and what you should be labeling your assessment is something like “Pure Software Business” (Google, Facebook, Uber, etc.) vs “Integrated Tech Company” (Apple, Cisco, IBM, Motorola…).
“Hardware vs. Software” as you lay it out, is sort of not really accurate enough, today. Again, in 1985, when a company like IBM really was a “hardware” company vs a company like Microsoft that was a “software” company, that comparison would have worked a lot better.
Phil Morettini says
Joe, Don’t want to continue to got point by point with you, because you don’t seem to get it. Couldn’t disagree with you more on most of your opinions. In fact, several of them are nonsensical. But everybody’s got one and you have a right to yours.
Joe Schmelzer says
OK. Just trying to help.
Reginald Knight says
In my opinion the focus of technology should be on understanding how to merge software and hardware in order to create or offer cost effective services. This is a balance of cost and cost effectiveness.