Was Copia wrong?

Jumia thinks not.

This is a short essay about one of the startups I really respect that filed for bankruptcy last year. It will also hopefully be the first in a series of essays I’ve wanted to write for a while.

---

11 some odd years ago, Copia Kenya was founded by Tracey Turner and Jonathan Lewis, barely one year after Kasuwa, later, Jumia, was founded. Both businesses couldn’t have been any different. Jumia was a rocket ship from the get go, with MTN Group and Rocket Internet (the German venture studio that gained notoriety for its so-called copy-paste startup building style) among early backers of the e-commerce venture. Copia was smaller and after a false start with a mobile app, the startup began selling to customers in small towns in Kenya, who selected orders from paper catalogues in nearby corner stores.

Image source: Facebook, Copia Kenya

But by 2023, Copia had raised roughly $100 million (Jumia had already raised multiples of this) and the startup had launched a new mobile app, and was planning a campaign to onboard new sets of customers to the mobile app.

I remember Copia CEO Tim Steel and another gentleman whose name, I’ve now forgotten, who had just joined Copia as CTO, explaining to me inside Copia’s HQ near Nairobi, how the company was hoping to find better ownership of its end-customer relationship by progressively moving its ordering process to the new mobile app.

The catalogues would remain, but Copia wanted to be more digital than it had ever been.

I downloaded the sleek app that day in early May 2023, but I didn’t try it out before I left Kenya later that month. One year later, the startup filed for bankruptcy in California and entered administration in Kenya.

Suffice it to say that the news of Copia going into administration didn’t go down well with me, even though I had accidentlally learned that it was in trouble before the news broke.

I’ve always wanted to learn more about what went wrong. Was Copia wrong about its entire thesis? Was Jumia right? I opened a draft in July 2024, but I never got down to filling it out.

Two weeks ago or so, I saw something that brought Copia’s story to the fore, so I decided to put pen to paper fingers to the keyboard. This is a first draft of my incomplete thoughts about an incomplete picture. So far it looks like Copia’s general thesis about ecommerce in Kenya was right but incomplete. The same is true for the capital it needed to execute, it was OK, but not nearly enough over a long enough period.

Pictured: a snapshot of planned essays on my Substack dashboard. I plan to write more this year.

Experience is the best teacher

Three weeks ago, Jumia Group put out a story on their Kenyan business that reads in part:

Vinod Goel, the East Africa Chief Executive of Jumia believes he has cracked the code to grow the e-commerce platform into an unrivalled firm in the region’s rural market.

The 2024 Black Friday campaign provided the light bulb moment as he revealed that the orders grew exponentially increasing by 70 per cent

Similarly, by December 2024, pick-up stations had increased from 220 to 300. He says the plan is to increase them again this year by between 60 and 100. Rural markets hold the key.

We are seeing a clear trend that outside Nairobi is growing much faster at two to three times.

The light bulb moment Goel describes is similar to what Tim Steel (and later Tracey Turner) wanted me to see when I first visited Copia’s Kenyan headquarters at Tatu City, just outside Nairobi.

The article I’m quoting from continues, “Previously Jumia was also sceptical about delving into the rural market. The platform would only consider opening a pick-up station if the area population was between 80,000 to 100,000 to guarantee economies of scale.” That has now changed. Some towns with only 35,000 residents apparently buy enough from Jumia that it makes sense for the company to open a, and it’s important to note that it’s likely just “a” pickup station.

But the bigger point to underline here is that like Copia, Jumia’s Kenyan boss is saying that there is an underserved growth opportunity in peri-urban locations which Jumia can tap into. And like Copia, he is also readying and already deploying an army of agents to support what his organisation sees as a latent opportunity to meet commercial demand for a range of products from beauty products and shoes to house appliances.

Jumia already has 25,000 agents (at its peak Copia had roughly around 40,000+ agents across the small towns it served in Kenya).

Jumia has deployed J-Force, a group of agents, whose role is to freelance and guide individuals to purchase from the platform. There are about 25,000 of such but on average 7,000 are active on any given day.

It is these agents who help skeptical first-time online buyers or those who still use feature phones to buy online…

Goel’s statements are more or less how Tim, or Tracey Turner, Copia’s co-founder, former CEO and later board chair, would have pitched the company to investors or FMCG partners.

How do you fail with $100 million?

In the comment section of a short biography of Copia on Startup Graveyard Africa, someone quipped, “How can you fail with $100m+ in funding, even if you are to stop the business and just pay salaries it will take you at least 50years+ to collapse.”

The second part of the sentence is obviously an exaggeration and perhaps betrays the comment author’s ignorance of the true scale of a Copia-type business. The snide comment about Copia founders not having the experience of rural living in Kenya to guide them, is also similarly discountable. But the question of how it was possible to fail after raising $120 million remains.

