Why did Substack suddenly start worrying about culture?

A middle-man business isn’t a manufacturer like JHS or a consultancy like my own company. It’s paid not for its value-added build logistics and supply chain as you are, Josh; nor for solving someone’s tricky problems as I am.

It’s paid to connect people. In this regard, it’s like a bank; an airline; a telco and of course other socials platforms. Users aren’t paying because they want to use the infrastructure and enjoy the ‘user experience’; they pay for whom the infrastructure connects them to.

So the differential value of Substack as a platform is in its firm infrastructure that connects people — especially, its referral algorithm, which Substack built around long-form writing.

More than the publishing services (which are plentiful and cheap everywhere), the quality of that referral algorithm justifies Substack’s cut of paid subs. To maximise profitability, Substack must maximise reach (market-size) and utilisation (paid throughput), which also explains why Substack recently invested in auto-translation services, and previously in leaderboards and category ‘rising’ buzzboards.

The problem is: autogen-AI undermines the algorithm by copying whatever people want to read, and serving up dilute versions of it. It can do that faster than the algo can change, and swamp the stuff readers would otherwise pay human authors for, leading to an exodus of the actual innovators.

The Strategy

There’s more than one way to counter it, but Substack was lured to the high-profile detection-tool outsource and announced it not like a trialled tactic, but like a strategic policy.

If it works, there’s plenty of potential upside: it’s inexpensive, looks good for the brand, promises continuous improvement at no R&D cost to Substack, offers cost-recovery from serial AI-users in future, while shifting any residual risk to authors — and authors complain anyway, right?

Note also that Substack has admitted to expecting false accusations, but has shown no sign of spending money to remedy them. It’s going with a feedback form instead. That doesn’t speak to cultural investment and service to artisanal authors: it speaks to bottom-line.

The reference below details the framework for this analysis.

Stabell, CB, Fjeldstad, OD, Configuring Value for Competitive Advantage: Shops, Chains and Networks, Strategic Management Journal, Vol 19, 413-437 (1998)