Regular readers know that I’m obsessed with the Future of Work — a catch-all phrase for the many ways in which the global labor markets are being reshaped by technology.
One visible element of this trend is the acute scarcity of digital creative talent, creating what I call “The Maker Moment“– the current golden opportunity for people who possess those skills to both create and capture a disproportionate share of value in the global economy. (I’ve also invested directly in this theme by backing GroupTalent, a “CAA for digital creatives”).
The darker side of this trend is the ongoing structural displacement of workers who lack skills relevant to the digital economy.
The book is a short, lucid read that perfectly captures the disconnect between traditional labor market analysis — which assumes that we’re in a cyclical labor downturn stemming from the 2008 crash — and what’s actually happening: a permanent reshaping of the labor markets by technology, in which an increasing number of job roles are simply being replaced by technological substitutes, never to return.
At a macro level, this trend has serious and scary implications for the future of the global economy and civil society: a world in which population grows and job opportunities shrink is one likely to experience significant hardship and social unrest.
While I don’t see any obvious macro solutions to this trend, I do think there are some near-term opportunities to create job roles that — while far from replacing previous full-time / high-wage employment — do create the potential for meaningful work and earnings for participants in the digital economy who lack hardcore technical skills.
Lately I’ve found myself recommending to my portfolio companies that they add labor back in to their service offering.
I know this advice may seem odd to many of my peers in the startup community, but it’s not — or at least not primarily — motivated by my concern for the health of the labor markets. Rather, it’s based on repeated observations of scale constraints in traditional Software-as-a-Service (SaaS) business models, combined with steady innovation in, and acceptance of, Talent-as-a-Service labor platforms.
As an early-stage investor I work with dozens of developer-led organizations that begin their entrepreneurial journey with an almost pathological fear of human inputs to their business models; the tacit product vision for most of these companies, at least at inception, is the “AdWords dream” of customer self-service with no human-to-human interaction required.
This approach can carry a company a surprisingly long way, especially if they primarily sell to other developers. Atlassian — a provider of cloud-based process management tools for software engineers — is constantly promoting the fact that they’ve grown their business to over $100MM in revenues without hiring human salespeople.
But for most SaaS businesses — and any that cater to the needs of non-technical customers — Google’s success with self-service at scale is a dangerous illusion.
The economic magic of the software industry — which is what attracts capitalists like me to the business in the first place — is its unique ability to support massively accelerating margins at scale by replacing labor inputs with machine intelligence. Each incremental customer for a SaaS business should — at least theoretically — only add pennies to your monthlyAWS bill but be worth tens, or hundreds, or even thousands of dollars of incremental revenue a month.
The problem with this theory is that most customers actually prefer to interact with humans to solve business problems, at least some of the time. A business that requires all customer interactions to be handled by machines can defend its margins, but only with an attendant loss in customer intimacy. This can work fine for routine / low risk interactions like online shopping (think Amazon.com)…
But when the service provided is new, complex, and/or strategically important to the customer, insisting that all interactions be handled through the browser virtually guarantees that the business won’t scale.
Ten years ago this problem was most often solved by hiring full-time sales and support people to handle complex customer interactions. And this is still a valid approach for business models with long sales cycles and very high customer lifetime values.
But this approach doesn’t really work for for the new generation of highly capital-efficient tech startups. Venture investors are quick to dismiss business models that rely too heavily on labor inputs as “services, not products”, and lean / agile approaches to company development demand small teams that can adjust quickly to changes in market dynamics.
So how can an ambitious, fast-growing SaaS business hope to win by adding labor back in to their business model?
There’s no single right answer to this question, but many encouraging alternative models are currently being explored by companies in sectors as diverse as online advertising (e.g, Trada), local services (e.g, TaskRabbit), and lifestyle retail (e.g, Stella + Dot).
To me, Trada is the most interesting of these because it attaches a labor-based service offering to the most self-service of all digital offerings: keyword advertising.
Trada’s insight is that running an effective keyword advertising campaign requires significant human expertise. These skills are relatively new and most companies don’t have in-house experts on staff. But rather than create a traditional consulting services business, Trada created a specialty labor market for keyword advertising entrepreneurs. Customers specify a target cost-per acquisition and set a budget, and entrepreneurs working on the Trada platform use their skill to acquire customers for less than the target cost, keeping the difference as compensation for their expertise (after Trada takes a cut for being the platform provider, of course).
The key insight — which can be applied by almost any SaaS business targeting high-value work processes — is to attach a contingent labor marketplace to your software platform that turns a DIY (do-it-yourself) offering into a DIFM (do-it-for-me) one.
This approach is made possible by the accelerating, technology-powered disaggregation of the labor market — there are fewer full-time jobs available in enterprise, and more smart people willing to work in non-traditional ways that allow them to apply and/or develop marketable skills.
If you squint a little, it’s actually not hard to envision a future in which most professional work takes place in these kinds of technology-assisted labor markets — with tasks put out to bid and compensation based, at least in part, on quantifiable job performance.
A quote from Race Against the Machine sums this future state up perfectly:
“The key to winning the race is not to compete against the machines but to compete with the machines.”
This isn’t just good labor policy, it’s good business.