The Black Swan Of Broken Business Models

If a company is in business and taking in hundreds of millions or billions in annual revenue, it must be built on functioning business and operating models. Unfortunately, that’s not the case, and we have a black swan event brewing. Thousands of business models are built on unsustainable assumptions and will struggle to survive the next 3 years.

I wrote this last year, but the ideas are more relevant, and the trends are more obvious today.

Access To Low-Cost Capital

The first bad assumption is the most talked about: access to free or low-cost money. Startups like Uber spent years running at a loss in the name of growth at all costs. The company exists due to VC and investor funding. Between 2013 and 2016, Uber raised billions in funding with no urgency to switch from growth to profitability. That changed after Uber went public in 2019.

Investors have pressured the company to provide a path to profitability and execute that plan. Early attempts have included buying other startups like Drizzly and Transplace. These acquisitions were only possible because Uber had access to cheap money.

In a tighter economy or monetary policy, Uber would have moved to profitability years ago, or they wouldn’t be in business today. I’m not picking on Uber. It’s a company I personally love, and I’ve used Uber for rides since 2011. My rider score is 4.95.

The amount I paid for rides in those early years was subsidized by VC funding. My cost should have been much higher. Thanks to Menlo, Jeff Bezos, Jay-Z, Google, and several other investors, I got to ride at a discount.

That funding is why Uber and thousands of other startups are in business today. However, access to cheap money has dried up for all except the hottest Generative AI startups. Many are having trouble raising funds, even at lower valuations. The longer interest rates stay high, the harder it will get for startups to operate at a loss. Most have 2024 targets for profitability, but I have seen more aspirational roadmaps than feasible ones.

The change is impacting big businesses, too. Many companies carry large debt loads on their balance sheets, and debt costs will rise dramatically as those come due. The names on that list will surprise you.

Verizon has $151 billion in long-term debt with only $104 billion in annual revenue. Verizon has $14.8 billion maturing in the next 12 months . High debt loads are common among wireless carriers. T-Mobile has $7.7 billion in short-term debt . T-Mobile’s interest expenses for Q2 2023 were $861 million, and Verizon’s were $1.29 billion.

The cost of maintaining that debt will rise significantly since much of it will be rolled over. Wireless carriers are stable businesses with assets to borrow against. There are much less stable businesses with a similar cost of capital increase coming. Last year’s thesis was that rates would stay high for a short time and drop sometime next year. It doesn’t look like that’s coming to pass, so refinancing debt will not be cheap.

Consumers are also being impacted by rising credit card interest rates. Most major banks report a slowdown in consumer spending , and US household debt is rising to record levels . The pie is shrinking at a time when companies need pricing power or spending growth to offset higher capital costs or support their path to profitability. Companies that thrived when money was cheap and abundant are in serious trouble if conditions remain tight through next year.

This is just the first type of broken business model. There are 2 more flying under the radar, and these hit closer to data science.

Access To Free Or Cheap Data

Facebook's very sweet advertising business was popular with marketers because Facebook provided high-quality tracking data. Marketers had a way of tracking ad effectiveness and justifying their ROI calculations. The problem was Facebook didn’t own the data. Apple did, and when the company changed third-party tracking policies, Facebook’s ad business took a $10 billion hit .

Facebook built much of its ad-supported business model on access to Apple’s data. Removing access to low-cost data revealed the true weakness of Facebook’s growth trajectory. Many businesses, especially social media companies, will stumble when their dependence on cheap or free data is removed.

LLMs are built on free data, and that’s already proven to be problematic. GitHub’s Copilot was trained on open-source code repositories . The people who created free training data sets aren’t all happy about that. There is a similar outcry over books being part of model training data sets .

Social media was another massive source of training data. In response, X and Reddit have raised the price for access to their data APIs. Websites are putting scraping protections in place. There’s a lot less free data for LLM training available now. The next generation is built on curated data sets that cost more to develop but produce more reliable models.

