Making JuaKali better š
A few months ago I set out to find a way to scale African informal manufacturing. I started with making improvements to the way Fundis (a Swahili name for the artisans) build products. By designing for their capacity and standardizing the production process we built reproducible products using locally available materials. You can read about that here.
We proved you could make consistent and well-engineered products with the Fundis. This was faster and more resource efficient. So, before I tell you about what we did next; Iād love to tell you Why and What in a little bit more detail. The next 2 sections (What & Why) give much more detail and may take a little while to digest. Start with Part 3 ā How and then you can always come back to it after the section January.
Part I - What?
Iām working on scaling African manufacturing. African manufacturing can be loosely broken into two segments: Formal and Informal. Formal manufacturing being large to medium factories employing 10 workers with high worker productivity. Formal plants in Kenya will make use of tooling and technology and can take advantage of economies of scale to make specific products.
Informal manufacturing (also known as JuaKali) are smaller production units making material specific products with a high labour intensity and generally low productivity. They make a variety of products with many being customer specific, their mostly hand-made with variance in the final product. The informal sector employs about 83% of the Kenyan workforce and responsible for 90.7% of total new jobs created in 2019. Many of the products Kenyans use day to day are made by JuaKali artisans, from pots to doors they make almost anything.
JuaKali artisans are incredibly resourceful, with many of their products being made from repurposed material using hand tools they make themselves. Majority of their processes are developed in house with their products being material specific, they segment based on what materials they know how to work with, wood and metal being the largest 2. This segmentation means that products that take advantage of composite materials require different artisans who because of non-standard processes will make inconsistent products both within the same workshop and from shop to shop.
Part II ā Why?
I started off looking for a way to apply what I had learned in University that would have an impact in my community and improve peopleās lives. I spent time with the artisans to get an understanding what problems they had and how I could solve them. From issues with materials to tooling there were numerous challenges that needed fixing. What was consistent though was that products made in these workshops were inconsistent. Two of the same product would have visible visual and measurable variations. That meant any products made at scale would have deviations that would compromise any guarantees on its actual performance.
I set out to prove that with changes in their processes they could make a reproducible products faster and more efficiently. By taking into account what capacity they had and materials they had accessible I designed a product that makes use of different materials made as discrete parts that would be assembled together, that meant working with different ātypesā of Fundis but still ending up with the exact same product in the end. It worked, I could make the same product, faster and more efficiently.
While doing that I realized that if I could scale this up then I could solve most of the challenges I had already identified. Take tooling for example, Fundis use hand tools and simple power tools because the cost of buying specialized tooling is too high. Availing consistent business would boost their productivity and give them the means to invest in their business.
Part IIIā How?
By building best-in-class products at competitive prices. I set out to find and design products that I could make with the available capacity and that people would want to purchase. The products needed to be close enough to what Artisans were already making but are different enough to be original.
I settled on furniture because it checked all the boxes and used different materials in its construction which would allow me to work with different types Fundis. Furniture is inherently bulky so I needed to find a market that was clustered around a small geographical location preferably close to the Fundisā premises.
University students are in this case the best demographic to target. They live in and around student towns and have similar tastes and needs. That would mean I could make designs that work for the majority of them. Another plus was that they are relatively easier to market to, word of mouth can go a long way.
So itās build student furniture at scale to get traction and pivot to the mass market and more diverse products all the while growing Kenyan manufacturing capacity.
Yes, thatās the plan.
Part IV ā January
Once I had the location, market and products figured out I set out to find out more about the Fundis in the student town I chose (Juja). There were many questions I needed to answer, from tooling to logistics, the dynamics of how different workshops built and fulfilled their products would be impactful on what products I could make and how they would be made.
I spent a few days talking to individual Fundis and their employees, seeing how they work and getting to understand the business they were in. It was a lot like the way other Fundis I had met before, with a few differences in their tooling.
Speaking with them was exciting! These folks are a wealth of knowledge. With decades of experience, they have large and diverse portfolios. This was their work but for many it was one part of a fulfilling life. They told me of children, hobbies and moonlighting. One of the Fundis I met has a YouTube channel where he posts videos of shorts skits about society. Check out Njuguna Wa Juja and thank me later.
Once I had a detailed picture of their capacity, I spoke with my target market and sought to understand what kind of products they wanted and needed. My first client was an Architecture student who spends hours designing landscapes on large studio tables. She needed a table to work from; large enough to fit A1 paper which is about 16 A4 sized papers but aesthetically pleasing. We talked about what she needed and I set off to design it to be built to her tastes and to be built to spec by JuaKali artisans.
Now, last year I established that Design-First works ; following a similar process, I made a Computer Aided Design(CAD) model of the table that let everyone involved to see and interact with the product before actual fabrication. We made changes to its dimensions and build from feedback I got from my client and the Fundis. Having the model meant we could make down to the digit estimates on how much the table would cost and optimize for the production process.
Hereās what the 3D model looked like.
