We're proud of Karma Go and the company we've built to build, serve, support and promote it. In this series, we'll be telling the behind-the-scenes story of how we got here.

Robert Schouwenburg, Karma's COO, isn't new to startups. His last gig was as a co-founder of Shapeways, a little 3D printing company you may have heard of. He left Shapeways to find a new challenge: "At the moment I feel like I can take a three month holiday," Robert says, "I feel like I'm not needed anymore." When he joined Karma in 2013, not long after the original Karma shipped, there was no danger of three month holidays.

From the start, Karma has had a DIY approach to being an internet provider. We're not slapping a sticker on someone else's product and services, we're creating a unique product and unique services because we want something different. Initially, this extended even to order fulfillment: we shipped our first 4,000 devices by hand before switching to a distribution center.

It was so we could experience the "pain and suffering" of getting devices to people, explains Steven van Wel, our CEO. Figuring out how to deal with lost and stolen hardware as it happens. It couldn’t last forever, but it was a start.

Karma's youngest employee, fulfilling ordersKarma's youngest employee, fulfilling orders

But as Robert explains to me, the supply chain begins far before a FedEx package arrives (or doesn't arrive) on your customer's porch. He likens it to a game: "It's not the hardest thing in the world. But it's hard to do it right."

The game is simple: have inventory as low as possible, but not out of stock.

"What's very important for a startup is managing cashflow," explains Robert. Having cash tied up in unsold inventory isn't good, and having no inventory to sell isn't good either.

It seems a little obvious, now that he says it. But, as someone who often forgets to pay his own phone bill, I figure I can handle a little more of this common sense money advice. Here are Robert's three biggest tips for playing, and winning, the supply chain game:

1. Understand your supply chain early

The big unavoidable constant is time. From the moment you place an order, it can take weeks for your manufacturer to source parts, free up an assembly line, and make your product. Then there's another week to ship by air, or six weeks by ocean. That's just to get it to your distribution center. If you're doing retail, or selling through Amazon, add another week or two at the end.

What you end up with is "inertia," as Robert calls it. "I cannot push a button and make production go twice as fast." The thing you're selling today is the thing you ordered 12 weeks ago. The ordering decision you make today determines what you can sell 12 weeks in the future.

"For most startups doing their first product," says Robert, "ramping up mass production is the hardest part." You know, just building the thing in the first place. So it's easy to forget just how much runway you actually need before you'll have a product in hand. "It's a lengthy process, so if you start too late, you can easily run into a one or two months delay in actually getting into the market."

The thing you're selling today is the thing you ordered 12 weeks ago.

And worries don't end once you've placed that first order: "Maybe your first batch sells out much more quickly than you expect, people have to wait another 6-8 weeks before the next batch arrives, and people lose interest."

It all begins about a year before shipping begins, when you negotiate a contract with the manufacturer. The terms of this contract are hugely important. The contract dictates how many units you promise to buy in total (a high amount gives you a lower price, a lower amount gives you more flexibility in the future), how many units you plan to buy at a time (the "minimum order quantity"), and the payment structure (paying as little as possible up front, and as late as possible in full, keep that all-important cash flow alive).

For a hardware startup, that big contract "is almost like a 'betting the company' prediction," says Robert.

2. Make your own forecasts

So, obviously, in order to draw up this big contract, and make each subsequent order, you need numbers. How much will you sell? And how fast? Don't trust the marketers, is Robert's basic advice.

"The secret, from an operations perspective," says Robert, "is to make your own predictions and forecast. I take input from marketing, but also look at past performance. My forecasts are not always in line with marketing's forecast."

Marketing and sales forecasts are typically "ambitious goals" but won't necessarily line up with reality. A huge chunk of Robert's job is figuring out what reality might just be. He delivers a monthly forecast to the manufacturer, predicting how many units we'll need in the next six months, so they can source components for the device. And then there's those all-important orders for actual devices. You know, the ones that tie up all the cash.

"My forecasts are not always in line with marketing's forecast."

It sounds like a lot of math. And, the more I try to grapple with all the variables Robert is inputting in Excel and ballparking in meetings and signing-on-the-dotted-line for, the happier I am that some people go to school for business and some people only have to be bloggers.

At Karma we place a monthly order for new devices, while a company that orders its product at a higher volume might place an order every week. We try not to have more than four weeks of inventory sitting in our warehouse. Robert is constantly looking 6-8 weeks into the future and trying to predict it.

And the funny thing about predictions? "They're always wrong."

3. Decide on acceptable inventory

Because predictions are always wrong, it's hugely important to figure out what sort of “wrong” is most acceptable. Do you want too much inventory and a cashflow problem, or too little inventory and a nothing-to-sell-to-people problem?

"Have the discussion internally if running out of stock anywhere in your retail channels is acceptable or not," says Robert. As a small company, right now Karma errs on the low inventory side. "I can tell from experience that the acceptability of running out of stock now and then enables me greatly to reduce the amount of stock available in our warehouse in any given moment."

"Did you miss the sale, or did you delay the sale?"

Ultimately, it matters what kind of company you are, and what kind of product you sell. "For instance," Robert explains, "with Apple, if they run out of stock people don't instantly go buy from Samsung. It's a very comfortable position to be in." Meanwhile, if you sell a computer mouse in Best Buy, any moment Best Buy doesn't have your particular mouse in stock is a lost sale — there are plenty of alternatives.

"That's the big difference: did you miss the sale, or did you delay the sale? If it's a delay, it's an annoyance to your customers. If you missed the sale they're lost forever."

These are the sort of decisions that mean supply chain management isn't a "hands-off" position. It requires close attention, and ultimate decision making, from the rest of management. "The decisions are too important to just ignore, especially for a startup," says Robert. "Because cash is king."

Robert also recommends startups find someone with 4-5 years of supply chain management experience: "There's too much money involved. There's too much at stake for the company."

Maybe, in a decade or two, we'll all be 3D printing our gadgets at home using "Shapeways 2000s" or whatever they'll be called, and the trials and travails of supply chain management will be a distant memory. In the meantime, the Roberts of the world will keep playing the supply chain game.

If you enjoyed this, you may well enjoy working with us. Come join the Karma team; we’re hiring.

About Paul Miller

That guy who left the internet for a year