Monday, February 2, 2009

Scalable Business Models

When a business model has the potential to generate growth in revenues significantly faster than its cost base, the business model is scalable. As growing revenues increases the operating margin, scalable business models have the potential for earning very high profits. The key to scalable business models is to have small Costs Of Goods Sold (COGS), and to get a demand driving revenues up.

Cost Of Goods Sold
COGS is the cost for each product or service, plus additional costs necessary to get the product into inventory and ready for sale, including shipping and handling, and is a variable cost associated with each unit sold. With low COGS, the contribution margin has the potential to increase dramatically relative to the fixed cost base, resulting in high profits. A product business with outsourced manufacturing can supply thousands of units with the same staff as it might supply hundreds of units.

A less scalable business model
When selling services per hour growth in revenues totally rely on personal output of staff, and there is a linear relationship between revenues and cost base. Growth is depending on the acquisition and retention of highly qualified people delivering the services, and those are not always that easy to find, even though there might be a high demand for the services.

Selling Software
A common example of a scalable business model is selling software. The development costs are high but once ready for launch, the costs for each new copy of the software is very low. When packaged and distributed in physical copies, software still has a COGS per unit.

The Internet enables zero COGS
Internet platforms with user-generated content, viral or word-of mouth marketing, and close to zero in costs for each new user has created very scalable and profitable businesses. An interesting example is Craiglist, a company generating estimated $100 M with less than 30 employees, having an operating margin of 90% . Another interesting example of scalable businesses is the rapid growth of companies producing virtual goods for online games and platforms. San-Francisco based Zynga apparently turned in 2008 revenues near $50M.

Designing a scalable business model is one thing...
It is easy to design scalable business models, especially using the Internet. It is not as easy to create demand and generate high growth in revenues. Perhaps the Internet bubble in 2000 was due to the fact that people only understood the first sentence?


  1. Thank you Anders for both an interesting post and a great effort in creating a valuable database for business models.

    Scalable business models obviously are the wet dream of many new software startups. However, something that I think is increasingly becoming a challenge subsequent to major successes, such as Craigslist, is certainly as you point out - creating a demand. Many of the highly successful internet applications (built on user generated content or participation) seem to rely to a large extent on attracting a large enough user base to make them profitable in the first place only to then evolve into something much more scalable. Many of these are usually preceded by a number of less succesful but similar examples (and sometimes also followed by some), e.g. Swedish Entropia which was less successful than Second Life (not to mention the promise from Swedish that still remains dormant).

    I guess that it is not particularly controversial nowadays to say that the majority of a technology/business model's value relates to the amount of users (alternatively customers). So it seems likely that the more a business model can construct and generate network effects among its user base the more the opportunities will arise for the business model to become scalable when customer demand augments (likely without large COGS increases).

    Structurally creating a sound framework where network effects are enough to reach the tipping point where "everyone" is starting to signing up (e.g. all of a sudden it seemed as everyone started using Flickr/Ebay/Facebook/Twitter on a daily basis) therefore remains a major challenge in order to build a scalable business model in my opinion.

    It will be interesting to see whether any reliable frameworks for how to structurally build foundations for generating network effects will be found/created in the future...

  2. With internet products and services the business model is not as important when the service or product is launched. It doesn't matter if you have the smartest most unique business model if you don't have any users. Therefore the most important thing to focus on is the acquisition of users. Then measuring what they are using your service to do (might not be the same as you originally thought - it rarely is), and from that develop a business model that meets the demands of the users.

  3. @Tobias
    Thanks for your input! As you point out successful internet applications are often preceded by a number of less successful but similar examples attracting the early adopters but not the early and late majority. There seems to be a tipping point when everyone not only start to use but also recommend a product or service for friends and colleagues. Just as P&G, Disney and Pfizer have found a recipe to repeatedly launch new billion dollar brands, there will probably be companies finding frameworks on how to repeatedly launch new successful Internet products and services.

  4. @Anders
    You have a point. However, I don't agree with your view on business models for Internet products and services, but I understand what you mean. Smaller initiatives without much investments, or perhaps VC funded start-ups getting funding before the financial crisis, could have the luxury of focusing on the acquisition of users, without a viable business model. But I believe the time of getting users without a working business model, with the hope of finding a business model or to be acquired by Google, is over - at least for this time. Also setting up services where you connect with different stakeholders such as Spotify with music, or with technology trading, requires an early business model thinking to even get the pieces you need to launch your Internet products or services.

    I totally agree with you that you have to be flexible in your choice of business model, and closely watch and measure how users are using the services and as you point out, sometimes users come up with new ways to use a service you did not think of. I also know that you have launched several successful Internet initiatives. Perhaps you have the business models you will use in your head, even though you don't start with the question "how to make money" but "what killer app would I recommend to everyone I know"...

  5. So you seem to focus on Internet companies only when looking at scalable biz. models. But one could argue that a hotel or cruiseship company is similar: one the hotel is or vessel is ready you have to fill it. With an occupancy of 25-35% you may cover your expenses but you start making a lot of money when you make that 70-100% as your operation cost hardly increase. The trick is - like Internet based companies - to get visitors and of course offer compelling value.

  6. Thanks for your comment Thorwald! You are of course correct in that not only internet companies can have scalable business models.

    However, it is important to not mix profitability and scalability. Take the example of an expert proving consultancy by the hour. Even though the services might be highly popular, highly priced and very profitable, if it’s difficult to train someone else it might be difficult to scale up.

    In the case with cruise ships perhaps it is difficult to find a cruise ship as nice as your own the day you have more passengers than seats? Or as good ship men? Or perhaps you have a small reef that is very popular that can only hold so many guests. Etc.