When is it appropriate to delegate overhead work?

Currently, I have both a start-up and a consulting business (self-funded), and having some trouble deciding when it is appropriate to delegate overhead tasks. The most common advice given to start-ups is to avoid employees and avoid costs. Little is said about deciding when it is best to incur them.

For example, the decision to have an accountant do all my personal and business taxes was a no-brainer: I estimated I would spend 20-25 hours on them and his quote was well below what I would make in the same amount of time.

Typically, I'm working 55-60 hour weeks. A portion of that (5-15 hours, it varies) is overhead in the form of accounting, correspondence, etc. The decision to have a friend build out my company website was more out of a need to save time - it wasn't so clear I couldn't get it done (but at a lower quality) and save money. It's difficult to gauge if I should spend time marketing or hire out, spend time running errands or hire out, etc. While I'm trying to figure this out, Seth Godin posts:

Self sufficiency appears to be a worthy goal, but it's now impossible if you want to actually get anything done.

All our productivity, leverage and insight comes from being part of a community, not apart from it.

The goal, I think, is to figure out how to become more dependent, not less.

...and he pretty much summed up my dilemma spot-on. There's a lot of pride in being able to do everything myself, but there's no problem in putting that aside if it brings me back my free time.

• What are typical no-brainer investments/costs for a start-up that provide a great return on time/effort?
• What are some overhead tasks that should be kept to yourself until you can justify the cost of employees to handle it for you?
• It's one thing to spend 5 hours a week on overhead, another to spend 15 - obviously, there's no fast rule on this, but what's a good measure of what "typical" overhead should look like?

The simplest way I can think to put this is to steal a quote from Rework by 37 Signals :

Hire when it hurts.

In the book, they suggest that you should look into hiring as a last resort and instead seek over avenues to solve your problem:

• Identify whether the extra work is even necessary. If you can get away with not doing it at all, it's not worth hiring someone to do it.
• Identify optimizations other than hiring that can save you time. For example, if you could employ the use of a cheap (or free) software package to save you time over the course of the week, it's a better investment initially.

Essentially, if you've looked into all other options and you still can't keep up with your workload over an extended period of time, it's probably time to consider bringing on some help.

As for what overhead tasks you should push off onto a new hire, it's going to be cheaper for you to push off some of the mundane and repetitive tasks. In addition, if you're the type of person who takes pride in getting your hands dirty, you're going to want to keep the more 'heady' work for yourself, anyway.

In terms of overhead time, I wouldn't use an amount of time spent on the overhead as your test. Instead, I would ask yourself these questions:

• How much time do you spend on actions that aren't directly monetizable?
• How much would it cost to pay someone else to do it?
• How much time would you spend managing another person? (For that, you might be able to get a general idea based on the 3/2 rule of employee productivity.)
• How much more work could you give the new hire that would be directly monetizable by you?

It should be pretty easy to figure out whether hiring someone would result in more money for your startup if you have those figures with which to work.

You may find that you can take half a step toward hiring by trying out some contractors to see if it helps. I like oDesk and Elance for finding people with a variety of skills. Once you establish that 1) it works for you to have someone else doing a certain task and 2) you have an ongoing need for the help, then look into hiring an actual employee.