* Opportunity expands for consumer businesses in Africa
* Micro factories viewed as nimble alternative
* High unit costs may deter smaller companies
LONDON/ACCRA, July 4 (Reuters) - Tucked in a corner of a Guinness brewery in a run-down part of Ghana's capital, flanked by industrial silos and crates of bottles, stands the Cube, a gleaming mini-factory that may point the way forward for global consumer goods companies in Africa.
The tiny blending and bottling plant for Gilbey's gin, housed inside five connected shipping containers, gives Diageo , the world's largest spirits maker, a way to test demand for new drinks while minimising capital deployment.
"What the Cube enables us to do is a large scale test. We can produce reasonable numbers, not very high numbers, but enough to give you a sense of what the market can take," said Preba Greenstreet of Guinness Ghana, owned by Diageo.
Small, prefabricated factories are also used by SABMiller and Danish dairy cooperative Arla Foods and are under development by Nestle to start production cheaply in markets where the likely demand is unclear.
Africa's middle class is growing fast, with 128 million households expected to have annual income of $5,000 or more in 2020, up from just 85 million in 2008, according to a report by global management consulting firm McKinsey & Company.
International consumer and retail businesses have noticed, accounting for 17 percent of foreign direct investment in 2013 from an average of 12 percent from 2003 to 2007, professional services firm EY says.
But many remain wary of starting up local production due to political risks, high costs for power and materials, onorous regulations and corruption.
In the short term, pop-up factories may not be cheaper in unit production terms because of their small output. This, plus their necessarily simple processes, means they are not for everyone, but for some larger firms, they are a low-risk way to work out where to make bigger investments.
Established markets like South Africa and Nigeria already have large food and drink plants so pop-ups are most attractive in more modest markets like Ghana, Mozambique and Zambia.
In an eight-hour shift at the Cube, 25 people dressed in lab coats and two managers churn out 8,460 bottles of gin that are packed by hand in the final step of a process that involves blending alcohol with flavourings and blow-molding bottles.
In Ghana, where spirits are a staple drink at family and community events, Diageo can reach a wider market with its local Gilbey's gin, which sells at around $2.50 per 75 cl bottle, than with its much more expensive international brand Tanqueray.
"It's really a new business model," said Chris Goddard, Diageo's innovation marketing manager for Africa, adding that the Cube, shipped from Britain, cost "significantly less" than a permanent factory. Diageo, like the other companies, declined to give exact figures for return on investment.
Success for micro factories is not measured solely on those terms since unit costs can be higher; SABMiller has found that successful micro factories can have a short life if demand takes off.
Other firms appear less interested in the trend. Asked if it had considered micro manufacturing, Unilever said it already had full-scale factories in Nigeria, Kenya and South Africa and planned to expand further but gave no details.
LEGO MODEL VS PARACHUTE
Nestle has been selling food products in Africa for some 60 years, with traditional factories and distribution models. But it told Reuters it now plans what it calls "modular factories" with elements like boilers, compressors and offices shipped in pieces and then assembled onsite.
Depending on complexity, a traditional Nestle factory can take 15 to 24 months to build and cost 30 million to 50 million Swiss francs ($33.5 million to $55.9 million). By contrast, Nestle says it can build a modular factory in less than a year, for about 12 million to 25 million francs ($13.4 million to $27.95 million).
"The concept of modularity already exists ... Lego invented it a long time ago," said Nestle's technical director for Asia, Oceania and Africa, Alfredo Fenollosa. "But for us it was a bit of a real mindset change in the sense that big companies normally do solid stuff -- built for 100 years."
He said the idea was developed by three young engineers, freshly plucked from university, who had no preconceived ideas about how Nestle typically builds factories.
They were given six weeks to come up with a concept and then six months to fine-tune the idea and sell it within the company, where some people needed to be convinced that you could maintain the same food safety and production standards.
"There are limitations to the model. It really is meant for markets that are small, and for processes that are relatively simple," he said. In Ethiopia, it could work. It's big enough, there is business going on, things are moving fast. Sudan is a bit more complicated."
In Abidjan, Ivory Coast's commercial capital, dairy company Arla is producing 25-gram sachets of Dano milk powder in a miniature plant housed inside three 40-foot containers shipped from Copenhagen.
Nicknamed "the Parachute," the facility was up and running in only two weeks, versus an average set-up time of two to three months for a real factory.
"You can parachute into countries where you want to test the waters," said Rasmus Malmbak Kjeldsen, who runs Arla's business in the Middle East and Africa.
Building a bigger factory can cost $8 million to $12 million, Kjeldsen said, versus costs of little more than $1 million for the Parachute, which employs a handful of workers and runs on solar power. If milk powder demand eventually warrants running the factory at night, it will use a generator.
The Parachute is a market penetration tool and it is too early to calculate return on investment, but in an ideal world payback would take between two and four years, he said.
Morningstar analyst Philip Gorham said of the initial costs of building a Cube or a Parachute: "That's a pretty inexpensive way to get into a market ... and gauge demand without any material commitment to (capital expenditure)."
SMALL AND SIMPLE
Africa saw capital investment grow by 12.9 percent last year, according to EY. Sub-Saharan Africa saw a 4.7 percent jump in new foreign direct investment (FDI) projects in 2013.
While micro factories provide investors speed, agility and low risk, they have their limitations.
Kjeldsen says for example that a regular Arla factory can produce up to 14,000 tonnes of milk powder while the micro factory can only produce 2,000 to 4,000 tonnes, which is insufficient to meet substantial demand.
Also, the facility is not sophisticated enough to manufacture the powder, but simply repackages it from 25-kg bags to affordable, single-serve sachets that can deliver one glass of milk when reconstituted.
When entering new markets, Nestle's Fenollosa said simple repackaging facilities let operators train employees and build business without heavy investment. When it comes to full manufacturing that requires lots of different wet and dry ingredients, traditional factories are needed, he said.
Africa's dominant beer maker, SABMiller, which began selling beer in the dusty gold-prospecting fields around Johannesburg in 1895, used small container factories to do pilot tests of a premium version of its Chibuku beer in Zambia and Zimbabwe in 2012. It has since moved to regular factories.
"We put it into an area where we're not 100 percent convinced about the economics," said Mark Bowman, managing director for SABMiller Africa.
"But the problem with it is that as soon as it's going well, we realise we have to make a significant investment and we eventually knuckle down and make the real investment," he said.
Even though the pilot lines only cost $1.8 to $2 million, versus $12 to $15 million for a full line, he said: "These always ultimately cost you much more than you realise."
Still, SABMiller is using them to test lines for Chibuku Super in Malawi and Mozambique and is about to try them out in Botswana. ($1 = 0.8946 Swiss Francs)
(Editing by Philippa Fletcher)