This was supposed to get mailed out last Thursday, before Friday's email. Drip's RSS-to-email rule was sleeping on the job, so this email (and future ones) are headed your way via ConvertKit instead. Yay for email list portability! Anyway...
Actually, before I move on to the next interesting part of that Hinge Marketing study, let's spend another moment with this chart:The green bars describe high growth firms (>= 20% growth), the red ones describe no growth firms.This chart sure makes it look like there's a positive correlation between several forms of specialization and revenue growth. Specifically:
- Offering highly specialized services
- Focus on solving particular challenge (aka horizontal specialization)
- Focus on helping particular role
- Geographical or regional focus
The outlier is vertical specialization (called Industry Specialization in this chart), which is more strongly correlated with No Growth firms.Should you therefore avoid vertical specialization?Well, it depends.I've long considered vertical specialization to be an ideal first step into operating as a specialist, especially if you have no other form of head start in your business. I still believe this, though the Hinge study suggests it might not be a place you want to camp out long term.About head starts...I was speaking today with someone who bought the The Complete Bundle + Strategy Session of The Positioning Manual.Before he became a self-employed developer, he spent about 10 years working for hedge funds.I asked him how long it took him to become fluent in the business side of hedge funds.He said about 3 to 5 years.That's a massive head start. That fluency is a real advantage when he works with a hedge fund client because he understands their lingo, their terms of art, and how their business works. He's a more productive developer and a more effective strategist because of this understanding.A counterexample: much earlier in my self-employment career, I landed a gig creating training and documentation for a large custom software project.The software was being built to allow a consortium of 9 utilities to manage the generation of power by the 31 hydroelectric dams along the Columbia River. The software would be a sort of client-facing control panel and Bonneville Power Administration (BPA)--the pseudo-government agency that actually manages those hydro facilities--would provide an API that this software would send generation scheduling requests to.The learning curve for me on that project was massive. I was trying to quickly come up to speed on a domain that takes years or decades to fully understand.In my 20's this would have been fun, but now I see it as ridiculous.It's ridiculous because all that learning went down the drain after that project. I didn't make any effort to build on the small head start all that learning represented.I didn't try to get a second and third project where I could leverage that understanding.To use my analogy of the skin: after that first project I was still in the "epidermis of expertise". The outermost, superficial layer.With repeated experience in the same vertical of power generation, I might have been able to get into the "dermis of expertise" where I would have been fluent in the terminology, concepts, and business operations of organizations that generate electricity.Anyway, back to the main point. Is vertical specialization a bad idea because it's associated with No Growth firms in this Hinge Marketing study?I don't think so, because if you want to specialize and you don't have some sort of head start that maps to a horizontal specialization, then vertical specialization is the easiest step forward.You might specialize vertically in an industry you find interesting, enjoyable, inspiring, profitable, or accessible. And once you move into the "dermis of expertise" in that industry, you'll probably start noticing patterns that suggest other, more valuable ways of specializing.You won't be likely to see those patterns without first picking an industry to focus on.OK, for real tomorrow I'll dive into this next chart:-P