Over-engineer to fail an online coffee business

Ten years ago I started a coffee roasting business in a good market with minimal investment and good margin, yet ultimately failed in terms of making money. I did that with lots of engineering but not enough business efforts. I would like to share with you my story and learnings, as they are relevant to whatever business that we are doing. Especially to entreprenuers with a technical background.

Photo by Paul Hanaoka on Unsplash

A Roast-on-demand Instant Coffee Business

It was around 2009, gourmet coffee houses were blooming in Asia cities. I moved to Taipei wanting to start a coffee business. As much as I love hanging out at cafes I never thought of starting a physical cafe, but something more scalable like a coffee roasting business, one I could sell B2B to cafes or B2C by fulfilling online orders.

There was a very popular Japanese invention called “drip bags”. They are basically sealed coffee filter bags filled with coffee grounds. You unfold the clever origami filter, put it on top of a cup, and make a good brewed coffee by simply pouring hot water over.

These coffee drip bags taste much better than soluble instant coffee, yet still relatively convenient. They are usually massively produced and sold in supermarkets and convenient stores.

Yet for coffees, freshness wins. Most of us have been served bad coffees so we may not realize what freshness is, but once you try a coffee brewed from freshly roasted coffees, you’ll remember the difference.

It’s like bread. Freshly baked vs five days old, you can taste the difference, and once bread is stale it gets moldy and you throw it away. But unlike bread, roasted coffees don’t get moldy when packaged, they just don’t taste good. So its perfectly fine for us to drink non-fresh coffees, they just, well, don’t taste good!

Now these wonderful drip bags are mass-manufactured in factories, sitting on retail shelves waiting to be enjoyed, and they are usually outside of the two weeks prime-taste period. They are not expired, just not as good.

What if we could manufacture drip bags from freshly roasted coffees? That will be a good blend of taste and convenience.

The Challenge

Now we are setting up a roast-on-demand coffee drip bags business. Ultimately we need to:

  1. Take orders
  2. Roast coffees
  3. Individually package the drip bags
  4. Ship

Taking orders and shipping are technically easy, I just built a website with simple order management functions and be done with it. Back in the days I I did that on Google App Engine.

Roasting coffee is another story. Roasting in a controlled heating process that lasts about 12 minutes, from room temperature to about 220 degree celsius, following a temperture curve. The process is highly temperature and time sensitive, an extra 30 seconds of over-roasting would totally burn an otherwise perfect batch. In other words, it needs repeated precision.

Packaging is labor intensive. It involves grinding the coffees, weighting them, individually packaging and sealing the drip bags.

Suppose we finish all these, this product yields really good gross margin. Drip bags retail for ~0.5 USD with coffee grounds that cost ~0.03 USD. With material costs its about 80% gross margin. Especially I’m playing the boutique roast-on-demand game, there is much more room to play with pricing.

Photo by Ochir-Erdene Oyunmedeg on Unsplash

Unfortunately, equipments are expensive

Yet equipments are expensive. Correction, requipments are not available for small scale manufacturing. Since I was doing a roast-on-demand business I needed to roast about half a pound of coffee per batch. There are small home roasters but they are not temperature accurate enough for consistent roasting. Batch after batch.

So, I built one. It is basically a 4000watt computer controlled heat gun that changes temperature according to a curve via PID controls. If you have an engineering background it is quite straight forward. Quality is by no means an industrial grade roaster but it serves the purpose.

Individual packaging is even trickier. Drip bags is such a niche product there simply aren’t ‘home-use’ equipment available. So I have to either hand pack, or hire a bunch of workers to hand pack, or home-brew a solution. Of course I did the latter.

Shot on iPhone 3GS, sorry for the low-res!

I rigged a coffee grinder, CD-Rom drive, and some metal parts to make this packaing solution. It packages and seals a drip bag every 7 seconds, and the machine retires after about 8000 iterations.

I did the MVP of small scale manufacture

Over engineering is bad. The goal was to start a roasting business so I did the absolute bare minimum to home-brew a roaster and packaging solution and I was off selling. I reminded myself not to perfect these equipments, avoid the fun and engineering high, focus on getting orders.

As a solo entrepreneur, I didn’t have much marketing budget to play with. And it was a product that you had to experience to taste the difference in freshness and convenience, so I worked with local cafes and offices to host workshops and events.

I got some orders from events I host, and usually people bought a box of 10, and gradually they come back to place more orders.

Things seemed to be working but then I realized it usually took about a month for customers to place a second order, guess people don’t drink coffee as fast as I hope. So it was growing, but very slowly.

So slowly to a point I could literally get around with fulfilling orders without home-brew automated equipments. In other words, I spent three months or so building a MVP that was still an over-engineered effort.

Putting business before engineering

Long story short, I was such a precise and inexperienced entreprenuer so I did all the math and aimmed to grow and profit enough in short months. Of course business did not happen as planned, so eventually I gave up.

In retrospect there are many things I could have done differently and in different orders, but business strategies, calculations, and marketing aside, the most important lesson is, are you not over-engineer when you think you are already doing to bare minimum?

While it is fair to say I did not over-engineer on the flimsy roasting and packaging solution, from a technical perspective. But if we look at the business as a whole, it WAS over-engineered, as that wasn’t even a large enough orders that required such automated solution.

I could have saved months to gain traction and learn from customers before getting high in automation. My MVP wasn’t minimal enough.

I’m an engineer and not to undermine engineering, but we should consistently ask ourselves what is it that we are making. It is so easy to nerd away and think we are spending the right and minimal efforts on building the right product, but is it? Maybe the product is right, but business isn’t, or the marketing isn’t. And from time to time, we should step back and look from the business side, what is it that we need to build so we could grow sales.


The more technical write-up for the coffee roaster I built.

Part-time gamer. Full-time indie game developer.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store