I’ve decided it makes the most sense to combine a company-search site such as this, with the Recruiterless site. So, Recruiterless will soon start offering a searchable database of IT companies based on both Whereco’s data and the Recruiterless job postings. It will provide company and contact information, usable by IT workers that understand the value in avoiding the recruiter middlemen and doing their job searches directly. The archive will be similar to whereco’s, except that it won’t depend on us IT folks to actually enter any data — we all know how much we love to do that sort of thing.

The question is - what to do with this blog. No matter how strongly I believe that IT hiring needs to get rid of the recruiter middlemen, there fundamentally is a limited amount to blog about it. As you can tell, most of my recent posts are veering off the IT searching/hiring path and are really not appropriate for a whereco blog. Instead, I’ll be starting up a more generalized blog.

So, it’s been fun on this site. A good learning experience all in all. So long, and thanks for all the fish!

Hackneyed but true - hadn’t realized how long since my last post. More coming soon….

Travelling is something I’ve never wanted to do for work. I don’t like being away from home for extended periods, I don’t like living out of suitcases, and I don’t like working in a generic cube that isn’t mine and has none of my stuff. I’ve always avoided it - in fact I’ve both declined job offers and quit jobs I already had because of it.

Until now. I got sucked in because I really liked the business domain and wanted a means to get involved in it, and because of a promise that the frequent initial travel would decrease to an infrequent level as the two merging companies began to work as one. Well, 6 weeks into a new job with accompanying 6 weeks of travel I have to say I’m learning that small things can make a big difference in how tolerable travel is.

In particular, there is one small thing that as far as I’m concerned is a make-or-break item. Something that sounds totally trivial. Something, well, as simple as… paint color.

Yes, I did mean that. Paint color. That’s the key as far as I’m concerned.

You see, I have two choices of hotel. Both are considered to be quite nice hotels. Both are pretty expensive (much more than I’d spend personally), both are good neighborhoods, have good-sized rooms, and have standard amenities. But after one night in one of them, I was ready to refuse all travel unless I could use solely the other. What was it that was so intolerable? It was just the color scheme. That’s all it took.

The first hotel was trying to be “avant garde”, and had a very dark coloring. Maybe they were trying to be mysterious, to be romantic, to be intriguing? I have no idea. All I know is that after one night with dark brown doors, dark brown trim, tark brown taupe walls, and only indirect lighting that pointed down to the dark colored carpet I was many times more depressed about travelling than when I’d started. Another couple of days in there and I’d have been ready to quit completely. What a relief to change to the other hotel with it’s white walls, trim, and wallpaper. I’ll trade staid but bright colored for unusual and dark any day!

Guess I’ll never be able to work in Alaska in the winter.

Now this is interesting. I started scraping both Dice and Computerjobs several weeks ago, for my principals-only IT job board. (I filter the scrape so that only direct postings are actually retrieved.)

I had expected that there would be a small number of direct posts on the boards, but I didn’t know how many since there are so many recruiter ads that it’s hard to take a guess as to how many direct ones are buried there. Well, it turns out that not only is it a small number of posts, but it’s an even smaller number of posting companies because most of the direct posts are multiples coming from the same company. Ouch.

So, now I have to ask… where do the small companies post to find their employees? Surely not Craigslist (at least, not in Atlanta) — that is as swamped with recruiters as the other big boards now. Our local user groups have some direct posts, but it’s not very many. Joel certainly has almost nothing for Atlanta. So…. is it all just word of mouth now for the small companies? I can’t believe they are using the recruiters.

Hmmm…. dunno. I need to find a way to poll IT managers of small companies to see what they really do to find their employees these days.

Oh, my - it’s been a month since my last posting. A combination of new job and additional work on the Recruiterless site used up all of my spare time, as well as all of my spare brain cells.

