... continuing from Scaling bloopers
There is a lot written on startup failures. You don’t need me to retell the stories, you can just google them. Hence, I will not go into too much detail and instead would link some of the most famous scaling issues to the points I made in Scaling bloopers and Scaling bloopers, part 2
Groupon allowed users to get discounted goods and services by buying as a group. It claimed to be the “fastest-growing company ever”. Whether it really true or not, the rise was astonishing.
They had an a very successful IPO in 2011, with a net loss of $420m in 2010. The revenue was skyrocketing – 2000% growth in 2010. 2011 was slower – with growth of “only” around 400%, but reduced net loss of $233m. 2012 hit them hard – revenue growth slowed to 45% and net loss widened to $373m. And that’s on $2.33 billion (!) of revenue. Share price dropped from $20 to $9 In just a few months.
At the end of 2012, Groupon had $1.2 billion in cash and cash equivalents and no long-term debt. That means good safety net and no real worries, right?
Well, the biggest issue was that business had a problem. They threw money at marketing promising new clients to their customers and delivered exactly that. Only, those were cheapskate customers hunting for one-time deals. Economics of the first visit were horrible for their customers, who were usually small businesses and had a really hard time dealing with this. But many did take this hit in the hope of getting return visits. When it became clear that return visits didn’t really happen, Groupon started struggling to sign up new business. Why would you lose money on the first promotional visit which also becomes last?
There was also an arms race to win – new hyped model and low barriers to entry usually lead to that. There were so many clones that it was hard to believe. I had a stake in one of the first copycats in Russia (one of very few cases when I actually had something to do with the Russian market). A couple of years later there were over 300 (!) similar ones on the market.
So, a proposition that wasn`t wanted (even opposed) by customers, spending an incredible amount on marketing and growth. In short - bad scaling leading to a pretty sour end. The company is still around, but I am not even sure what they do these days.
Internet darlings
Oh, the good old dot com bubble and crash. One classic premature scaling story after another. So similar to each other that I put these two together – you can add most dot-com failures to this list.
Webvan was a US online grocery store. Apart from new ways of shopping (on the Internet), one of the big promises was delivery within 30 minutes. Investors: Sequoia, Benchmark, Yahoo, Etrade, Softbank, Goldman Sachs. Real pedigree.
Boo.com was a European online shoes retailer started in London, at one of the iconic Internet startup addresses that saw several very famous names of the era. One of the things I remember was that users could view a rotating 3D model of the shoes. Online. Wow. Investors: JP Morgan, Goldman Sachs, Bernard Arnault (then chairman of luxury group LVMH), the Benetton family. Another wow.
Main reasons for failure: not knowing what they were doing and trying to aggressively scale it. All that in a hyped-up bubble which then burst leaving nothing but thin air.
These were the days of “those who don’t move fast lose in this transformational thing called internet and e-everything”. So, they went for a land grab. Boo.com actually didn’t sell anything before they opened offices in multiple countries. Founded in 1998, in October 1999 it employed 400 people in eight offices, and was building a state-of-the-art logistics business across multiple countries. Webvan went for 10 markets right away and built infrastructure in them (expensive infrastructure, may I add).
Then they realised people aren’t buying. Back in the late ’90s neither technology nor customers were ready for either of these propositions. Nobody was buying things online yet. Rotating an online 3D model is impressive in principle, but a lot less impressive via a slow dial-up internet connection, which was the only option then. And even if someone bought, delivering often was tricky. For example, Webvan’s promise of 30 minute delivery was simply not possible – putting an order together and getting through traffic in, say, Manhattan in that time wasn’t happening.
Result – paltry sales and high cash burn. Boo.com made £200,000 In its final two months – while selling across 18 countries and after about $100m investment. The cash burn was just crazy (by the standards of that time, chicken feed today…). To add insult to an injury, the market was going vertically down after the burst of the dot com bubble. Dot com and internet were out of fashion and dead. No more investors’ money for it.
One could argue that if they tested their model in one location on a small scale, found the problems, fixed them (or adjusted proposition) and then started scaling, the story would have been different. It would be – they both would have been operating on a small, more reasonable scale. And they would have been outspent and outmarketed into oblivion by someone else throwing cash at the issue. Those were the days; such was the hype-de-jour – invest or be left behind.
So, I don’t really believe a business could have become successful by behaving sanely in that time in an environment like this (with few exceptions such as Amazon). And the whole industry wasn’t going to change its behaviour. As often is the case with bubbles – it’s seen as the time for winners with guts and lots of money changes hands before the bubble bursts.
"We were too soon and trying to do too much, but those were the times" - Mr Malmsten, co-founder of Boo.com
Of course, today we do all of these things very routinely.
Blackberry was an early smartphone before smartphones were a real thing. It was targeting business users and basically gave access to email on the go. Revolutionary stuff. It dominated the smartphone market for a while, but couldn’t keep up with changes and failed to innovate. I don’t know why they didn’t think touchscreens would be useful. They also had a proprietary operating system and a closed system, i.e. no ecosystem around them. They essentially ignored both and lost to Apple and Android devices.
