Hacker News Reader

Comment by krapp | Hacker News Reader

If they were really "top coders" wouldn't they be writing their own agents?

(Altay from 10x here.) Not exactly sure what you mean, but if you're suggesting this is a problem that can be solved with software, I actually think one of our competitive advantages is that we're approaching it as a high-touch, people-oriented, service business rather than a software problem.

Obviously, this makes it less scalable, but I do spend my nights writing software for 10x to automate the things we find ourselves doing more than once. (I spend my days working on deals... still working out the balance; see PG's essay about maker/manager schedules.) And one of the things that's always at the back of my mind is how to maintain the high-touch, personal approach as we scale.

It's just a pun. >_>

And I regret nothing.

:-). I lol'd at this

I got it. :-)

No, no. You have a point. A programmer could simply develop a system that would scrape the web to get leads. Then sort them in a way that the best gigs are on top. Then an automatic email would go out to these leads introducing the coder to the company. Then, depending on the answer, the coder would simply follow up on that introductory email.

Magic? Nope. I have such system in place.

Just out of curiosity, how many automatic emails do you send on a monthly basis?

Unless you have a stable of subs (which I recommend -- to the adventurous, only!), how do you manage the dialogue volume?

At first, it sent around 300 a month, but then I dialed it back to less than 50. Then I added the ability to automatically follow up on people who I had contacted before, but did not a close a deal with (did this monthly, to keep leads warm). I never really used subs, but tried it. It was too much hassle and I'm very lazy.

In terms of dialogue, I searched for key terms in replies, and only replied to those that fit. I did miss on good leads, but the follow up email got me back on their mind and they would always reply.

Are you interested in commercialising that system?

I've thought about it, but the code is shit. Cleaning it up would not take much time, but I'm very busy with like a thousand different projects. This is an old version of the Nuuton crawler (V0.2, I think), and is way buggy. And its overly complicated being that it was a crawling/aggregating system for a search engine. I would re-write it in Python (like I did with Nuuton), and take off all the crap. Use something like Scrappy to make development easier.

Though I wish I could do it with someone else (a co-founder, if you will). This sort of thing could make a lot of money in the right markets. They do exist already, but most are antiquated. The advantage that I have is that I do my own email campaigns. So the emails that get sent out are actually A/B tested for performance.

I do plan to offer the actual emails that I sent to get clients for around $20 soon. Its an easy way for freelancers to get clients without much hassles. If you are interested, then shoot me an email.

OT: Don't you just find bottom feeders, or have you been able to generate quality leads? What's the global vs local leads breakdown like?

EDIT: Finally one, how are you crawling? Aggregating lists or web crawling? Tips?

I used Nuuton to do all the crawling and aggregating, well, the early Lisp system (that was buggy as hell). In terms of leads, I usually got good leads by simply doing cold emails with data gotten from Linked-in. Though I did crawl craigslist, twitter, and other boards.

I don't use it anymore, because I have a long term contract.

Are there a lot of sites on the web that post opportunities for freelance/contract jobs ?

Except for specific sites like elance, this isn't something I've seen much of.

Reply on news.ycombinator.com