Swvl's terrible unit economics

Good afternoon,

Swvl’s market cap as of December 23rd is less than $15 million. This is what a lot of pre-seed stage startups in the region raised money at, last year - and some even in the first half of this year as well. Hovering around $0.10, its stock has lost almost 99 percent of its value YTD, and is among NASDAQ’s 15 worst performers (YTD) of the year.

As Swvl struggles to turn things around, its investors, particularly Agility and Zain who invested millions of dollars just before the firm went public, are left with no choice but to wait and see if their investments will recover. Currently, every $100 invested in Swvl is worth just $1.

(Agility had a very interesting role in Swvl’s debut on public markets but more on that some other day).

The early investors including the likes of Beco and Vostok who have funded the company across different rounds will also have to wait and see how the situation plays out.

Vostok still seems very optimistic. In their latest financial report, they note, “We continue to believe in the company’s upside with a special emphasis on the company’s core B2C markets, which we believe have the potential to build high barriers to entry through the presence of strong network effects.”

Since the report came out (October 20), Swvl’s stock is down 82 percent (until December 23rd, 2022).

The answer to how the company, which was once the hottest tech startup to come out of the Middle East, has ended up here would be different depending on who you ask. In my opinion, its poor economics and rushed debut on the public markets are the biggest reasons.

There was never a doubt about the company having a product-market fit. It was clear from the outset. It doesn’t imply that the company offered excellent service. What it apparently meant was that there were no better or cheaper alternatives. Public transportation in the markets where Swvl operates is subpar and Uber and Careem are relatively more expensive.

What was not clear, however, was how Swvl would make the (terrible) economics of their business work.

Uber, Careem, and the likes, never really rented cars or hired drivers on a fixed salary. Well, yes, they did offer very generous bonuses to drivers in the beginning, to drive supply to the platform but apparently nothing beyond that.

Swvl’s model on the other hand, basically, involved renting (or engaging) buses with drivers, for a fixed monthly fee and then running them on fixed routes.

(There are a lot of other differences too between Uber and Swvl. For example, Uber can be requested at any time and it takes the rider from point A to D. What Swvl offers (or used to offer) were rides on fixed times and generally from point B to C. The user has to figure out other options for their first and last-mile. But we’re not discussing that in detail today. This piece is more on the differences in the business model).

Swvl used to offer users a highly subsidized fare, in the beginning. Nothing wrong with that. It is (or used to be) the standard playbook for every startup to get early traction - exceptional service at highly affordable prices. And that's what the Egyptian company was doing. At one point, it was cheaper to use Swvl than the government-run buses in Pakistan (Lahore & Islamabad).

But what they apparently miscalculated or didn’t pay much attention to was that they would never be able to increase the fares to a point where the economics would make sense and if they did, usage would drop because at those prices, people might prefer slightly more expensive alternatives like Uber, Careem, or a private taxi.

I’ll try to make it simpler. Uber and Careem are marketplaces that connect supply (driver) with demand (riders), and take a cut from the earnings of drivers whenever they complete a trip. Swvl is more like a trading firm that buys inventory (bus seats) from companies or individuals that own these buses (by renting a bus or maybe an entire fleet of buses on a monthly basis) and sells them to consumers (riders). To make money, they should sell these seats at prices higher than what they’re paying but what has happened until now is that they’ve been selling them at significantly lower than their purchase price.

I explain this in more detail in the chart below.

In 2019, Swvl’s revenue was $12.4 million and its cost of sales (basically the money it spent on renting buses) was $33.8 million. They’ve been able to bring this number significantly down but even after all the changes they’ve made, their cost of sales is still higher than their revenue in the first half of 2022.

Let’s assume that Swvl sells a bus seat for one dollar. In the first half of 2022, this bus seat cost the company $1.21.

(What’s very important to remember is that we’re only talking about the cost of the product that Swvl is selling. Marketing, payroll, administration, and every other type of expense are not included in these numbers.)

In 2019, the same bus seat that we’ve assumed Swvl sells for one dollar used to cost the company $2.73. It was able to bring this down to $1.53 in 2021, $1.28 in 2021, and as explained above, $1.21 in the first half of this year.

Swvl would’ve perhaps continued doing this without worrying too much of the impact on its bottom line if it had access to an infinite supply of money but life is not a fairytale.

To fix this issue, Swvl has been shutting down B2C businesses in different markets to rely more on what it calls TaaS of Transportation as a Service, its corporate offering, as a source of revenue.

Its share has increased from 48 percent of total revenue in 2019 to 61 percent in 2022 (H1). TaaS had helped Swvl generate over 60 percent of revenue in 2020 as well but that was apparently due to the stalled growth of B2C revenue as it was Covid year.

TaaS helps Swvl solve a few things - including better predictability of demand. When they onboard a new company or an educational institute, they know how many people would use the bus on a regular basis so they can optimize the supply based on that. But it has little tech advantage as with this, Swvl is replacing traditional vendors who offer transportation solutions to corporates. Yes, Swvl apparently offers a nice dashboard for the admin of the companies it onboards as clients but that’s about it.

Pakistan generated a significant portion of the company's revenue in 2021, with over 25% coming from the South Asian nation. Despite this, the company made the decision to end its Daily (B2C local rides) business in Pakistan earlier this year, and then fully withdrew from the market last month (which means that it was also shutting down its business offerings in the country).

It suggests that the economics of its business in Pakistan were worse than some of its other markets and it could not bleed more money here. More importantly, it also confirms that B2B offerings would also not work in all its markets. And it is going to be even more difficult for the company to build a moat with that as there are no barriers to entry in this category.

Swvl also let go of over 50 percent of its employees (across its markets) as part of its ‘portfolio optimization program' last month. A large number of employees who were affected by this were part of its engineering team. (This comes just six months after the company reduced its workforce by 32 percent earlier this year).

The company said it is taking these measures because of the uncertainty in the global economic environment and volatility in the capital market, “[It] potentially impacts Swvl’s ability to generate sufficient cash from operating activities and external financings to fund working capital and service its commitments.”

It is uncertain at this time if these efforts will help the company navigate the shitstorm it is in right now. But for now, the wheels on the bus have gone flat.

---

Termsheet Numbers

Termsheet may get a new name in a few days.

The number of subscribers has increased to 3,988 - and the open rate for the previous issue was 48 percent.

The previous issue had a poll asking readers what they like the most about Termsheet. Here are the final results.

I should’ve perhaps highlighted that I wouldn’t know who voted what to generate more votes. A very small number of active readers participated in the poll but I think the results wouldn’t have been different even in the case of a higher number of participants.

---

Please don’t forget to hit the like button and share Termsheet with your friends and network.

Happy holidays.

See you soon,
Zubair