The short answer is that $120 million spread over 11 years to cover short-lived regional expansion, and the launch of Copia-branded products (on slim margins) is not enough.

The long answer is that while Copia built everything to serve its primary thesis on peri-urban markets, it had unwittingly locked itself out of the benefits of agglomeration that urban areas and metropolises typically offer. So while it focused on serving peri-urban and rural market demand with some level of predictability, it lacked the relative stability that a denser urban market would provide.

This could have meant that its capital expenditure and cost of sales would tend to be relatively inefficient compared to expenditure from traditional ecommerce companies.

Ecommerce works the way it works for a reason. The ‘e’ in ‘ecommerce’ is simply how technology tried to recreate the density of options and relatively near instant nature of transacting in any physical marketplace, mall or supermarket. Ecommerce is artificial densification. And the e-commerce difficulty scale increases or decreases in tandem with the potential for agglomeration effects.

Still Jumia is now doing what Copia was, or at least that is what it looks like. And while it’s still early days they seem to like the results enough to brag about it.

According to Jumia’s Goel, while the company was initially skeptical about operating in towns with populations fewer than 100,000 people, that filter has come down twice—to 80,000 and now 35,000, this despite Kenya’s economic malaise.

That said when I read, “What we found out is the moment we open a pick-up station in any town even with a population of 35,000, it starts to give enough orders to sustain it,” I cannot help but feel that this is Jumia Privilege speaking.

What is this “Jumia Privilege”?

“Rural” Copia was not enough

NB: Copia was not really a rural ecommerce platform. It served small towns that were outside of the larger cities and had a few truly rural agents which it served as well. Most clients were in what Copia called peri-urban areas.

The “Jumia Privilege” is the incremental advantage that large platforms gain from existing operations that enable them to launch in new directions, copy the competition, or seek to test and exploit new markets, as they become more efficient and dominant in their existing space. In essence, some of Jumia’s current market knowledge and processes can be redirected to test out new strategies while its current “safe” business continues to generate cash.

This type of privilege means that any opening up of market share in a promising but untested space will be more quickly filled in by the Jumia-type company than by an upstart, especially one with limited resources, impatient or mismatched expectations for capital investments to pay off. Jumia has more of this luxury (even though it’s constantly on the edge) than almost anyone doing exactly the same time in Africa today, and there are not many Jumias.

---

Image source: Facebook, Copia Kenya

I first wrote about Copia in February 2023 in a piece titled, “Agent commerce: how e-commerce in Africa is changing”. My basic argument was that a lot of investors and entrepreneurs had come to believe that ecommerce in Africa needed to become streetwise, hence the birth of TradeDepot, Wasoko and RejaReja by MarketForce among others—but that Copia’s approach stood out.

I wrote:

Despite growth in mobile phone and internet users, and an increase in the number of people using formal financial services, early online shopping ventures cast in the mold of Amazon have fallen flat. More than any, Jumia’s ongoing struggle to establish itself as Africa’s online retail market made it clear to investors and entrepreneurs that replicating Amazon’s online success (itself waning) in Africa is a costly undertaking. And in the last three years, African e-commerce companies (and investors) decidedly turned towards competing with (some say complimenting) the middlemen who sell to street retailers.

I was right. Investors (public stock investors) and private venture capitalists had soured on the Jumia model. Every Jumia press release and official filing was watched for clues for how much better (or worse, depending on the reader) Jumia was faring.

In 2021, it was easy to sell revised theses for investing in ecommerce in Africa. And in more ways than one, that idea—that serving corner shops is easier than direct-to-consumer ecommerce platforms, is correct. But ‘easier’ doesn’t always mean a better business. It might mean a slightly less expensive model, but that might also be at the expense of scale.

Attempting to scale this slightly less expensive ecommerce model runs the risk of greatly increasing capital expenditure along narrow absorption lines on top of a smaller customer base. This becomes a problem when/if the fixed costs grow and revenue from sales fail to significantly offset growing costs.

More money may not have saved Copia. But was $120 million enough of a price to learn the lessons that will now allow Jumia (and maybe Copia 2.0) to serve smaller towns better? We will see.

Update:

Someone reached out via WhatsApp wondering how Jumia would pursue the agent commerce model.

I don’t sit in on Jumia meetings, but for as long as they can conjure revenue from small towns and the costs remain sane, Jumia will likely continue to play with agent commerce and see how far they can go.

It’s “extra” sales. Not a primary business strategy and that is my point. What Jumia can afford to do as a side strategy, Copia was doing as its main business.

Having the option to switch gears is an underrated power that $120 million did not or could not buy.