How I recovered from a giant manufacturing mistake

Andrew Becker founded DAT Media after identifying a solution for companies in the retail sector that need digital signage. By using LCD televisions and other types of technology, DAT provides a digital signage solution that can be changed and altered to whatever the client wants, whenever they want it.

The company is experiencing solid growth, with revenue of $3.5 million during the 2010 financial year.

But earlier in DAT’s history, the company bit off more than it could chew and accepted a rushed order for 100 signs that it had to scramble to fill – without a debugging process. The software broke down, and Becker says the strain provided a good lesson on making sure products are up to scratch before release.

How has the business been travelling?

Business is going very well. We won the tender for the ABC Stores nationally, and we’re currently rolling out digital signage and digital media for them. There is a good amount of work on the horizon.

So what happened when your machines broke down?

At that time, the status of the business was quite good. We were recording strong growth, and we had some major clients who we were providing some in-store music and media solutions to. Through our conversations, we came into a close involvement with one of them, and they were expressing some concerns they had about a particular app.

What was the project you took on?

It was a specific tech solution they wanted. There was a particular off-the-shelf solution they could buy, but they wanted something quite specific in hardware and software and it needed to be heavily modified. And so I asked them to explain it, we were given a brief, and then we put together a prototype.

And how did that go?

The prototype was installed and worked quite well. It was very much a hand-built unit, with off-the- shelf components, we did the whole circuit boards and LED lights and so on. It was what you might call a mad scientist device.

But it was to act as an interface with our existing infrastructure, so that we would communicate with it. It was a very interactive design.

But then what happened?

We won the tender based on the prototype, which met all of the objectives. But once we started to manufacture these on mass it was done by hand. That was when the issues started to present themselves and there were 100 machines to build.

Why did the bugs start to occur all of a sudden?

Well, like anything, you can build a prototype, and get a one-off on a test bench, and then it works fine. But then you put it into a real-life environment, and then bugs start to appear. We had plenty of bugs – manufacturing, specification bugs, so on and so on. The software was buggy in many of these machines and they weren’t operating properly.

The problem was that what they wanted to do changed slightly based on the location. And in retail stores, not all of them are the same. There were variances from cable location to distance, and all of these impact all of the devices.

Was this because the manufacturing process was rushed?

You have to keep in mind that at this time, we were still hand building everything. It didn’t have that continuity in design or high-tech finish that you require from a factory built unit. If you build something and do it by hand, they get impacted by certain things – they were impacted by transport, different temperatures, and the installers at the stores weren’t even trained in these particular devices.

So we had lots of calls and there were installers confused about what it was they were actually doing with these machines. They didn’t have the uniformity of being mass manufactured.

How much of a strain did it put on the business?

You build something, they like it, and then they say they want 100 of them, and we still had one guy building these units. So you had one guy making these components. It was a big strain on cashflow, and we didn’t have the time we needed to test everything properly. It was a very challenging time for the business, but we were able to keep out of external funding by keeping in contact with the company and making sure we stayed cashflow positive.

So how did you approach the problem?

I had to approach it in a three-step process. The first was to clear everything off the board and say, first – do we have the specifications correct? So we had to go right back to the design of the thing, and think about circuit board design, and so on. And then we had to source a manufacturer that could actually do that for us.

The next part was the installation. There were problems there. They weren’t packaged properly, so we had to address that. There were plenty of installers telling us some valid concerns they had, so we had to package them correctly and ship them off as well, and teach the installers how to use them.

What was the solution for manufacturing?

The next focus, of course, was manufacturing. We had a great contractor building these things, but I had to put him aside because we simply needed to get these manufactured somewhere on mass. We found a fantastic manufacturer and came up with a solution, because we really needed to get this products made uniformly.

What lessons did you learn from taking on too big a project at once?

We needed to secure as much testing time as possible, so that’s the key here. We didn’t have enough time to do that. You also need to think of all the other obstacles your product is going to encounter – such as our installation problem. They will be operating in environments not known to you, and that can cause problems. You need to think about what those will be, and then focus on them straight away, not go back and fix them later on.