Seth Godin writes that the only two cases, as a freelancer, that it makes sense to charge by the hour are:
- “If the time is precisely what I’m buying … Freudian therapy, say, or a back massage.”
- If the project involves a lot of revisions, so that your time is effectively the thing being purchased.
This falls right in line with my advice on charging for freelancing services. But we need clarification on his first point, and a guideline to his second.
Hourly Rates vs. Hourly Units
Charging an hourly rate for a professional service is not unusual in some industries, but it shouldn’t be confused with being the thing you’re actually buying.
Godin’s example of a back massage isn’t actually the purchase of time. It’s just the de rigueur service unit within the industry.
In other words, masseuses could charge by the body part: $30 for the back, $10 per arm, $15 per leg, etc. But most are selling an experience, and somewhere along the line the industry decided the easiest way to relate that to a purchasable unit was via time.
So it’s not really the time you are buying. Yes, you pay for an hour and you get an hour, but the fact that it lasts an hour is not really what you are buying. It’s like a roller-coaster ride that’s $5 per ticket. If the ride takes 5 minutes, you’re not paying $60 per hour for it; you’re paying for $5 worth of fun that just happens to take 5 minutes.
The better example is Freudian therapy. Or a trial lawyer’s time.
How much effort a lawyer needs to expend on any one trial is unknowable until the end of that trial.
He can guess on the basis of previous case results, his opponent, the judge hearing the case and the initial evidence, etc. In fact, he’s probably going to ask for a retainer based on such an assessment.
But the actual effort required for the trial is impossible to calculate until the trial is complete. It may be that the lawyer winds up spending only a handful of hours working for you; or he may spend twice as much time as expected, making appeals.
That’s what the retainer is for: To provide a base assessment of the likely cost of the project, with the understanding that it may cost more or less than quoted.
This plays to the second point he makes, revisions.
Manage Hours, Manage Expectations
I remain against the idea of open-ended revisions in a project, and I tend to be wary of contracts that require them. It’s been my experience that customers who want a lot of revisions come from two molds:
- They’re never satisfied with anything. While a mercenary spirit might consider that a good thing (unlimited billable hours!), nothing wears me out faster than an unspecified, and therefore unattainable, business goal.
- They have no idea what they are actually after. This is basically the previous issue on steroids: It’s not that the client is never going to be happy, it’s that the client has no idea what “happy” even is. They only know that they are unhappy, and chances are that if they can’t figure out what will make them happy, there’s no way you will, either.
These are bad customers. Trust me: You may make a lot of money off their dithering and complaining, but you and they will be miserable the entire time.
When I write a project contract, I always build into the quote a couple rounds of revisions; and my pricing includes the time I expect to spend on those revisions, as well as deadlines for when they can be requested and when they are due.
Suppose I am making a Web-based store locator application. In that case, I am going to maybe include two rounds of revisions.
The first round will be for correcting information about the store, such as its location, hours of operation, etc.; the appearance of the map and markers; and the functionality of the UI. The second round of revisions will immediately follow the completion of the first round, and it will be restricted to ensuring that all first-round revisions were completed correctly.
I do that because otherwise, it is too easy for the nature of the work to change.
I don’t care if every single stakeholder enthusiastically signed off on your storyboard; I don’t care if the client pointed to an example and said, “I want it to be exactly like this.” You have my personal guarantee that once you show it to the client, he’ll want to change it. Unless you hold his feet to the fire with a revision schedule limited in scope, time and number, that is.
Now, it is true that some work — notably, Godin’s example of a logo, or other branding work — is essentially a back-and-forth. But even so, that back-and-forth needs to be managed, and the best way to do that is through a specific number of revision rounds, each more limited in scope.
That is, maybe there are five revision rounds. The first two are free-for-alls; anything can be changed. The third doesn’t allow for changes to the shapes or patterns, but does allow changes to text, fonts and colors. Round 4 is restricted to color only. Round 5 is to ensure all previous rounds were met.
That said, I’m 100 percent in favor of service contracts. Many professional IT projects require a post-project service contract of some sort. And those almost always make sense to bid out as straight hourly time.
For example, if you sell WordPress websites, you should probably sell your customer two hours of your time, every month for at least two years, to perform routine maintenance and upgrades. And you should make that a retainer. (Most companies are going to want to pay retainers as an annual fee, if they are an annual contractual obligation.)
Depending on the nature of what you’re being retained to do, your retainer can be non-refundable, refundable or bankable.
A non-refundable retainer expires at the end of the retained period.
For example, if I have a retainer for 2 hours of my labor every month, and the client only uses 30 minutes in a given month, the client forfeits the other 90 minutes.
This is an appropriate arrangement for routine service which usually takes a known amount of time. For example, if you need to update a database once a month with new information from a CSV, and you know it usually takes an hour to do that, an annual, non-refundable retainer of one hour per month makes sense.
A bankable retainer allows your unspent time to “roll over” to future periods.
For example, if I am retained for 2 hours per month, but the client only uses 30 minutes in January, the remaining 90 minutes would be credited to February, giving the client 3.5 hours of retained service for February. In theory, this could be unlimited banking, or a bank could expire after a month or two.
This arrangement makes sense if you know that your services are needed occasionally, with the expectation that a handful of occasions will require more time. The previous example of WordPress maintenance is a good example; most months will require maybe a couple theme or plugin updates; but one of the 3-4 major WordPress core updates in a year might require you to invest several hours in re-coding a template file or changing the functionality of a plugin, etc.
A refundable retainer returns the customer’s unspent money at the end of the service period.
This makes the most sense if you are unsure how much support the customer will need over the lifespan of the service contract.
For example, maybe the customer has a year-long marketing effort that you think is going to require about 10 hours of your time every month, so you bid on a service contract of 120 hours. But at the end of the year, you only used 90 hours. In that case, you’d send a check representing 30 hours of your time back to the client.
A final note on retainers: Most clients understand that retainers cover the minimum expected labor, and that all service not covered by the retainer is still invoiceable.
That is, if I am on a 10-hour-per-month retainer, but work 15 hours, I can and will invoice the remaining 5 hours for payment above and beyond the retainer fee I was paid.
Although this is the common understanding, your retainer contract should clearly articulate that hours in excess of your retained time are independently invoiceable, and should also include the rate(s) that you charge for non-retained work.