A classic example of a one trick pony (albeit with multiple devices)
A classic example of a one trick pony (albeit with multiple devices)
Younger generation doesn´t know that Yahoo was “the Internet” at some point. It started as "Jerry's Guide to the World Wide Web" and kept that “all about everything” theme for a while. This is Yahoo in 1999
And list of their acquisitions from 1997 to 2001
See if you can figure out what the company´s core business was, between a job search site, an internet service provider, direct marketing, Internet radio, E-commerce and of course news.
Clearly, too many things at the same time and undeniable similarity to what I talked about in “Too much too early or too little too late”. They are often remembered for failing to buy Google for $1m and later trying to buy Facebook for $1bn. But to be honest, I am not convinced that really would have made them a major player with Yahoo’s extreme lack of focus and really failing to select what their core business was. Instead, it probably would have killed Google.
An extremely well covered story. $47 billion valuation in 2019, lots and lots of locations, big name investors, big scaling effort, a failed IPO, ousted founder (with some questionable personal deals with his own company).
Fundamentally, what I believe led to a problem here is bad unit economics multiplied by aggressive growth. “Capture the market” (sooo familiar!) by locking down long-term leases on premium locations. While tenants, i.e. revenues, were short-term. Plus poor core economics – which I would guess became totally horrible as the company got into trouble, because once you have fixed costs that you need to cover urgently in order to survive, you start looking at incremental value rather than profitability of each user (“they bring cash we wouldn’t have otherwise which helps to cover the costs – better than nothing”). And that incremental value in the period of soft demand or economic downturn has a tendency to decrease as a function of market conditions and your desperation.
Then there were people issues. Neumann (the founder) was charismatic and a god at talking to investors and talking vision. But - he had a very lavish lifestyle, and self-serving approach, e.g., leasing personal properties to WeWork in my view wasn’t even close to the right thing to do. So, a board – founder conflict erupted, which is not only very disruptive for business operations, but it severely limits attractiveness to investors.
Artimi was a company building wireless communications on Ultra-Wideband (UWB) technology, a remarkable tech with military origins. The idea was to have a huge wireless pipe for data transmission, which could be applied to video, wireless Ethernet and all kinds of things. I am not going to go into the fascinating and sad story of the UWB industry, which I watched unravel over many years. I will, however, use this for an example of incorrect “weight loss” aka cash burn reduction. And I chose this one because I had an insider view for many years.
In short, the whole UWB industry struggled to produce a fully functioning product for a long time and, in the process, required a lot of funding. Not surprisingly, by the time the products were starting to work, the investors were tired of the promises. Raising money was hard to impossible. Sales were desperately needed, but they were slow for a number of reasons. As an inevitable result, cash burn reduction was required to survive.
But – the product just started working, was impressive and many potential customers were dabbling in it, trialling it, talking about where it could fit and so forth. Since several possible applications were on the cards, each needed to get the product tweaked to fit their particular needs. Which needed people, who in turn need food and hence money. The burn was high. Moreover, in such situation the companies are always reluctant to cut costs – “we are just about to hit ramping of sales, look at all these customer engagements! They will sign imminently and we will need people to deliver. So, it’s silly to make people redundant now”. The hard fact is that sometimes it’s true. The flip side is that making people redundant is the only way to meaningfully reduce costs in most startups. It also means letting go of some activities – which in this case meant many of the possible applications, hence walking away from the hopes of sales in those applications. Hard when you don’t know which one will take off first. An immensely difficult dilemma in the absence of a crystal ball.
There is no right answer, unfortunately. But when it comes to life vs death some hard decisions need to be taken. And quick. As I said elsewhere about weight loss – better to do it “before” (i.e. earlier) than to do it “after” which may be too late. Not being decisive enough often starts a downward cycle. If you have to make staff redundant, you want to keep your best people, but they are the ones who can find other jobs the easiest. And redundancies don’t increase the feeling of job security, do they? That’s the main reason to cut once and deep if you have to. Few good people stay pass the second round of cuts and even fewer after the third round. This is the situation when one is a million times better than three or more.
Plus (obviously), bigger cuts mean much lower resulting cash burn and hence longer runway, which comes with a higher ability to see which application picks up first.
Artimi didn’t really have the guts to cut deep in expectations of imminent sales. As a result, this is how we cut the workforce (numbers are illustrative, I didn’t go to dig out the exact ones from old board papers):
- Round 1 of cuts – 15%
- Round 2 of cuts – 10%
- Round 3 of cuts – 60%
See my point? After round 3 everybody was seriously questioning not how much trouble lied ahead, but for how many weeks the company will exist. Not good.
In the end, the company turned to delivering services as a way to survive. Which it ultimately didn’t.
By the way, I am not saying that things would have turned brilliantly if the cuts were done in a more decisive way with just one round. There was a plethora of other factors at play both for Artimi and for the whole UWB industry. But that’s a story for another day.