I want to go out on a limb and tell you a little bit about the production process and how we solved a problem that would result in orders of magnitude savings in material usage and efficiency. You see, we needed to cut out pieces of material from a larger rectangular piece. We would buy a large board and cut out what we needed. Now, cutting is pretty easy the challenge is the order in which to cut out the pieces. If you donāt get the order right you could very easily end up being unable to get the sizes you needed or end up with A LOT of wasted material
This problem isnāt unique to JuaKali artisans and is well described as a āNesting Problemā think; how many marbles can you fit in a square jar. Toyota has this very same problem when cutting out parts from sheet metal, and so does the baker who makes cookies with a cookie cutter from rolled cookie dough. Solving nesting problems have real world impact, for Toyota that means material savings and for the baker it means more tasty cookies.
Back to JuaKali, because I knew the sizes I needed and the standard size material available in market I could find the best possible order in which to cut out the pieces we needed with the least material wastage. For materials left over to be reused, they need to be in larger and single cross-sections. Ideally, you want one large piece instead of many smaller ones. Using an open source nesting algorithm, I developed the best cutting plan which optimizes for material usage and another slightly less efficient plan that also optimizes for the cutting time.
With the nesting done and the design complete, we started building. The table was built by three different workshops. We leveraged their expertise so parts were made by the fundi who could do it best. While the construction could easily have started all at the same time at the different workshops, it was important I took time to be part of the fabrication process for the first product. By walking the Fundis through the steps they would need to follow to make their part I could ensure reproducibility from product to product. For the most part I didnāt really need to, the Fundis jumped right into it with the occasional clarification on an obscure detail.
What needs to be mentioned here is that when making these products with Fundis, you need to speak their language. They are extremely skilled folks and while some of them have been formally trained, most have learned their crafts on the job as apprentices at other workshops. This meant that standard views and diagrams of the product would only get you so far. I needed to know how they think and provide enough information to make a reproducible product but just enough that itās not patronizing. It was a balancing act and for the most part needed more listening and understanding than anything else
The table frame was first built at a metal-working workshop, the top and drawer was assembled at a woodworking workshop close by. The frame was then powder coated (which is like painting but with plastic instead, the metal gets a more textured feel and is warm to the touch) at a workshop a little further off.
I should mention that on the whole the table took less than 2 hours to build, because we had all we needed at the start getting it together was more of a logistics question than anything else. I spent more time getting the parts together than actually making it. While this is challenge, it means making the product has a larger economic impact(more on this later). This production time can get much shorter because we make use of different shops to make the parts. If the 3 workshops all start making their parts at the same time the time would be cut in half. That does not account for transportation time.
At this point we had our first product! And it looked so good. As a bonus the design is completely disassemble able (flatpack; you can put it all in a box). We could ship it as its parts and needed just a screwdriver and a friend to put it together in under 20 minutes. If you find yourself moving house, then this makes moving it a breeze.
We delivered it to its new home and assembled it. I touched base with the Fundis to find out what it was like to build the table and what design changes we could make to improve the product and the process. They all had brilliant ideas and the next ones will have those changes.
Part V ā Impact š
I want to talk a little about the economic and social impact of this process on all the people it interacted with. Why? Because why else do we do the things we do.
The value chain consisted of 8 people directly employed in the fabrication. Now, this is the Fundis in the shop who were involved in the building and the Boda-boda rider who shipped the product. Indirectly, the local shops I purchased materials from had easily >15 staff. Thatās a total of over 23 people economically impacted by a single product.
Because we distribute the production we distribute the gain. Much like industrial towns support the growth of support industries(think industrial towns like Thika and port towns like Capetown).
Naturally, the question is does this result in a price competitive product? Yes, Its competitive from the start and profitable too. Manufacturing locally is much cheaper than most imports provided you use locally available materials and can scale. Like they say the proof is in the pudding. Its important to compare to the competition and with similar products being priced at over 2X the price its hard to see why not build local(This table cost about 60usd which is around 6000ksh).
When it comes down to it, this only works if the product not only meets customer needs but excites! That you buy a product you love and if its production is impactful then thatās a huge bonus.
Part VI ā February
Looking back weāve again proved that you can build well-engineered and reproducible products with JuaKali artisans. What remains to be shown is if the model can scale and remain profitable and impactful. I believe it can.
We have already started feeling the growing pains, the products we want to build here require fasteners and tools that are hard to acquire locally. We need to get the tooling to make the products and the tooling to make the tools.
Thatās why this month we go to market and allow you to purchase our products, yes thatās in plural. Weāve been busy. There are new designs complete that youāll love. You can check them out here. We know that buying these products is you believing in us, thatās why any products purchased this and next month will come with a years guarantee. Donāt love it? Any issue? Itās on us for free!
Itās clear we will need to get some space and financing at some point so if you know any early stage funds, incubators or accelerators please send us a link here.
Part VII ā Thanks
To everyone who is a part of this, it means everything. Thank youā¤ļø
