9.17 Early Dotcom Failures

The early dotcom years were hard for everyone, and it was some time before viable business models were developed.

Many ecommerce sites were amateur affairs, costing little money and effort to set up, and losing little when they were quietly dropped. But many of the larger sites, receiving outside funding and professional management, are still in business, though often under new owners.

Flouting sound business rules was the overall reason why the larger Internet companies failed. In the 'new economy', mundane matters like cash flow and long-term profitability were set aside for first mover advantage, market position and brand identity. Below are a few stories from Sean Carton's The Dot.Bomb Survival Guide: Surviving and Thriving in the Dot.Com Implosion, though different conclusions are sometimes reached.


AllAdvantage paid surfers to look at adverts served up by a specially-installed browser bar, and gave them an extra 10 cents/hour for each new member they introduced. Income was to be generated by advertising revenues and by selling the information collected. A modest 30,000 members in the first quarter was the goal on launch in March 1999, and Softbank Capital Partners put up venture capital of $135 million.

Like many pyramid selling schemes, the service was initially a great success. Membership soared to 2 million in the first 8 months — landing AllAdvantage with a monthly bill of $40 million. Unfortunately, advertising revenues never met that figure, indeed amounted to only $10 million at the end of the first quarter. The company considered an IPO in February 2000, but the dotcom bubble had burst. The IPO was pulled, the payments cut, and a sweepstakes model introduced. But by December the company had cut 300 from its workforce, seen its membership fall to 547,000, and in February 2001 was forced to close.

Was AllAdvantage a victim of timing and its own success? To some extent. Few foresaw the Spring 2000 crash, and the sweepstakes model was viable, though the switch came far too late. But the real problem was the advertising. Adverts work when targeted properly: at the right people in the right way. That was something AllAdvantage couldn't offer, and advertisers understandably held back.


1. Be commercially viable, and don't count on cash injections or share flotations.
2. Remain in control, of both income and expenditures.
3. Test your assumptions carefully.
4. Research key aspects, taking expert advice where necessary.


Backed by Idealab! (see below), and launched in June 1999, Eve.com was an ecommerce site offering 50 exclusive brands of cosmetics, backed with makeup advice and fashion information. The selection was later expanded to 120 brands, but Eve.Com didn't come good, and was closed in October 2000 with the loss of 164 jobs. What went wrong?

The first problem was the size. Sales of $1 million/month in a $30 billion/year market could not attract funding in the difficult 2000 climate. The competition was fierce, with over 100 online cosmetic retailers, including some big names — Glass.com (Estée Lauder) and Reflect.com (LVMH, Moet Hennessy, and Louis Vuitton with $80 million venture capital). The second problem was the nature of the product. Customers need to feel and smell cosmetics, but by not stocking the leading brands that could be tried out in the shops, Eve.com denied customers this essential experience.


1. Either be largely self-funding like FragranceNet, or incorporate the competition in the the 'one stop' shop.
2. Understand the customer: retail is detail.


BizBuyer was a portal site aimed at the small and medium sized business. Buyers {3} posted their needs in an online reverse auction. Vendors bid for the business, paying a $3-$10 fee to BizBuyer, which then posted the bids on to the customer. BizBuyer started in September 1988, raised $69 million, but closed in December 2000, returning $35 million {4} to investors. Why the poor performance?

B2B exchanges were a good idea.{5} They make commercial sense, and commentators were correct in thinking they would become increasingly important. But exchanges are still not the way most procurement is done, and attitudes change slowly. Equally important is the need to focus on small market sectors, particularly if advertising revenues are sought.


1. Plan for the long term, even for the very long term.
2. Keep improving the technology, but remember that customer attitudes change slowly.
3. Target market niches rather than 'offer everything'.
4. Improve advertising revenues by refining target audiences.


Many companies have excess supplies or capacity: unrented cars, airline seats, theater tickets. Why not post details of these and other (unwanted) items and get customers to bid? That was the idea behind Priceline, where customers named their price, and the seller had to accept or reject that offer, either within a certain time frame, or as initially agreed with Priceline.

