Discover more from Frames
This month's Frame: Exit, voice, loyalty
A framework that helps to understand the relationship between a product and its users as a dynamic equilibrium.
In the last quarter of 2021, for the first time in 17 years, Facebook lost users. Not many, about half a million—a blip compared to its impressive 1.93 billion daily active users.
Yet that dent in its growth curve has been taken extremely seriously, both by Meta's leadership and by investors, who heavily sold the stock in the days following the announcement (monthly active users and the stock have since then bounced back).
How can we explain that such a tiny drop in users caused such a large drop in market value?
The key concept is time. The valuation of stocks, and much more so tech stocks, is a function of their future earnings. The thesis is that digital platforms like Meta can leverage network effects as a moat to fend off competitors, which guarantees the longevity of their business. The thesis assumes that users will decide to stay on the platform rather than to switch to competitors. Which is why most digital platforms betting on network effects make the costs of switching to other platforms hard.
So it is precisely because Facebook is designed directly (eg. the UI is nudging users to deactivate, rather than delete, the account) and indirectly (eg. social costs associated with not being on the dominating social networks), to prevent users from leaving the platform, that it is particularly worrying when they nonetheless do. Feedback loops work in both directions: if they turbocharge the growth of a digital platform, they can also precipitate its collapse.
The financial models used to value stocks are predicated on assumptions, one of the most significant ones being, precisely, the rate of user growth. However, they are incapable of providing guidance as to how to evaluate this assumption. How do we know if a tiny drop in users is an anomaly, or the beginning of a sustained new trend?
Albert O. Hirschman, an economist, developed a framework that can help understand the relationship between users and a product over time as a dynamic equilibrium. This can prove useful not only to help financial analysts predict users growth rates more accurately but also to help product managers design products that users will want to keep using.
In 1970, Hirschman published "Exit, Voice, Loyalty", a book in which he puts forward a simple framework. People who engage with an organisation of any type can fall within three categories:
they are satisfied with the organisation (loyalty)
they are dissatisfied with the organisation and try to modify it from within (voice)
they are dissatisfied with the organisation and thy leave (exit)
This is equally true for consumers. If you think your broadband has become slow and unstable, you can voice your concerns and hope for your provider to fix it, or you can leave. If you were a fan of the original Coca-Cola recipe and don't like their new flavour, you can switch to Pepsi, or you can express your discontent and hope for them to revert back to the old one (something which famously happened in 1985).
Hirschman's analysis is primarily concerned about the mechanisms that organisations put in place in order to enable constituents to change organisations from within, rather than exiting. He sees this as the most efficient way to foster engagement, and ultimately loyalty.
Looking primarily at political organisations and their members, Hirschman argues that rigid organisations that don’t have any “voice” or “exit” mechanism risk falling apart before their leaders even notice it. This is because rigid organisations tend to be characterised by:
the absence of signal as to how loyal its members really are
a lack of understanding of the characteristics of the political organisation the members aspire to live in
When such organisations fall, they do so without warning.
Using the framework
What is true of political organisation can also be true of the relations between customers and products. It's no secret that many companies are intentionally making it hard for their clients to abandon their products (and this largely predates the digital era). Famously, to end a New York Times subscription, you need to call them during business hours and be prepared for a long wait.
A recent research project we did on privacy revealed that a majority of digital platforms are creating artificial barriers to prevent people from deleting their account. A classic one is creating a two-step process: users have to first "suspend" their accounts, and only after a lengthy wait (often a month), can they finally ask for it to really be deleted.
So what can product strategists learn from Hirschman's framework?
First, rather than making it hard for users to exit, follow the "easy in, easy out" design principle. If anything:
it guarantees that users' churn will be a clean signal, unequivocally communicating that your product is falling behind your competition
it's a forcing function ensuring that the organisation is geared towards building the best possible product
Second, create open communications channels with your users, either in the product itself, or outside of it—and when they voice their concern, pay attention. Seen through this light, the recent rollback of Instagram's new interface isn't flip flopping: to the contrary, it's a sign that the organisation is listening to its constituents.
In Ernest Hemingway's novel The Sun Also Rises, a character is asked how he went bankrupt. “Two ways,” he answers. “Gradually, then suddenly.”
As Hirschman theorised, such is often the fate of organisations that (1) fail to develop mechanisms for their constituents to voice their discontent and (2) artificially retain unsatisfied members otherwise ready to leave.
As members' frustrations remain unvoiced, they grow, unseen, until they reach a tipping point and then users exit anyway. From there a negative feedback loop kicks in (eg. in the case of companies: less users, less revenues, less investment, worst product, less users) and things can quickly collapse. Gradually, then suddenly.
We're delighted to be back in person at this year's EPIC conference in Amsterdam.
Come and say hello if you're at the conference, or email us if you'd like a pre-print of our paper on how teenagers build social resilience through virtual worlds.