Managing little but important data: a little tech report on our tools.

We’re going to be looking for help to fund a major redesign and technology change on June 12th so I thought I’d explain what we have built so far at Good Night Lamp.

When we sent our first lamps back in April 2015 and until this spring we literally ran the information system linking customer’s information to their lamps on a Google spreadsheet. We just about covered our costs from our first sales so getting a developer to build something early on was an expense we couldn’t afford.

We captured essential information related to the sim cards that help power the connectivity of each lamps (EMEI & ICCID for the geeks out there) and we captured for each customer which lamps they had, Big and Little. This quickly became not so much a nightmare as a Frankenstein of a spreadsheet as we tried to capture the status of each lamp when we were repairing them (which doesn’t happen often but yes does happen). Were they in transit back to us? Were they being repaired? If so when were they received? And when were they sent out with what shipping number. Shipping numbers initially didn’t come into it at all as the first lamps were sent with a courier company who didn’t provide us with unique shipping numbers.

We then tried to capture errors and connectivity issues (yes we have those too) in order of date they happened and trying to keep track of what was happening to find patterns and feed them back to Eseye, our connectivity partner, for them to make changes remotely to the firmware or software they use which we don’t control nor have a view on.

This was all done by me initially but as the months went by I realised I needed help. I had gotten Tom Armitage to work on some technology strategy for the lamps a year before we shipped so I knew he knew what I was trying to do and so we chatted about what a barebone management tool might look like which he built very patiently, with the feedback from Laura who I had hired last summer to help out with what one could consider ‘first line support’.

We’ve ended up with our now 4 month old ‘Manager’ which is a digital space where we can now do all sorts of things. There are 2 ways for us to look at our business information: by lamps or by customers. We can look at a lamp, look at its current status (but not geography) the firmware version and the lamps that are part of its family. We can leave notes which are archived and we can update customer information and add shipping details. We don’t yet capture everything an e-commerce site like Shopify would give us, partly because that changes. People’s shipping address is a terrible way of guessing where a lamp might be as many of our customers give lamps away to family around the world, so we don’t really integrate that information. We simply capture the order number that links to a Shopify order. It also keeps the two systems separate, ie vaguely safer. When you look at a customer view you see the lamps they own, all of them, regardless of the order numbers as we do have repeat customers which is fun and super rewarding.

This is fairly simple in a sense but saves Laura and I quite a lot of time and energy trying to capture what happens to a customer or a lamp. We also invest now more and more in helping people give us the right kind of information very quickly so we can help them quickly too, pushing them to fill in a little questionnaire so we immediately know what lamps they’re having problems with, where those lamps are and what behaviour they are experiencing.

This is something I’m sure we could provide with a chatbot environment eventually but initially it just helps our little team of two to come back to people more quickly.

But now things are changing. We built the Good Night Lamp to use 2G which at the time was the right choice for our customers but 2G is being shut down in plenty of places around the world now as more modern systems are put in place like 4G, LTE and 5G. So we have to re-engineer the product completely and we’ll be going to crowdfunding to help with that.

I’m in Nice today and tomorrow to find out who could help us re-engineer the electronics so we can cover more customers and countries while still using our little organic potato farm style backend.

Changing tools is tricky and requires us to remap how we build good support tools, but I’d be curious to know if anyone has built the kind of tool we have already. There are tons of these ‘middleware’ companies out there but unless you give them a case study like ours you don’t know from the outside if they’ll be useful.

It’s going to be exciting and scary changes ahead. Wish us luck and follow our upcoming Indiegogo campaign!