The concept was initially popular. Venture capitalists poured in $100 million, and the $16 shares soared to $80 on the IPO of March 1999. Priceline branched out to groceries with WebHouseClub, which signed up 15,000 customers in its first week. Nonetheless, Priceline's share value had fallen 42% by September 2000, and the company posted $100 million losses over that quarter. Losses mounted to $192 million in the third quarter of 2000, though a small profit was returned at the end of June 2001, and something more appreciable in 2002 and 2003. {9} The company remains in business, with 30% owned by Cheung Kong (Holdings) Ltd. Why the early difficulties?

A heavy drain was WebHouseClub, where the manufacturer refunded the difference between what a store charged and the customer paid — rather like coupons, but with shorter lead times. Not many manufacturers signed up to WebHouseClub, but the company still had to provide discounts on a wide range of items to keep the site popular. Customers saved some $4 to $6 per order, but those savings cost WebHouseClub over $1 million a week, which proved unsustainable: the company closed.

Then there was the hassle of booking a car or airline seat with Priceline. Not everyone wanted to hang around waiting to see if their 'bid' had been accepted (usually after several adjustments to preferred time and date), particularly after competitors like Hotwire.com and Orbitz.com had simplified the process. Nor was Priceline always the cheapest, {6} or innocent of 'hidden extras'.


1. Online commerce has to add value. Customers will suffer to save some tens of dollars on a flight, but not cents on a bottle of ketchup.
2. Better technology was waiting in the wings — which Priceline had not the time or financial resources to implement: look at the larger picture.
3. Americans are not used to haggling, and found the process irritating once the novelty wore off: know your customer.


SixDegrees seemed hardly a business at first, more an extended network {7} of friends. With over 1 million members by September 1998, and supporting itself by selling its accumulated demographic data to advertisers, the site went on to offer expert services and the chance to build online shops with the shopnow.com software. By December 1999, the company had spent only $150,000 on promoting itself, but was the acquired for $125 million in stock by YouthStream, getting major investment from News Corp and 3 million members.

Youthstream marketed to college and high school students, and attempted to expand its small mybytes.com teen community through sixdegree's user base and by offering better membership features. Other deals with teen-orientated websites followed, but the business wasn't prospering. Youthstream tried a B2B model, selling the software as connection technology, but marketing and software development created losses amounting to $28 million on revenues of $11 million in the quarter ending September 2000. A quarter later the stock had dropped from $32 in January to around $2, and Youthstream closed both SixDegrees and its application software provider company Sodalis. Youthstream itself was acquired by Alloy Inc for $7 million in August 2002.

Online communities are fragile things, and prosper only when there's common purpose and trust. SixDegrees extended membership by asking members to sign up their friends, who were in turn contacted. YouthStream had more bills to meet, however, and increasingly indulged in hard sell and spam, which only alienated people. Kids are not easy to sell to, moreover, and perhaps only 10% will use chat boards. A community service got highjacked by hard-nosed business.


1. Business and sentiment don't mix: a wholly free service is not a business.
2. Respect your customers' intelligence: spam is seen for what it is.
3. Too much competition: members simply switched to other online communities.


Furniture.com began life in June 1998 as furnituresite.com, one of the many furniture stores online, in this case the shopwindow of the 50-year-old Empire Furniture Showroom company owned by Steve Rothschild. Six months later the high-flying Andrew Brooks had taken over, raising $13 million in funding, moving the company to plusher surroundings, purchasing the furniture.com domain for $1 million and launching a $5 million advertising campaign. Sales rocketed. CMGI added another $35 million of funding, and by the end of 1999, Furniture.com had achieved net sales of nearly $11 million from 2 million visitors.

The IPO filed in January 2000 was expected to bring in $50 million, but Goldman Sachs pulled out when another IPO seemed risky in the overcrowded Internet marketplace. The company was now low on funds, and the dotcom crash of April 2000 didn't help. The $11 million sales of the previous year had been against losses of $43 million, and the company posted losses of $3.1 million the next quarter. In April the company laid off 12% of its workforce, and in May stopped payment to some of its creditors. More staff were laid off in June, and though CMGI injected another $27 in cash, and the company had achieved $22 million in sales by September, the losses continued. On November 6 the company closed down — to be bought out by former employees and resurrected in April 2002: as a smaller, leaner company. {13}

What had gone wrong? Almost everything imaginable. The order tracking software didn't work, and neither customer nor rep knew the status of orders. Some deliveries took 8 months to arrive. There were over 200 different furniture suppliers, with no way of managing them. Finally the company, desperate for business, offered free shipping, which only hastened bankruptcy.


