>Lots of long-term freelancers will drop the hourly stuff ASAP too.

Hm. See, I've been thinking about this. I'm a sysadmin. A computer janitor. The essence of my job is that when something breaks, I fix it. (I'm actually way better at that part of the sysadmin role than at the architecture part, though I can do that, too; there are many sysadmins who are better architects than I am. However, there is this phenomena; the more senior a sysadmin is, the harder it is to make him or her carry a pager. I carry a pager, and in /that/ set, well, I look pretty good.)

So yeah, what I'm charging you for is something you probably won't need much at all of (I mean, I'll help you apply patches and do backups and other simple bullshit, but I'll probably make my PFY do that, or train up one of your kids.) My real value is that when something breaks at 3am on a Sunday morning, I'll wake up and deal with it. Really, unless you are twitter, I've got enough spare hardware that your stuff could literally catch fire and as long as I've got backups, we're good.

Problem is, how to charge? I've got a bunch of folks wanting to pay me $50/month. Which I could do if I automated all the PFY work. (which, I could do if I didn't give you root, and I only let you use my versions of the programs I chose to support.) but it's one of those things where I'd need thousands of customers to make it worth my time. At that point, we need to do shifts, etc, etc... I mean, it's a realistic business, but a deep/hard one.

I mean, I'm going to want like two to four grand for the two or three days of my time that a full "shit caught fire at 3am" cleanup is going to take (not that you'd be down for two or three days; I'll get you back up the first day, but I'm not going to be particularly useful for the next few days... which presents something of a problem if these things happen in succession.)

But that's the thing, I can do a better job if I'm there all along, and I can say "don't do that" when you are setting the thing up, and I can make sure you have accessible backups and stuff.

I've sent out some feelers at the $500-$1000/month range, were I keep a spare going (which means more ongoing work for me, keeping the spare up to date, but it means much less pain when shit does break.) But, eh, I haven't gotten interest, and frankly, I'm not sure if that's enough to make it worth the brainspace. And that's the thing, if you get up to $4000-$5000, well, you can get a midwesterner or an indian or a russian mostly kinda sorta full-time, and yeah, at that rate, they aren't going to be as good as I am, but there is a lot of value in having someone mostly full time (which I would not be, even at that rate.) So maybe the right model for me is "wake me up, place a five grand deposit; I keep the money if I fix your shit in the next X hours" or something. I could be talked into that, if I had some reasonable assurance of getting paid.

I think the root cause might be that most small companies use "the cloud" and don't feel they need that kind of sysadmin experience (and spare pool... "the cloud" vastly diminishes the market value of my giant spare pool, probably even more than it does my SysAdmin experience.) and larger companies would prefer to have someone on-staff full-time.

Also, well, I'm very good at marketing "value" - e.g. most of my experience is competing on price (yes, yes, I know I'm behind the curve. I am working on it. You do not understand how much I am working on it.) and, well, this is not an area where I want to compete on price. I'm pretty okay getting woken once a week. If I get woken 3 times a week, for more than one week in a row? things start to get bad. Wake me up every night? well, I've had times in my life when all I did was sleep and respond to emergencies; I can do it, but I'd really prefer not to go back there. So yeah; I have a very limited number of "wake me up" slots I can sell. They need to be expensive.

> So yeah; I have a very limited number of "wake me up" slots I can sell. They need to be expensive.

Retainers, man, they're wonderful. You can oversell to a small extent, which is unlikely to make your life fun when everything falls over on the same weekend, but is otherwise very lucrative.

Figure you oversubscribe by 50% and split half of the difference with the customer in savings - $1500 * 8 customers for what would normally cost them $2000 for your time, or something like that. You can arrange it so at most you're waking up 4x over the weekend, mean of 1x and median of <1x.

The way I've done it in the past, it's a flat fee for having your time available, and then X hours at a substantial discount, further hours at the normal flat rate, that way they have an incentive not to burn your time unless they need you - I used to just do 10 hours a month for free with retainer but then you get mickey mouse work just to use the hours at the end of the month.

> and, well, this is not an area where I want to compete on price.

Definitely not, since you are local to them, you speak their language, you can be there to work on their stuff, and you are part of their network which they can hit up to find other useful people.

I have several friends who run local IT businesses or do IT freelancing. One thing I've noticed is that it's very easy for them to make enough for their needs by maintaining systems, but they all understand that it's optimal to be seen as a profit-builder.

When I was doing freelance IT work (for a short period while my web stuff built up) I noticed that I could often do cursory research on the fields my clients worked in, and easily pitch them on things that would help them out. For example, I would say, "hey, have you seen this new CRM software for non-profits" and "hey have you seen that this web hosting firm offers free hosting for non-profits" and I'd build projects off of that--rolling in a system upgrade, doing in-person training, etc. If I were to go back to that today, I'd make sure they knew that they needed to pay me an extra $ monthly to support the new cool thing. People are used to that and want to invest in profit-makers. Both sides benefit.

One thing I can highly recommend is giving your clients "good, better, best" estimates. For example tell them, "I can fix it with duct tape for $, I can keep this from happening again for $$, and I can hook up this cool new solution called X for $$$." You might be surprised how many clients will go straight for the top. The clients who always stick around the bottom will become more obvious about it, and you'll learn to work with them more effectively, too.

The pager is quite an interesting thing for sysadmins. Your value is because you can fix what breaks when it breaks, or when you're so senior that people only bring you in for solutions architecture and setup.

In the corporate world, you're not ever the only guy on call, so you only get woken up for a week. If you're full freelance, you have flexible hours after some late night work, and you have multiple pagers, so it's less of an issue to add one. I haven't figured out how to rectify side work and being full time corporate with the customer wanting an on-call break-fix guy, so ultimately I've just said no to those requests.

>In the corporate world, you're not ever the only guy on call, so you only get woken up for a week.

oh man. the /worst/ pager I've ever been on, by far, was at a fortune 1000 in the mid to late 2000s.

Yeah, it was only one week out of four (seven days, not five,) and only 12 hours a day. When I heard that during the interview, my response was something like "Twelve hours? I can do that standing on my head"

I mean, the thing you have to understand is that the pages that really hurt are the ones that come when you are tired already, right before you are planning on sleeping... you are most likely to mess things up then, and it interferes with your sleep, which means you are more likely to screw things up the next day. So limiting it to 12 hours is pretty goddamn nice. You have 12 hours that you can sleep, which sounds super easy.

So yeah, I've been on 24x7x365 pager for more of my life than I haven't... but the worst pager I've ever been on was only 12 hours a day for 7 days a month... why?

For most of my career? the pages that actually wake me up have a frequency between once every month or two and twice a week. Sometimes it gets worse, but then you look at what the root causes are, you fix them, and it gets better.

At this place, though? during that 12 hour shift, there was a serious (like 'fix this now or we lose tens of thousands of dollars an hour in revenue serious) issue every 30 minutes. Most of it was not technically difficult, but you had better be fast.

Essentially, you'd work one 84 hour week out of four. And actually work those 84 hours; like literally you had to find someone to cover for you when you had to take a shit; (or you just bring your laptop) it was that bad. (On top of that, then you had your usual architectural responsibilities... but realistically, you didn't get any of that done during pager week, and you were usually pretty useless for the next week or two. This is one of the reasons we didn't have more of it automated.)

so yeah. In corporate land, yeah, you usually get shifts... but if someone fucks it up in corporate land? (In this case, there was... cultural pushback against automating the failures. We were supposed to diagnose hardware problems as we went. Hahah. Yeah right.) things can get really nasty, really fast.