Presentation on Typical Errors in Early Business Model Development – the case of intouch Group
Yesterday in the Biz Barcelona conference, an event which attracts thousands of innovators, entrepreneurs, and investors, I presented the topic “Renew your business model with the business model canvas: typical errors in early business model development” on behalf of La Salle Technova. These errors are based on my 20 years of experience working for startups (Fan Asylum, intouch Group which I give as an example below, and Vision Integrated Marketing), as an intrepreneur where I was finance manager for MCI Systemhouse (we bought Apple’s data center when they were nearly bankrupt in 1996 and filled the excess capacity with a $40 million business), as a consultant at DiamondCluster creating digital businesses for large companies, as a “founder” of startups like viajeria.com (and other projects based on affiliate marketing) and allysguesthouse.com (the quotes are because I’m not a billionaire yet), and as Associate Profesor at La Salle Barcelona, Director of the Máster en Digital Entrepreneurship, and mentor at La Salle Technova. In my current roles I see over 50 new business ideas a year at various stages of development. My primary focus is on early business model development and testing.
The basic conclusion is that there are bad ideas (I give several examples like Olestra, and an energy drink called Cocaine) but usually it’s a problem with the business model, which can be represented with the 9 areas from the business model canvas.
Joan Magretta in the seminal Harvard Business Review article “Why Business Models Matter” identifies 2 main tests: the narrative test (the story doesn’t make sense), and the numbers test (the profit and loss don’t add up). To illustrate a company that failed both tests I use the case of intouch Group, a company that produced an interactive music kiosk in the early 1990’s for sampling music in music stores. I was Controller in the early days, and sourced the over 200 parts for the product (before the Web was available). This is the first time I’ve told this story, and as far as I know the first time it’s been told with this type of analysis.
intouch built the largest music sample database in the world at that time with over 40,000 albums and up to 5 songs per album. The iStation kiosk was extremely advanced, with a touchscreen, Mac computer, Laser Disc player, and Cd-Rom jukebox. The proprietary software was quite advanced as well. Each member received an i-card (yes, we were “i”-ing things before Apple!) which served as their membership card. Listeners could scan the barcode of a CD to listen to the tracks on it. They could rate the songs, and receive recommendations based on the songs they liked. Data analysis was very sophisticated, with each scan of a cd, press on the screen, and rating registered in the database.
In spite of having an innovative and promising idea with a disruptive business model, producing over 500 kiosks by 1995, and raising over $20 million in venture financing (and nearly achieving an initial public offering), in the end the company failed to attract sufficient support from partners and customers to reach critical mass.
Here’s my version of the business model canvas for intouch Group:
The basic issues that led to intouch’s eventual closure were the following:
- Expensive hardware. The initial units of the kiosk cost about $50,000 to develop. Then the Internet exploded! Our partners began asking us why we were developing such an expensive piece of hardware instead of streaming the songs over the Internet. Good question!
- Expensive “artisan” recording of samples not scalable. We had over 70 musicians ranging from opera singers to punk rock guitarists listening to each track to choose the right 30-second sample.
- Story for music stores was difficult. Music stores paid about $300 per month for each iStation they carried. The iStation resulted in some additional sales, but mostly of long tail cd’s. Members would listen to the popular artists from the radio, and quickly realize that while the hit was catchy, the rest of the songs were terrible. But they also had the chance to try other cd’s from artists who weren’t receiving as much radio play. We saw an increase in the number of sales of these “lesser artists”, since hearing the song beforehand lowered the perceived risk of dropping $20 on a cd. But this created a problem for the music stores and music labels. It made inventory planning more difficult when they would normally carry 10 copies for big artists, and only 1 or 2 of the smaller ones.
- Story for music labels was all or nothing. And they said no. Music labels were both partners (rights to music samples) and customers (paying $.05 per sample). The data showed that people tended to buy more, but not the major artists the labels were promoting. And the music industry was (and still is to a large extent) an oligopoly run by a few large labels – Warner, Sony, and Universal. The music labels were used to being paid each time someone listened to a song. Now they were being asked to pay for each sample in exchange for tremendously valuable data that they had no way of gathering prior to the iStation. Their perception, however, was that they were essentially being asked to pay to disrupt themselves, and their combined answer was no.
You’ll find the rest of the presentation on slideshare:
Paul Fox @profefox