Regarding Recruiterless, it is still up and kicking. We had a minor revelation on extending the site charter about a month ago. As I talked about in my prior posting, we had a serious chicken-and-egg problem with our principals-only job board — in order to convince IT folks to use it, we needed a serious number of postings. And we needed the eyes in order to draw enough direct employer posts. We could tell there was interest, but getting the site over the initial startup hump was proving difficult (if not downright depressing).

So, we went back to the drawing board… and realized that we can extend our scope and by adding a “filtered scrape” ability. We can add aggregation, as long as we are careful to filter out all aggregated posts so that only the principals-only (recruiterless) job posts are linked. Was this even possible to do? Turns out that it is, and now we are! Yes, believe it or not, there are actually direct postings on the likes of Dice and ComputerJobs… although you’d probably never know it since they are so swamped by the recruiter ads. And we can find ‘em.

This extension holds to our site’s principals-only principles, but it extends our source of postings in a way that both helps us break the chicken-and-egg issue and also (more importantly) add more value for our users. Now we can offer not only Recruiterless-postings, but more general principals-only postings aggregated across multiple sites.

No, I’m sure we aren’t done yet with Recruiterless. There is a lot more to do. But I think we are getting closer to where we need to be, which is nice to see.

So, what are my take-home lessons from all of this?

1. If you start a site where you depend on your user-base for content as well as eyeballs, it is very hard to get over the initial hump unless you already have a known user-base ready to sign onto your effort.
2. You have to be flexible with your site - always check to see if there are other ways to provide value, and always be willing to add them when you find them.
3. You need to be able to add features fast, and be willing to add them piecemeal. No one is going to wait on you while you figure things out, but they may wait for you if they see you are in the process of adding new (and necessary) features.
4. This really takes a lot of work, especially if you are doing it on the side of a normal job. Be sure you really, really, really, really (did I mention really?) like your basic concept and want it to succeed. If you are wishy-washy about it, it simply isn’t worth the effort.

It was bound to happen sooner or later. Someone’s automated spamtool found my blog, and has been postings trackbots to his “buy these drugs for cheap” website. I’ve set up moderate/nuke for the posts (luckily it is filterable since he’s not too imaginative with his verbage), so they don’t show up on the site. But… it’s continuing to hit the site despite the fact that none of the hits is published. Got up this morning to about 70 “please moderate” emails from wordpress.

Looks looks like he’s hiding his real ip. According to the logs, each hit is coming from a different ip address, and none of them are registered with whois. But I did manage to find an isp abuse email address for one of the urls he is trying to direct to.

Hopefully the abuse-report can get him shut down. We’ll see what happens…

After several months of effort on our new (FREE) principals-only job board, Recruiterless, we are finding that we may have to shut it down. Although we have gotten a relatively good response, it has not been enough to cross the threshold into self-sustainability. A site like ours requires a very strong grass-roots support effort, and so far we seem unable to get enough of it.

*sigh*

I think we are having several different problems in gaining enough traffic:

1. We have had significant trouble getting the word out. Almost all of the current public sites that might be used for letting people know about the site already have their hands in the IT hiring pie. There is a lot of money being made by people hosting job boards — both big boards and niche boards. Almost all of the highly-trafficked sites these days (blogs, community, news) have their own job boards which bring in a good amount of money. A free centralized alternative with a goal of hooking up direct job-searchers/employers is viewed as unwanted competition, and this effectively shuts down most of the currently available internet mechanisms for getting the word out.
2. To survive, a job board requires a *large* audience - both of job-seekers and employers. People don’t go to a job board to browse or hang out; they go to job search. Therefore, the audience size you need to reach is pretty big since only a small percentage of those you reach are actually going to be job-searching. If your daily traffic goal is X, you probably need 5000x people to be aware of the site in order to get your desired traffic. That is a large number of people to get involved; more than I had originally realized would be required. And you need this audience size almost immediately, since a job board lives or dies almost solely on traffic.
3. Much as I hate to admit it, there are evidently a large number of people — both employer and job-seeker — that simply don’t see anything amiss with the current 3rd-party recruiting techniques. In my opinion, this approach leads to cog-in-the-wheel ads and cog-in-the-wheel responses… it is a big flight toward mediocrity in the IT business. Although a lot of the people I know personally feel as I do, it doesn’t seem that we represent as large a proportion of the IT community as I had thought we did.
4. The people that do understand the value in going direct may already have significantly evolved networks in order to get around the recruiter-in-the-middle issue. And so they don’t feel a need for a direct site such as ours, or at least not as much as I’d thought they would.
5. IT folks really don’t change jobs all that frequently any more. The bubble days are long past, and people are staying put a lot more than they used to. From the job-seeker perspective, dealing with 3rd party recruiters may be a pain in the neck and may result in you being unable to find the job you really want because you don’t have enough information available to search properly… but it just doesn’t happen that often. So supporting a site that would resolve the problem may not be all that big of a priority to most folks.

