Because I enjoyed running the software business more than I enjoyed running the consultancy. Partly that is for difficult-to-explain personal reasons (I feel the sense of satisfaction with operating a software company that some people feel when painting things). Partly it is because of reasons which are endemic to consultancies, even successful consultancies. Feel free to ask if you're curious.
I also have a quirky relationship to money as a goal. Rather little would change about my life if my salary increased by $800k. My hobbies would remain books, video games, taking my wife and daughter on walks, and posting too much on HN.
I'll take a stab at the "why not build a consultancy business"...
It's not scalable in any shape or form. Despite having aspirational goals early on, every consulting company falls into the trap of becoming a body-shop. It happens every. single. time. You begin to feel less satisfaction about your services and proposals to clients because you emphasize "our talent speaks for itself" and it's polarized by your clients desire to simply want to pay X for Y and don't care how, what, or who delivers it. Being a one-man shop means that right from the beginning your client's expectations are set on you being the right man for the job and they'll get Y because they're going to pay you X. At scale (I'm looking at you IBM, Accenture, etc) you end up fighting with cost center budgets, and spend more time trying to squeeze quality and scope within a set budget, or risk overspending budgets and end up looking like a bad services partner. Building a consultancy sounds awesome at first, looks awesome at first, and then it slowly burns you down.
Another way of saying this is that scaling a consulting business involves a peculiar set of challenges which may or may not be animating for people like Patrick. :)
I enjoyed the process of scaling Matasano; I only left because Erin and I were more engaged with the idea of making a dent in recruiting.
Things one does to scale consultancies:
* Refine a recruiting and hiring strategy to make it easier to staff larger number of projects.
* Build ramp-up processes to take new consultants and get them to "billable" status quicker.
* Play with service definitions and rate books, so that clients that need Patrick-level delivery can pay extra to get it, while clients that just need deterministic competent execution of tasks Patrick was good at 5 years ago can pay less to get someone without the name brand.
* Develop systems to cross-train consultants so that more people get an opportunity to (a) watch Patrick deliver a project, (b) deliver a similar project themselves, and (c) help bring other consultants up to speed.
Because revenue is so directly tied to performance and delivery in a consultancy, I think it's a pretty good lab for learning how to effectively manage people.
It's also true that a lot of them just turn into body shops.
Horizontal scaling in consulting seems to be what you're so put off by, but I think at most lower levels, vertical scaling is still very possible.
For me, at least for the forseeable future, I'm pretty solidly convinced that I can earn more money by studying and improving my tech skills than I could ever earn by taking a side-job that actually pays me dollars in my hand.
Maybe one can "max out" and just know enough to get their job done, but I'm not there yet -- I frequently find myself ditching other opportunities for making short-term cash in favor of learning something new. That's why I would favor "building" over running a less profitable software business.
Not snark... serious question... its not clear to me what sort of services anyone could offer for 30k per week... what sort of thing commands that kind of price tag?
Making $20 million a year SaaS companies into $22 million a year SaaS companies. If you have reasonable expectation that I can actually deliver on that, then you're pretty much willing to say yes to any rate I quote.
If you're curious "How does one turn a $20 million a year SaaS company into a $22 million a year SaaS company?", there exist many options depending on the specifics of the company and virtually all of them are doable. For example, if you increase the rate at which the SaaS company closes deals by a modest amount, that could reasonably hit that by itself. You could do that by some combination of A/B testing, better email marketing, tweaking sales processes, etc, alone or in combination. I've blogged pretty extensively on all those topics if you're curious.
These are not necessarily all that technically complicated. If you can figure out ActionMailer, an if statement, and a for loop, you have every technical capability you need to be an engineer in charge of email marketing. The interesting bits are "What emails do I decide to send?", "What words go in those emails?", and "What goes in the if statement?"
The actual work is exactly as dominated by meat-and-potatoes execution ability as you think it is.
What books can I read to, at the very least, attempt address the questions in your 3rd paragraph, given how unique every business is? Isn't anything you find a book on sales or marketing so abstract that it's 2 or 3 layers of assumptions before you can talk about projections? How does one begin a career in SaaS sales / marketing? What makes for a good "portfolio piece" in that domain?
One of the things that has made Patrick a "tech. celebrity" is that he has written extensively and openly about this very topic. In particular his brand of uniqueness was to take a lot of things that engineers don't like and then turn it into an engineering problem. He didn't guess, he tested.
Here is the amazing thing, he also did all of us a huge service and wrote it up and put it on the internet for everyone to find!
If you haven't yet, spend as much time as you can diving in his blog. Every post he's written is worth the price of admission, but you could definitely limit your search to the ones that focus on the SaaS sales and marketing aspects.
I wouldn't call Patrick a 'tech celebrity', I would call him a 'marketing celebrity' in tech. His writings on how to market yourself are without doubt the best essays I've ever read on the subject.
Those are the standard skills of marketing people I've worked with, getting paid around £30000/year. I find this very peculiar, maybe the companies particularly dysfunctional, but there seems to be a missing variable.
Strategic advisory work. As an industry analyst, I would get paid up to about $18K per day from large clients. (Really a day plus travel plus prep plus some follow-up. And sometimes two of us would do the engagement.) But, still, $5K-$10K per day although we didn't get that level of engagement all that frequently. $10K-$20K is also the rough going rate for speaking at sales meetings, etc.
$2K or so per day was more routine for reports on a variety of topics (competitive analysis, strategic options, etc.) Expert legal work is probably in the $400-$500/hour range.
I can't give any details, but the largest bill I've seen was $1M for ~2 person-weeks worth of work; it was clearly demonstrated ahead of time that it would save the customer about 10x that, and the customer had tried twice already to do it in-house and failed (it was way outside their core expertise).
Imagine you are running a business making $30 million in annual revenue. You run into someone that reliably assures you they are able to increase your sign-up rate by 1% by optimizing your sign-up flow, SEO, and other general magic. This 1% increase in sign-up rates is easily worth $1 - 3 million to you. As a business man, the only question in your mind should be "When can you start?"
Overall, is it that being closer to the product feels better? And did you really shake all the "weird feelings"/impostor syndrome about charging a high rate?
I also have a quirky relationship to money as a goal. Rather little would change about my life if my salary increased by $800k. My hobbies would remain books, video games, taking my wife and daughter on walks, and posting too much on HN.