Thoughts around Kinnevik’s half year report and the e-commerce industry in Nigeria.

Much ado about these numbers. I was supposed to be on a call about them with the guys at TechCabal earlier, but fate and Etisalat conspired to shame me. Disappointing, because I was looking forward to it.

I should preface the rest of this article by saying Konga is a company that’s quite close to my heart. I think the founder/former CEO, Sim, is a great guy that has shown excellent vision, courage and leadership with the three companies he’s started in Nigeria and I truly believe his primary motivation is a desire to see Nigerians live a fundamentally better quality of life. What he’s trying to do with Konga is by no means easy nor will success be clear cut, but it’s pretty commendable.

Here goes.

So, first off, what did the report actually say? Two things:

1. Konga has 184K active customers (people who have purchased on the platform in the last 6 months); and
2. Kinnevik thinks its 34% stake in Konga is worth SEK 101m ($12m at current rates), which implies the entire company is worth $35m.

First reaction: those numbers look quite ‘small’ (although, to be fair, the active customer number has been published for the last 3 or 4 quarters, so shame on us all for not paying close enough attention).

A useful valuation benchmark for ecommerce companies is 1–2x GMV. There are issues of net vs. gross GMV, gross profit vs. revenues, and the question of what exactly Konga’s revenue model looks like, but at a high level, I’d guesstimate Konga’s GMV is roughly $35m (let’s call it N10bn).

If you hunted around on Kinnevik’s website, here’s what else you might have found:

1. Konga’s active user numbers are declining quarter on quarter. 184K is 10% less than the number was at March 2016, which was 2% less than the one at year end. Active users grew at a pretty decent 17% between September and December 2015, but it’s tough to estimate how much of that could be down to ‘Yakata’ promotional activity.
2. To date, Kinnevik has invested a total of SEK 209m ($25m) in the business. That means they are carrying the Konga investment on their books at an unrealised loss (0.5x money). A year ago, they valued the company at $140m (2x money), up from $52m about a year before (when cumulative capital from both investors was $38.5m), implying an unrealised return of 1.4x for the investors after just one year.
3. Kinnevik has changed its valuation basis in its last report. Instead of last transaction value, they now use their estimate of ‘fair value’, based on market multiples. I don’t see the change as particularly noteworthy. The last transaction is almost two years old, valuations have come off a fair bit across the industry (tech bubble this, tech bubble that) and the macro environment in Nigeria is fundamentally different. Frankly, that’s what market multiples are for — to estimate an asset value. What I find more interesting is the fact that Kinnevik has a similar sticking point with Rocket Internet, which uses Kinnevik’s previous approach to Konga for its own reporting. Kinnevik has traditionally taken a more conservative approach, saying this was based on market multiples.
4. Konga hasn’t raised money for almost two years. In the tech/ecommerce world, that’s a long time for a company that’s got a high cash burn (inherent in the business model) and is growing. Its last round ($40–60m according to Crunchbase) was in October 2014. We know that Naspers led the round, and it doesn’t look like Kinnevik participated too significantly — it was diluted from 41% to 34% at some point in early 2015, according to Kinnevik’s reports. I don’t think the timing of the adjustment means there has been a secret round somewhere, I suspect it‘s just about the time taken to technically close the round.
5. Between launch in 2012 and the end of 2014, Crunchbase estimates that Konga has raised a total of $79m; they believe the right number for the Series C raise in October 2014 was $40m. Looking at Kinnevik’s reports, I size the round at closer to $50m, so let’s say they’ve raised $89m since launch.

And Jumia?

Jumia reported 1.3 million active customers as at March 2016. They are in several more countries than Konga, but Nigeria is the largest, followed by Egypt then Kenya (by GMV). It’s reasonable to use GMV as a proxy for customers, and I’d hazard a guess that those 3 countries are 60–70% of Jumia’s business. Again, complete guess, but I put Nigeria at 30–40%, so let’s say 400–500K active customers.

Jumia’s active customer number grew by about 100K between December last year and March this year. Over the same period, Konga’s fell by 13%. That growth could have come from outside Nigeria, but given that it happened on a base of 1.2m, if Nigeria is 30–40% of the business, you’d need a fair amount of growth from other countries to make up for a decline in Nigeria. My guess is a lot of that growth came from Nigeria. At the minimum, Nigeria is probably flat.

Transaction volumes grew 8% in Q1 2016 compared to Q1 2015, GMV fell 8% but gross margins are improving. They also talk about an ‘acceleration of the shift to the marketplace model.’ All here.

I think what looks like Konga’s relative underperformance is explained by three things: (1) some disruption to the business with the transition to a new CEO, Sola Adekoya. This is not unusual and perhaps to be expected somewhat, given the ‘ambassadorial’ profile Sim had in the market; (2) The fact that they haven’t raised capital in a little while, putting pressure on marketing/customer acquisiton and other spending (they’re letting people go); and (3) macro headwinds which continue to hurt local consumer confidence and the broader Nigeria equity story. Konga is a single country play (vs. a Jumia, for instance) and Nigeria is a tough place to raise capital for at the moment.

