Hacker News Reader

I co-built and ran a successful consulting company that was sold to a client many years ago, am now running another successful consulting. So I've been at this game for many years.

And for what it's worth, tptacek's advice on billing, especially this advice on billing daily, is one of the foundations of our success in this business. When we were just getting started, a lot of tptacek's advice helped us to learn and grow what we are doing, and I don't think we would've been anywhere near as successful in growing, or in understanding the business, without it.

Billing daily is a no-brainer. Almost every client is absolutely fine with it. This depends somewhat on the exact field and the exact client base you have, but if they're not fine with it, it's a sign that you're targeting bad clients and should change your aim.

I recently switched to billing by the half day and it has been a big improvement in my quality of life. I don't think I wanna bill by the full day because some days I wanna be able to take off after lunch, even if I have more work to do, without feeling guilty. On days where I go into and stay in a deep focus and work from early till late, I can bill for a day and a half and feel both productive and uncheated.

Perhaps full day billing is better for people who work on a more regular schedule, but for my nomadic lifestyle I think half day billing strikes the perfect balance. It certainly is a massive QoL improvement from billing in 15 minute increments.

You shouldn't feel guilty about this even a little bit. Consultants are a steal for clients. The alternative is to hire someone full-time, committing to their salary and benefits for years at a time. That you're willing to bill them a single day, and then get out of their hair? Unreal. Your bill rate should be an integer multiple of (your salary / 250).

These are recurring clients and it is less about what is fair for them or how much work they have for me (which is usually more than I want) and more what is good for me mentally and helps me manage the unreasonable expectations bI place of myself.

With half day billing, if I work the first part of the day and have more work to do (that isn't time sensitive), it is easier for me to stop working and go do something fun. When I do work a full day or put extra time for a day and a half, then I don't feel like I gave away my time and it helps me feel fully satisfied with my productivity.

You can still do that with day-rate billing! Your client cares about (a) the calendar date your project is completed on and (b) the total price tag. They don't want to care about anything else. Asking them to care about whether you worked a full or half Tuesday is counterproductive.

> A lot of the [...] perverse incentives vanish when you bill days instead of hours.

I'm confused how that's supposed to eliminate perverse incentives. Doesn't it exacerbate them? Don't you now have an incentive to spread out every client's work over every day possible? Like doing a few minutes of work for each client every day round-robin?

If you bill hourly, you have a strong incentive not to invest in process improvements that will cut the number of hours you bill. Most process improvements worth investing in are measured in at least hours. If you bill daily, you have a lot of headroom to make yourself more efficient before you get to a place where efficiency is cutting into your returns. That's the most mechanical way to respond to the question, but I could keep noodling on reasons you'll work more efficiently billing days than hours.

(If you can bill weeks, you should bill weeks!)

I too see how on paper it doesn't seem to change anything, but I think culturally it's a huge gap.

The incentives are not always to bill the most, and you also might have to pitch a time estimate before starting the task.

Having a day as the minimum unit removes the need to decide if it's a 2h fix or a 5h fix, an estimate that could take 2h in itself to reach.

Then a lot of scrum team will report advancement daily, and it will be way easier to communicate with the client with a billing that fits that cycle (this is also where you probably don't want to report small time sliced increments to the client, if they're already not very happy with your 20 days estimate for changing a page's footer)

> A lot of the pathologies involved in metered billing, time tracking, and perverse incentives vanish when you bill days instead of hours.

Billing daily and not tracking time doesn't make lazy contractors less likely to be inefficient. If anything, it gives them more opportunity to obscure what they're doing.

The only model that really eliminates the perverse incentive to be inefficient is project-based billing.

Project based billing just creates a different perverse incentive of churning out garbage that technically meets requirements

It also makes on fly design changes harder and makes managing scope creep critical.

Not really. You can bid a project-priced project with T&M overages, which is what every project at Matasano did; Matasano was founded by @stake people, which is where we got our project structure from; @stake was Cambridge Technology Partners people, which is presumably where they got it from; and so on down the line all the way back to, I assume, Samuel Adams or Patrick Henry or some other famous Bostonian who came up with the concept.

Just break the project out by week (or by day if it's short) and provide a "total project price" line, and then a clause saying you'll charge overages at a pro-rata day rate. Of course, you can decide whether or not to actually charge OT; if you're OT because of something you did wrong, you're best off not charging, while if the client did something flagrantly wrong, like not kick off the project until 2/3rds of the ostensibly billable time was over, you're well within your rights to do so.

This stuff isn't hard, and I've been a little confused for, like, about 17 years as to how message boards manage to make it seem complicated.

If this is news to you, and you start doing it yourself (you should; it's easy, aligns incentives, and will probably make you more money) you have to credit me, but also cite the whole Apostolic succession (Matasano, @stake, CATP, ..., Paul Revere) I just did.

I think the difficulty in software comes from the fact that it’s very hard (impossible?) to provide estimates because you never know what problems you’ll run into until you do. And so how can you tell “project will take 8 weeks, times weekly rate, for a total cost of X” under these conditions?

A project can have fixed time and cost, but you won't know what the results will be. Hopefully you can define the project so that complete failure is highly improbable.

Like I just said: you give your best estimate, and bill overages at a pro-rata day rate.

The way you phrased this makes it sound like you're in favour of using a design after you've discovered it's inferior to an alternative.

Isn't "churning out garbage that technically meets requirements" a motto of IBM Global Software Services (or whatever they name themselvs these data).

But jokes aside one of my friends seen this kind of setting degrade to "our lawyers will talk to your lawyers about this change request" on every piece of communication between client and big software consulting company.

I'd always prefer project billing to T&M billing, but even if you're stuck doing T&M billing, if you're in tech, you can always eke out an improvement by not billing hourly.

Do you double bill client days? Many client tasks may just be a phone call or emails?

No, I think double billing crosses an ethical line.

I guess what I’m trying to get at is, as a practical matter, there are many tasks that are <1day of work, how do you bill for those tasks? Hyperbolic example: have a 1hr client call, charge for day’s worth of work, pursue non-work interests the rest of the day?