1. Construct a proper plan: if profits don't exceed costs there is no business.
2. Ensure the technology works before you go live.
3. Market position is not a substitute for profitability.
4. Start small and expand as the market allows.


PointCast was going to take the Internet to new levels. Instead of surfing for information, users would have their own selected channels of news and information automatically pushed to them. The 1992 software that allowed Compuserve members to create their own newspapers from multiple sources was redesigned in February 1996 into PointCast Network (PCN), a development that received the enthusiastic backing of Netscape and industry pundits {8} everywhere. By August 1996, PCN users totaled 1 million, $36 million financing had been found, and advertising revenues of $2.5/month were being talked about. Competitors were frantically developing their own brand of push technology. {9}

PCN had its problems — it was slow, overburdening the many corporate networks that had installed it — but PointCast rejected an offer of $400-450 million from NewsCorp in 1977 and filed an IPO in May 1998. Other Internet IPOs were phenomenally successful at this time, and PointCast expected to raise $235 million. In July 1998, however, PointCast was obliged to pull its IPO. The technology didn't really work, and the company had lost $58 million.

A letter of intent to purchase for $100 million was signed by a consortium of telephone companies in December 1998, but PointCast laid off 220 employees in March. Eventually, PointCast was purchased for $7 million {10} by Idealab! (see below) and the technology developed into a EntryPoint, a 700K, well-behaved toolbar that delivered news, stock and weather information. The toolbar itself mutated into Infogate delivering personalized news and stock information, when EntryPoint merged with Internet Financial Network in October 2000.


1. Don't fall victim to your own rhetoric when better technologies exist.
2. Be realistic and accept a generous settlement when offered.
3. Think more towards applications, {11} as Backweb did with McAfee.

Volatile Media

Rather than suffer the inconvenience of downloading music from the Internet, customers of EZCD could get CDs individually filled with their recordings of their choice. A similar service for games was offered by the co-owned EZGamer.{12} Deals were made with all the large record companies, and the 24 hour shipping service was faster than downloads over a normal modem. EZGamer partnered with GameSpot. Yet neither service prospered, and both were shut down in August 2000.

'Technology' is the one-word reason. Customers could afford to create their own CDs when prices fell for both blank CDs (from $10 to 10 cents) and CD burners (to little more than $100). Broadband Internet access also became more generally available, and many obtained free downloads through Napster.


1. Look ahead: if a digital product can be made faster and cheaper, then the history of computing and the Internet suggests it will be.


Everything about Clickmango was something of a record. Launched in April 2000, the online health and beauty site had raised $4.4 million in 8 days. The media darling Joanna Lumley {13} helped as promoter and site guide. The site was favorably received, and page impressions reached 750,000/month. Yet the company couldn't raise more money in the difficult trading conditions of 2000, {14} and closed in September, four months after opening.

Bad timing? Not entirely: customer acquisition costs were the real problem. At its worst, ClickMango got through some $730,000 per month. Supposing that the site received 250,000 visitors/month (3 page impressions per visitor), and 2% of these ( the norm) purchased something, then an average purchase of $146 would have balanced the books. But sales in fact were only $12,000/month, an average purchase of $2.40 on these calculations, showing the large reality gap that confronted investors.


1. Do your sums properly, employing industry averages.
2. Plan for the long term.


Niche marketing is often the clue to ecommerce, and the QuePasa portal site — with free email, auctions, chat rooms and ecommerce — focused on the under-represented Hispanic community. Gloria Estefan kicked off the high-profile advertising launch in 1998, and the offering share price of $12 reached $113 on its IPO in June 1999. Unfortunately, in line with increasing losses ($6.4 million in 1998, $29.3 million in 1999) the share price price fell to a dollar in April 2000, and the company came close to liquidation in December 2000. {15} Most staff were laid off the following year, but the site was reactivated in 2002 and acquired the search and retrieval technology of Vayala Corporation.

Three things caused the hiccup. First was the assumption that Hispanics would prefer Spanish to English language sites. They don't: a study by Espanol.com {16} found that only 8% preferred content in Spanish. Second was the digital divide: comparative fewer Hispanics were online, or inclined to purchase computers. And third was the view of the Hispanics as a monolithic community. It actually consists of various subgroups, each with very different cultures.