Here are the real issues.

(1) Market size, (2) business model, (3) competition. These are the big, fundamental, strategic questions. The operational (what to do to stem the losses, or how to grow the customer base, for example) are less important, to my mind.

Market size. Yes, long-term, indicators support an expectation of more and more online purchases. And, yes, ‘Amazon didn’t get there overnight’, but does the Nigerian market really have the capacity to support both Konga and Jumia at scale?

Business model. Neither company is running a sustainable business model yet, neither is profitable and I’d wager neither is cashflow positive. Both are in the middle of a pivot to marketplace four years after they started, and in both cases, the pivots have been ongoing for two years. Jumia recently consolidated all AIG’s businesses under a single brand. That tells me the cost of marketing several distinct brands in each market was starting to look scary.

Doing eCommerce Amazon-style is a low-margin, high-volume game. The problem with that is you have to outcompete the other guy, and to do that, you need to outspend him. Your customers aren’t loyal. Your inventory is full of commodities (like books or mobile phones) and you can therefore only compete on price and/or service quality. To do that, you need to invest in promotions/marketing, your technology/back end and fulfillment/last mile. None of that comes cheap, and discounting is a race to the bottom, so you’re also polluting the market/teaching the customer to behave in a certain way in a bid to ‘win’.

Many of those phones you see on Jumia are being sold at little or no gross margin (yes, sometimes at a negative GM). That’s great news for the customer but a nightmare for an investor. When Konga announced its last raise, the press said some of the proceeds would be used to build out its logistics (to be clear, this meant buying fleets of vans and scooters). The company was quoted as saying DHL and the like couldn’t keep up with how quickly it was growing. That logistics burden is huge in a market like Nigeria (probably a lot heavier, relatively speaking, than any ‘dumb’ infrastructure burden that even Amazon faced).

Competition. Jumia’s parent is well funded, has a number of powerful, well-capitalised strategic partners, and desperate to make a success of it. Nigeria is Jumia’s single most important market. Rocket Internet’s stock price is trading at less than half its valuation at IPO. The company is facing heavy criticism from investors and the success of its equity story will lie in its ability to find sustainable models in emerging markets, and what’s clear so far is that’s easier to do in some parts of the world than others. It needs Africa, and consequently, Nigeria, to work.

But Amazon took 20 years to get ‘there’.

Yes it did, and it spent a heck of a lot of money along the way pursuing a ‘winner-takes-all’ strategy in one of the most sophisticated markets in the world from a customer perspective, so arguably comparably easier markets than somewhere like Nigeria. It was almost 20 years before Amazon delivered one dollar of profit from ecommerce, despite operating for the most part in countries where people are rich (and probably a lot less price-sensitive than Nigerians), educated, connected to the internet and generally open to new experiences. Countries with credit card/digital payments cultures and decent distribution infrastructure.

Amazon has now taken that global brand, massive balance sheet and all the expertise gained from making mistakes and spending hard, to India, where Flipkart is desperately trying to defend its ‘first mover advantage’. Flipkart has raised $3.2 billion in 12 rounds over the last six years and is still at risk of losing the market to a new entrant with deeper pockets. I don’t know how much capital Rocket has put behind Jumia in Nigeria, but Konga’s $89m pales in comparison to Flipkart’s invested capital, even when you scale the capital requirement down for relative sizes of the opportunities in India and Nigeria.

It’s easy to insist that “it’s too early to judge”, that the trajectory for Konga or Jumia in Nigeria should be similar to Amazon’s, until you remember that in the two years between 1995 (when Amazon started) and 1997 (when it IPOed), 34 million people in the US came online, joining 24 million that were already connected. The average American grew her income by almost 10% over the same period. By the end of its first two years, Amazon had 1.6 million customers. The US had a population of about 260 million at the time.

The thesis that goes “because going to informal markets is so inconvenient, millions of Nigerians will shop online” has had four years to prove itself. In two years, Amazon grew its active customer base from zero to 1.6m in a country of 260 million. In four years, Konga and Jumia have (by my estimate) 500–600K active customers in a country of 180 million, many of which are probably not unique (so let’s say there are no more than 500K people in Nigeria that are currently actively shopping on Konga or Jumia). Flipkart has 20 million monthly active users 9 years in, with “no breakeven in sight”. The cost of changing behaviours is high, never mind the cost of building out infrastructure. The unit economics of some of these businesses in Nigeria are probably really ugly.

Jumia’s new Nigeria CEO recently said she thinks it will take another three to five years to reach profitability. That would be quicker than Amazon, but perhaps.

But even if that proves to be true, the fundamental question remains, does the Nigerian market have capacity for two profitable companies in this space?

E.