Privacy concerns and consumer sentiment are eliminating many sources of free data. Concerns about OpenAI’s data policy reduced early corporate usage and shrunk the feedback data sets available to improve GPT. Cookies have been largely taken out of the equation.

First-party data is one of the few reliable sources of low-cost data left. The impacts will continue to play out as more data sources get taken off the table or are priced more in line with the value customers get out of them. This will hit startups especially hard. Early-stage startups have the least access to first-party data and are already challenged to raise pre-seed funding for data acquisition.

Access to data-generating processes will become a significant competitive advantage. It will enable companies to be first movers with new foundational models and deliver high-quality data products to partners and customers. However, startups that are only viable when the data they need is cheap or free are in trouble.

Few established businesses will fail because of rising data costs, but companies will see margins shrink. Starting a business or launching an advanced machine learning initiative will be harder. Products and businesses that are profitable today may not be feasible in the future. Combine the end of cheap data with the end of cheap money. That’s putting a lot of pressure on businesses that have gotten used to high margins.

Access To Cheap Labor And Technology

The third factor is also well discussed, but only in isolation. Rising labor costs alone have business impacts across industries. Rising costs have been passed along to customers, and that’s now trickling down to technology providers.

Oracle raised prices by 8% last year, and they are one of the lowest. Average software costs are up by 12% . It’s not limited to the US. Microsoft increased prices between 9%-15% for EU customers and 20% for customers in Japan. Rising enterprise software prices are helping to offset some of the rising costs startups and big tech companies are challenged by.

It’s essential to look at all 3 factors because they create conditions for a black swan event. Technology initiatives are more expensive than when transformation budgets were set back in 2021 and 2022. They are rising due to rising software, data, and labor costs. If the business plans to finance these initiatives with debt, the rising cost of capital piles on.

As consumer and eventually corporate spending slows, enterprises will be forced to cut back somewhere. Pulling back spending on forward-looking projects is at the top of the list. The software, infrastructure, and talent are expensive. Transformation should deliver incremental returns, but most aren’t broken down that way. The growth is a promise that’s a long way off. It’s an easy decision for CxOs.

That’s one way the dominos could start falling. A slowdown in technology spending and transformation hits technology companies that are facing rising costs from three directions.

What Can We Do About It?

The problem isn’t some systemic failure of businesses or the marketplace. We’re not staring a Great Recession in the face. Businesses and products are out of position for the changing economic conditions. That risks turning a year or two of low growth into something worse. We should be repositioning and planning for strengthening impacts from these three forces. The good news is that data science can offset all three forces.

Transformation initiatives must deliver short-term returns on the way to longer-term big growth numbers. Budgets will shrink if the data organization isn’t part of reducing costs and growing revenue next quarter, next year, and 3 years from now. Deliver incrementally and avoid projects that will take a lot of time to set up. Use these smaller initiatives to put one or two pieces in place that set up long-term growth drivers.

Delivering internal tools and automation to improve productivity is one type of quick-win project. If we can break the business’s reliance on hiring new people to scale it, we protect margins from one of the largest pressures on them. Each internal initiative enables us to put in a little more data gathering that can set the table for bigger initiatives. Save the business enough with small wins, and there’s a budget left for more expensive infrastructure.

Consumer and enterprise spending are slowing but not falling off a cliff. Wallets will open for best-in-class products. Data and machine learning-supported features can deliver functionality that competitors will struggle to replicate. The data team can deliver features to keep existing customers from leaving and taking market share from competitors. Data products are a high-margin revenue source that doesn’t take long to bring to market.

We’re seeing threats that the data team can help businesses avoid and opportunities it can help take advantage of. Build a partnership with C-level leaders and position the data team as a solution to emerging challenges.

?

I’ll let you in on a secret. I host the largest set of data and AI strategy, product management, and leadership content (over 500 articles) in a different place. I post selections here months after I release them. Be first in line for my best insights.