In the theme of comparing the AI bubble to the dotcom bubble, I thought there might be some merit in looking back at the irreverent book ‘F’d companies – spectacular dot-com flameouts’ to see what lessons it might have for the AI bubble and post-burst period. Published in 2002, the book judged many of the ventures too harshly, as it is now apparent that many things failed simply because the timing was not right, typically with technology market development or delivery out of sync. In time, many of the dreams have, in fact, come true.

Dotcom start-ups were not a big thing

In the dotcom bubble, there were lots of small ventures doing crazy and not so crazy things, but the scale of these ventures was limited – we are generally talking tens of millions of $ where now the reference is billions. The dotcom companies did not vacuum capital in chunks in the way that AI is doing now.

The scale was tiny vs AI’s ambitions

The great capital vacuum was the telecom operators and equipment companies, the likes of JDS Uniphase, Lucent, Cisco, Nortel, MCI Worldcom et al. – drawing in hundreds of billions of $, but not trillions. Much of it was debt, much of it share issuance, but, in contrast to today, relatively little of it was cash from profit.

Product market fit needs work and patience – retail

There were plenty of retail dotcom ventures and most struggled to get product and market to fit.  Webvan, Pets.com and Furniture.com were all derided in the book for their errors, but two decades later the same basic business models exist and we all understand that the dynamics of online retail can be just as nuanced as they are for traditional.

Webvan
Webvan infamously ran through over $1bn trying to build online food retail and delivery.  The supermarkets generally seem to have settled on a way to deliver this service now, each finding a suitable mix of delivery, automation, warehouses and stores. Notably, the costs incurred to reach this point will have been many times Webvan’s derided extravagance.

Pets.com
Pets.com was one of several online pet supplies retailers that together worked their way through several hundred million dollars before Petsmart was eventually left standing with a business and delivery model that works. Others now prosper too, as it turns out people are happy to buy pet food in bulk and to pay premium prices for special niche foods and accessories.

Furniture.com
Furniture.com was just one of many online furniture retailers that failed to meet the challenges of technology, managing a very 20^th century Mom & Pop supply chain, and delivery costs that frequently far exceeded the sales price. Grand in ambition, none of these furniture retail ventures amounted to any great scale, with funds raised rarely into hundreds of million dollars. This model now works but it is far from achieving the scale many initially hoped for.

There will always be some things that are just plain daft

iharvest – just daft
Some applications just didn’t seem to understand what the internet was, iharvest being a case in point. iharvest allowed you to save your favourite websites, not just bookmark them, saving them to a remote server/drive. Truly a problem that did not need solving. It was stupid but it only got high single figures $m in funding.

MusicMaker – competing against free
Backed by EMI, MusicMaker raised close on $120m when it floated in mid-1999 with a $600m valuation. Rising more than 70% on the first day of trading, its revenues were barely into six figures $.  MusicMaker allowed customers to create their own CDs online for $10 each. It had a limited back catalogue and the competitive challenge of Napster delivering music content for free. MusicMaker closed in January 2001.

Go.com – just copy?
One business plan that is always challenging is ‘let’s just build what someone else has and somehow hope that we do it better or get lucky’.  Go.com was Disney’s attempt in 1999 to recreate Yahoo!, building upon its investment in and the acquisition of Infoseek with access to Disney’s content. Disney shut it down in January 2001 with a write-down of just over $800m.

Digital currencies – doomed to fail!

At the end of the nineties, Flooz.com and Beenz took up the challenge that had defeated DigiCash earlier in the decade, creating a digital currency.

Some might now think it odd that when they rolled over and died no one was pointing to the lack of encryption, let alone ledgers, as a reason for their failure. Combined, the two businesses raised no more than a couple of hundred million dollars, perhaps not that much in comparison with their grand headline grabbing vision, and certainly not the scale or traction we see from the current generation of digital currency offerings.

Application Service Providers – beware the overreaction

Application Service Providers provided software functions online, primarily intended to be delivered via websites/browsers. They were the precursors to SaaS.

BidLand provided online auction functionality for websites while Lipstream provided voice over IP functionality for customer service websites.  BidLand’s market did not grow big enough fast enough, while Lipstream could add no value as successful VOIP calls required wider network infrastructure that was simply not there.

It is difficult to imagine nowadays that anyone wanting these kinds of services would follow any path other than to use a SaaS product.

In 2002, F’d Companies fumes at the madness of these companies each raising just over $20m and employing nearly 100 people. But with a 2025 perspective, their demises clearly suggest that many sensible long-term stories were lost to the downside overreaction and an investor base that stopped believing.  It seems almost inevitable that the same will happen when the AI bubble bursts.

UK not immune, and not really that different

The book focuses on the US, but the UK saw its share of dotcom-era dramas, mostly telling the same stories.

• Boo.com, the sports-fashion retailer, was ahead of its time, with ambitions of 3D websites and interactivity on 56Kbps.
• Freeserve demonstrated that UK companies could create innovative business models even though it was doomed by the transition to broadband and the awakening of BT.
• LastMinute.com didn’t make money while listed but still survives as a brand, and it is hard to argue its founders didn’t spot a long-term profitable opportunity and model.

And as with the US, it was the telecoms operators and equipment providers, notably Marconi and Bookham, that hoovered the capital and really destroyed the value.

Ian Robertson
07768 276595