A big queue of those stuck at the border
The international expansion theme has produced an infinite number of stories, startups or big companies. I´ll mention a couple of big names and keep it short. Stories below have two similar themes:
- Implicit assumption that foreign users are the same as home market ones
- Expanding too quickly before testing the hypothesis
Tesco and the US
In the UK, everyone knows Tesco – one of the biggest supermarket chains in the country. When Tesco opened its first shop in the US in 2007, it was the third-biggest retailer in the world. Tesco wasn’t a widely known brand in the US, so the company used the name Fresh & Easy instead. As any self-respecting big company would do, they went for multi-store rollout, launching with 60 new stores in the first few months and quickly opening more.
They should have read this blog and tested the proposition first.
Americans are used to large stores, while Tesco’s Fresh & Easy stores were smaller, similar to those in the UK. There were issues with product range and self-service tills which didn’t appeal to American consumers. Additionally, some strategic decisions were plain strange, to put it politely. For example, many stores were in working-class areas, while products were for more well-to-do customers.
And the timing of the launch, with recession starting in 2008, didn’t help.
Tesco left the US in 2013, at a cost of almost $2 billion.
Starbucks and Australia
Australia only has around 30 (or 50 by another source) Starbucks stores, compared to 13,000 in the US and 3,400 in China.
It turns out Australians prefer local coffee shops that have the setup and the drinks that are part of the local coffee culture to a different global Starbucks. Plus, it was too expensive. So basically, Starbucks started the rollout assuming same cookie-cutter products and shops would do – and they didn’t. At the peak there were only 87 Starbucks, of which around 60 were closed and the rest are now operated by the 7-11 chain.
IKEA and the US
Contrast the examples above with IKEA coming to the US. In 1985 IKEA opened one store with an explicit goal of testing things out. Then it opened one more in each of 1986 and 1987. Then 2 more in 1988 and none in 1989 getting to a total of 5 stores in the first 4 years. In 1990 it finally ventured into their second state (Wisconsin) and in 1991 into their third state (Georgia). I am guessing they considered this enough testing and by 1995 IKEA had 19 stores in 11 states.
So, they didn’t commit a sin of scaling too fast and not testing the proposition. Which was rather lucky because the sin of assuming things would be the same as at home they are guilty of. When IKEA first entered the US, its products were rooted in European markets – and I guess for a company that made standardisation one of its main strategic advantages, it made sense to go with the standard offering. But it was too small by the US consumers, both in perception, but also including anecdotes of plates not fitting into kitchen cabinets. Talk about knowing your consumer, eh? Lucky for them, they ironed things out on a small scale before spending on expansion.
This isn’t really a failure to scale up, but I listed it to include one of the obvious problems.
I think everyone knows this example; it’s been so widely publicised that I don’t need to say much. I will just say that sometimes “fake it till you make it” smoke and mirrors approach goes way to far – in this case as far as to fraud.
Fraud is hard to protect against while in the moment, although later on everyone looks back and says “how could we not notice, it was so obvious”.
In the end, Theranos didn’t revolutionise healthcare, but it did make a major contribution to the history of startups by being the most visible case of a founder charged with criminal offense and going on trial. Amount of smoke and quantity of mirrors was assessed by judicial system in the end.
This one is actually not a failure either, but rather a demonstration of the landgrab that happens in some industries.
Delivery Hero has raised a total of approximately €5.5 billion in funding. You kind of have to when you want to be among the top B2C players in an industry with huge economies of scale. This means you are involved in an arms race to grab as much territory (users) as you can and you are doing it across many markets (Delivery Hero operates in approximately 70 countries worldwide). For illustration, one of its competitors, DoorDash, has raised a total of $6.6 billion and there are many more that raised similar mind-boggling amounts.
There was also a flurry of M&A activity, since this is another way to enter new markets or gain stronger position in existing markets. Delivery Hero has completed 35 acquisitions over 21 countries. One of the types of “new market” strategies is adding product categories to increase the economies of scale in markets where your position is good. Think of Uber Eats or Amazon Fresh, which basically capitalise on Uber and Amazon huge user base.
M&A deals roughly fell into 3 categories: market entry, strengthening position, consolidating position (the latter is basically getting out of countries where they didn’t get to the top). ChatGPT says “Delivery Hero sold its South Korean operations, Yogiyo, for approximately €1 billion in 2021, and previously sold a stake in its German operations to Just Eat Takeaway for €930 million in 2018. These sales are part of Delivery Hero's strategy to focus on markets where it has a stronger competitive position while generating capital for further investments in other markets”
Here’s the list of their acquisitions:
As a result, Delivery Hero operates a portfolio of brands including foodora, talabat, DámeJídlo.cz, donesi.com, efood.gr, foodpanda, Glovo, Hungerstation, mjam, NetPincér.hu, Pauza.hr, PedidosYa, Yemeksepeti, Yogiyo, foody, and InstaShop.
And here’s some recent news – very clearly showing Deliver Hero getting out of some markets, while strengthening others:
I think this illustrates how landgrab strategies in some industries work.
Do you like this post? I do other ones every now and then. Subscribe if interested.
© 2024 No to Startup BS