It’s a real shame — the current method of recruiter-based IT hiring does not serve either employer or job-seeker very well. But that may end up being the way it stays, anyway.

I went to our Unix users group meeting this week. They had a speaker from IBM discussing their new Deep Computing Capacity on Demand (DCCOD) service.

A lot of companies need a lot of cycles, but they either can’t or don’t want to build out a big enough IT infrastructure to handle it. Grid computing is becoming a big thing these days — many highly compute-intensive programs can split apart their computations amongst a set of machines and then combine the individual results for a final result. An example of this is in the financial arena. Alll of those Monte Carlo simulations to assess trades, positions, and portfolios are very compute-intensive… and also very distributable.

To do big-time grid computing you need a big-time grid. It’s a lot of machines to house, cool, monitor, maintain - a big investment in hardware, personnel, and space. Very expensive to build out. So, along comes IBM (and many other companies, actually). They will provide the machines and cycles. As much or as little as you need, for whatever time periods you want. You can purchase cycles in monthly, weekly, daily, hourly time frames. You get a front-end machine that is yours to put your software on, and they will hook up the grids for you when you request it.

Sounds great! A lot of companies are very interested in purchasing time this way, including some companies that you think might be worried about exposing their sensitive data since the providers are going to great lengths to clean the machines after use. It’s the new, cutting-edge way to handle large computations. Very cool. Except… ummm… well, I have to say it. This sounds very familiar. Remember back to batch processing on mainframes - buying cpu cycles and sharing the hardware? Granted, the software and hardware have hugely progressed and the ability to distribute your program amongst a number of separate, swappable machines makes the whole approach work much better. But the concept of timesharing, purchasing cpu time, and centralized computing centers — isn’t that kind of 70’s?

I guess what goes around, often comes back around. Mini-skirts anyone?

I used to think that anyone that used an IDE was hiding incompetence behind a tool. Years ago, if I interviewed anyone that learned to program Java with a tool such as JBuilder, I was pretty sure that they fundamentally didn’t understand much of Java. They knew how to click IDE buttons, but didn’t understand what was underlying. And if you took their IDE away, they would probably be unable to produce a working program and definitely incapable of producing a well-designed one.

I was the Linux/Emacs junkie at the time, sure in my understanding of the fundamentals because I used only an editing tool. Yes, Emacs is a pretty high-powered editor, but fundamentally it is an editor. I was proud of the fact that my peers were moving to Eclipse, but I remained a die-hard Emacs junkie.

Well, that was then… and this is now. A little over a year ago I finally buckled and tried out Eclipse. I didn’t want to at the time, but I was forced to because somehow knowledge of Eclipse (or Idea) had become an actual job requirement. That’s right, all those recruiter checkboxes included the (proper) IDE experience.

I grudgingly started using it, and over time grudgingly admitted that it could be helpful. The continuous compilation was a big immediate win; it sure decreased the code/compile cycle. Then I started using the auto-completion, and found that it beat the Emacs dynamic abbreviations. The auto-imports came next. Then the refactoring tools - those were sure faster than repeated egreps/etags through the source.