1. Market research is critical.


Where venture capitalists waited for business opportunities to come to them, incubators started with the good idea and built it into a viable business through know-how, premises, money and financial contacts, often taking a stake in the company it nurtured. Business incubators allowed MBSs to spread their wings in the 'new economy', and were popular until the 2000 Internet crash.

Among the best known was Idealab!, launched in 1995, and the source of many household names: eToys, CitySearch, FreePC, CardsDirect, Commission Junction and Goto.com/Overture. {17} Its founder, Bill Gross, had created the multimedia company Knowledge Adventure, and in 1995 produced CitySearch, which was sold to Ticketmaster for $260 million. The Idealab! model was simple and successful. Each venture got an initial $250,000, and Gross was paid handsomely for his money, time and expertise when the venture went public. When eToys was floated in May 1999, Idealab! had a capitalization of $1.5 billion, which jumped to $5 billion when Goto.com went public.

Unfortunately, not all ventures were successful, or stayed so. Gross lost serious money when share prices declined sharply, and investors soon worked out that Idealab! products were not always a long term investment. In an effort to raise $1 billion in early 2000, Gross guaranteed 50% of Idealab! if the future IPO of Idealab! didn't go well, but pulled the IPO in December 2000. He also embarked on a spending spree: $200 million on CarsDirect, $60 million on Homepage.com, $20 million on Goto.com to a total of $800 million by August 2000. By December, Idealab! had only $50 million in cash {18} and the following year started laying off staff.

The difficulties centered on the US Company Act of 1940, by which Idealab! was obliged to take a stake in its offspring to avoid being classified as an investment company and so subject to restrictive regulation and reporting standards.


1. Regulations, however sensible, can create their own problems.
2. Take cash where you can.

General Conclusions

The dotcom bubble was like all such flights from reason that periodically excite the business community. What led this one was the spectacular fortunes of Yahoo and Netscape. If a couple of Stanford students could make themselves billionaires, and a group of software developers create something in the big league, then the way was open to everyone. And hadn't Amazon shown that you didn't need profits? Just keep expanding and the stock-market would do the rest.

So, in general:

1. Internet companies were not reckless but got locked into the logic of the market and their business goals.
2. Advertising spend was the great killer, followed by software development costs. Both were unavoidable once the business was launched.
3. The pressure of the market (i.e. investors) caused most problems: the land rush mentality, dreams of instant fortunes and the slack accounting.


1. Give a brief history of the early dotcom failures.
2. First-mover advantage can be overrated. Discuss.
3. Describe in some detail the history of two failed early ecommerce ventures. With the benefit of hindsight, how would you have saved the companies?
4. Take one of the companies and replace its operations with current technology, including marketing techniques. What would the estimated profit and loss accounts look like?

Sources and Further Reading

Most of the original sources for the sections above are no longer live, and are not therefore listed.

1. The Dot.Bomb Survival Guide: Surviving and Thriving in the Dot.Com Implosion by Sean Carlton. McGraw-Hill. 2002.
2. Ghost Sites. Extensive listing of sites that failed: mostly just screenshots.
3. The Truth about Priceline.Com. Angelfire. July 1999.
4. Six Degrees. eOpinions. January 2005.
5. PointCast Adds Bells and Whistles to Its Push by Miguel Helft. Wired. June 1997.
6. Wary Microsoft Partners Take Pragmatic Path by Chip Bayers.Wired. March 1997.
7. PointCast Coffin About to Shut by Craig Bicknell. Wired. March 2000.
8. BackWeb Revs Portal Server Share by Renee Boucher Ferguson. eWeek. April 2002.
9. Ezgamer.com unveils a new CD fulfillment solution for Talonsoft. PRWeb. June 2000.
10. Clickmango goes under. BBC News. September 2000.
11. Funding dries up for European e-commerce sites by Paul Mungo. Computerworld. August 2000.
12. Quepasa.com to Liquidate Assets by Carol King. Internet News. December 2000.
13. Ideas Aplenty From Idealab Head by Joanna Glasner. Wired. July 2002.
14. Incubators On Life Support: Idealab! by Lisa DiCarlo. Forbes. January 2000.