The setup
acareer.com started from long conversations with my mom. She had a degree in marketing and advertising, and she placed job ads for clients for a living: a client would call, define the role and their budget, and she would translate it into the tight, acronym-heavy phrasing of a classified ad and spend their budget across the right targeted newspapers to get them hires. My dad, meanwhile, worked for TRW at the NUMMI plant in Fremont, the site that later became the Tesla factory, maintaining the assembly robots after being trained in Japan. Listening to how my mom worked, it clicked. This whole thing belongs on a website.
We named it acareer.com very deliberately. The main way people found new sites was Cool Site of the Day and its alphabetical listings, so a name starting with “a” sorted to the top. For a sense of how early this was, I had been an early IRC user, back when gopher and FTP were about all there was.
What I owned / built
I built the product. We set out to make a dynamic, database-backed website for job placement, the online version of what my mom did by hand: employers define a role and a budget, job seekers find and respond to listings. As I recall it, the back end was Microsoft SQL Server, the front end HTML, CSS and JavaScript, with Perl CGI generating the dynamic content. A database-driven web app in 1996 was not a solved problem; you built most of it yourself.
The hustle: keeping the lights on
Arsenal was lean and self-funded, so we ran several lines at once: one-off websites, mostly static, for small businesses; reselling bandwidth, since we had dropped a T1 into our building and sold access to other tenants; and consulting, which became the main thing after acareer. We also did a fair amount of volunteer work, including running the ethernet cabling for the Oakland Habitat for Humanity office.
The fundraising
I pulled down one of the famously short, legendary business plans as a model, wrote ours, and went around the Valley. In 1996, fundraising was often literally “drop your business plan in the basket and maybe we’ll call.” We had a handful of real conversations and one genuinely good one. The hard part was not our idea; it was that we frequently had to sell the VCs on the internet and websites at all, since most of them were still focused on biotech and hardware.
Outcome
We launched acareer.com and got almost no traction. We had been early to exactly the right idea and watched it get capitalized elsewhere. I pivoted Arsenal into a consulting firm, which is what carried us: contracts at Bank of America, Intel, and then Silicon Graphics, where I brought in and grew an onsite team to four people including myself. The company later redirected its product ambition again, from acareer to Axis, which has its own entry.
People
- Dan Wascovich, co-founder and partner.
- Seed capital came from family.
- My mom, whose job-placement work was the entire origin of the idea.
Tech & craft
Dynamic, database-backed web development in 1996, as I recall it on Microsoft SQL Server with Perl CGI, HTML, CSS and JavaScript; static site building for small-business clients; running and reselling a T1; hands-on networking and cabling.
The story
Our office was in a building called the Fashion Mart in San Francisco, right across from CitySearch. It was the scrappiest possible version of a startup: build the product, build client sites to pay rent, resell your own bandwidth, wire up a Habitat for Humanity office on a weekend, and pitch VCs who needed convincing that the web was even a thing.
Throughline
First time founding something, first time raising money or trying to, first time being right too early. The pattern of seeing where things are heading before the market does, and of building the practical thing that funds the ambitious thing, starts right here.