After a while, I found that I preferred the IDE for most tasks. Although for global replaces I’ll still go to Emacs, for most other things I stay in Eclipse.

I accepted this state of affairs. It didn’t mean I was a bad programmer; all it did was speed a lot of mundane tasks up. Right?

Well, I worked on a small programming project this weekend, and decided to go back to Emacs. I didn’t think it was worth the time to bring up Eclipse and start a new project. And then I realized… I was having a lot of trouble coding with the editor! I had to go back to the javadocs to find where the imports were. I was slow grepping the files to figure out where I’d placed subsidiary classes. Moving files around and adjusting their references was a problem. It was hard to remember the exact method names of classes I’d written and wanted to call. Aaaccccckkkkk!!!! I was dependent on the IDE now. I could have finished the project without it, but it would have taken me twice as long. I went back to Eclipse to finish.

I don’t know what to think about this. On the one hand, I’m quite depressed to know that I need the IDE now. On the other hand, I’m quite quick with the IDE, and my understanding of the fundamentals has not changed. Is this a problem? A friend of mine thinks it is - he may be the last remaining java-emacs developer, and proud of it.

Perhaps I should rejoin his ranks….. Naah

This last week I have been going to local IT user groups meetings in order to talk to my peers about the Recruiterless site, both in order to spread the word as well as to gauge how much grassroots support there will be for a dedicated, direct-post only job board. In the process, I saw some pretty big disconnects between the groups.

At the development-related user groups (java, ruby, linux), folks seemed quite interested. After each group got the message, we got hundreds of hits on the site. Evidently I’m not the only person that would like a dedicated venue for principals-only IT job ads. I’ve also talked with my peers that have become managers and my previous managers (all of them were excellent developers and managers) and they all thought the Recruiterless idea was great idea. They saw a lot of value in direct hiring, and in being able to place their direct ads in a venue where the ads wouldn’t be hidden amongst the many generic 3rd-party ads.

So, I was very suprised when I went to a users-group of enterprise architects. These are the folks higher up in the organization, responsible for high-level architecture decisions as well as much of the staffing and managerial aspects. The idea of hiring on an individual, direct basis seemed to be out of scope. I saw no interest in the value of direct posting, or a direct posting venue.

The topic of the meeting, ironically, was approaches in handling offshore development. This actually was helpful, in that it eventually led to general discussion of hiring issues. This discussion was eye-opening. It mostly centered on head counts and finding large pools of applicants to hire from, as well as how to manage the people that eventually were hired. It was pretty clear that the hiring processes primarily involved using 3rd party recruiters, or full outsourcing of the development personnel.

In the midst of this discussion, someone stated that when he interviews a developer he always asks them what they contributed to the bottom-line of their prior company. He wanted to know, essentially, if they could specify what the ROI was for their employment. This was after he had talked about his attempts to “farmsource” his development team!

The whole discussion was eye-opening to me in several ways:

1) Obviously, there is a big disconnect between how the enterprise architects view hiring and IT personnel, and how the technical user groups view it.

2) Most of the other folks in the architect meeting actually argued against the ROI manager. They said that all they wanted was a “coder”. They didn’t care if the person understood the business, end-users, maintenance issues, etc. They just wanted someone to pound out code. This isn’t how most development staff views themselves.

3) Most of the architects discussed bulk and outsourced hiring techniques. No one ever discussed hiring for excellent developers, versatile developers, or in general, any particular characteristics of developers. The discussions mostly centered on producing good software from good management, not from hiring the right people.

4) Not a single person in the room was interested in the concept of recruiting by direct-postings. It seemed to be simply outside their realm. There was no recognition that you can attract a different, possibly better, pool of applicants by this type of hiring approach.

The disconnect between this group, and my previous managers and peers, was huge. This group was composed solely of large and mid-sized companies. Are all the differences due to big-companyitis? I’m left to wonder… whose perception of IT development is the most common.