March 26, 2011


Tags: ,

I quit my PhD studies and gave up the associated scholarship at the start of September 1997. I needed to find a job fast and I wasn’t sure how to go about getting one. A friend suggested some agents and I was soon interviewing at various places. The first couple of interviews went badly (including the agent sending me to one for which I was completely unqualified – at least it was short). I began to calculate how long I could cover my rent before resorting to selling stuff. Then in late October I found myself talking to Alain at the agent’s offices. It was supposed to be a proper interview, but Alain’s company didn’t have any offices and it was more a friendly chat than the interrogations to which I was now accustomed. He offered me a 3 month contract starting a couple of weeks later. I accepted.

Alain was a one of three co-founders of Electrolley, a company which at the time existed only in name. The plan was to create an online website for grocery shopping. A small supermarket had agreed to supply groceries and now they needed the website. Alain had a technical background in database development. However, he kept his full-time job at another company didn’t have much time to spend on Electrolley. The other two founders, Tim and Dennis, had a successful background in business. They would provide the funding and day-to-day management, but had no coding skills. As Electrolley’s first and only employee, my task was to create the website.

In early November I turned up to my first day of work in Alain’s garage. Admittedly he had converted it to an office, but it used to be the garage. I was sat down in front of the computer and Alain showed me what he had created – a simple product list page backed by a product database table – all created in a Borland product called Intrabuilder. It was a Javascript based development environment for websites combined with its own webserver. Intrabuilder no longer exists now, disappearing a good decade before server-side Javascript became cool (as it now seems to be with Node.js. I was also given a book to self-learn, I think it was “Teach Yourself Intrabuilder In 21 Days”.

Luckily it took me a lot less than 21 days to learn and Electrolley was largely complete by February. Customers could login, browse products, create lists of favourites, put products in their shopping cart and checkout, with various admin tools for us. It wasn’t pretty, but it worked. By then I worked in a proper office in South Perth, near our associated supermarket. Alain was largely absent, but Tim and Dennis came in every day to run the company. I liked them as people and envied their lifestyle. I used to joke they only started Electrolley to have something to do in their early retirement. They would come in later than me, leave earlier and have longer lunches. In their defence there wasn’t much for them to do at this stage. The business model and marketing had been worked out, they just needed the website finished.

Once Electrolley started trading at the beginning of 1998 there was, for a time, even more to do. In addition to the inevitable bug fixes and new functionality, I also handled the day-to-day operations, working as a system administrator (Intrabuilder only worked on Windows) and liaising with the supermarket. I remember sticky-taping a plastic cup over the power button of our production server after it was accidentally bumped a couple of times. Our partner supermarket also had no net connection. When an order came in I had print and fax it to the supermarket to handle. Also a few times a week I would walk round to the supermarket and download onto floppy their product list and pricing. It was on these trips I found a taste for a lunch of chicken & chips or a chicken roll from the fast food place next to the supermarket. I gained weight fast.

The business process in our office was normally quite smooth – problems tended to occur once the order was faxed to the supermarket. The supermarket would pick the ordered items and then leave them for a courier to deliver. Unfortunately, there were often errors in the picking – unnecessarily missed items, or just plain wrong items. If this occurred Tim or Dennis would drive around to the customer with the correct product and apologise. It was decided we needed to track that the proper products were selected. To that end I wrote a program for a handheld barcode scanner that ensured the correct items were picked. As the product was taken off the shelf, the picker would scan its barcode – any illicit substitution or something missed would be flagged as an error. It was interesting little project as it was lower level coding than I’m usually doing (memory and speed was severely limited). However, the plan fell apart because the cost of the scanners was too great considering company revenue.

Although growth was slow and steady, not enough revenue was always a problem. The profit growth line was linear for most of the time I was there, but we needed to grow much larger in order to negotiate better pricing on groceries and deliveries. The couriers were a particular issue. Because we were only making a few deliveries per day, it was too many to do ourselves (I didn’t even have a car at the time), but not enough for a dedicated delivery service. Standard parcel couriers were expensive meaning our delivery charges to customers was often 10% of the total bill. Our margins on the groceries were so low we couldn’t absorb the costs and thus we were noticeably more expensive than local supermarkets. Still, we found a reasonable number of people would pay for the convenience of home delivery. Although Perth’s low population density played havoc – it is a very spread out city. The further out from the supermarket we delivered, the higher the delivery costs. While the catchment area was large, the population of that area was low. To grow, delivery costs need to decrease or more supermarkets needed to be added. I know the owners were working on both, but by the time I left there was no progress on either.

Towards the end of my 9 months at Electrolley, work was slowing a bit. Bugfixes were becoming rarer and new functionality was problematic. Intrabuilder was a product ahead of its time. While server-side Javascript is hot right now, back then it was looked down upon (it’s still not on my CV), mainly because it was slow – very slow. Maybe I could have done something to speed it up, but I wasn’t experienced enough. Instead I recreated some core functionality using Java Servlets (this being before JSPs) and it was nearly an order of magnitude faster. The owners were rightfully concerned with the possibility of a rewrite, but in my naive innocence I thought it was the only way to go. To keep me coding and happy, the owners had me write a classifieds system, but they never put it live (as far as I remember).

After that my motivation began to slide. I had calculated the profitability of Electrolley and extrapolated based on the trend. Our gross profit wasn’t even enough to cover our Internet connection and marketing, let alone me (the firm’s greatest cost) and wouldn’t for many years. I didn’t foresee a rosy future. At the same time some old university friends were saying how great things were at Rio Tinto, and that they were hiring. I asked the owners for a large pay rise – not quite double my initial rate. My housemate, Leigh, said this was the market rate at the time and my research suggested it was possible, but optimistic (although that was for an experienced developer, not one with only 9 months under their belt). The owners balked and talked instead about the options I would be given when Electrolley was a success, as they had on many previous occasions. My thought was that options in a failed company are worth nothing. I applied to Rio Tinto, got offered a job and handed in my notice to Electrolley.

I left in early august 1998, after a week of training my successor. A year later the site was relaunched after being rewritten using Microsoft ASP technology. For a while they became moderately well known around town, but they soon seemed to disappear. I know they survived the dotcom bust, as I met with Tim and Dennis in early 2001. They were beginning to specialise in bulky high-value products like nappies and flowers, but I guess that didn’t work out. I can’t find any evidence that Electrolley still exists. The old web address goes to a domain holding page and I haven’t heard anything about them for a long time. It was a good first job, I learnt a lot.

comments powered by Disqus