Companies are the most theoretically interesting when they're still figuring things out, the most lucrative to own when they've figured out the important stuff and want to scale it, but that value mostly accrues in corporate middle age, when they're getting close to the point that there's nothing left for them to invest in and the big question is what capital structure is optimal for returning as much cash to shareholders as possible.

The exit state looks different for different companies. The usual shorthand investors will use is that they'll pencil in GDP-level growth, perhaps with an adjustment based on the company's industry—if it's an airline in a middle-income country, for example, they can probably add a few points of revenue growth on top of the GDP growth number because travel spending grows disproportionately as countries get richer, while an industry that declines as a share of GDP, like a water utility, might get a small haircut to that number.

In some cases, the business is an asset (or a collection of assets) that are expected to deplete over time. For a mine, or an individual oil well, you'd want to model some kind of decline curve that eventually reaches the point where additional resource extraction is uneconomical (with a mine) or a low, predictable trail of production (stripper wells). But these are actually very different! The oil well turns into an oil-denominated annuity. But the mine is now an option, which can go back in the money if the demand picture changes—uranium goes in and out of style—or because changes in the regulatory or technological picture suddenly make it worth owning again. Decline curves must be modeled for pharma assets that will go off patent as well. 

For subscription-based companies, maturity means they've reached whatever level of market penetration their model naturally supports. Spotify is still growing, because a lot of the world wants to listen to music and has willingness to pay for a big enough library and a good enough user experience—the business is a testament to pricing power in that when it started, torrenting music was pretty easy and legally low-risk, but required either patronizing various user-hostile sites that used maximally aggressive ads and sometimes tried to get their customers to download malware, or ingratiating oneself into the topsite-aware community and probably ratcheting that legal risk back up.

Consider a business like Spotify, or Netflix, that eventually mostly saturates its primary market. They can spend enough on marketing to mitigate customer churn, but they've realistically tapped into the entire movie-watching or music-listening audience out there. People will keep aging into the service, and other people will age out or die, but there isn't much they can do.

Or is there? A steady subscriber count for a consumer subscription product doesn't mean a fixed userbase—stability is just the point at which churn is equal to gross adds. Their analytics are no doubt much more sophisticated than this, but there are probably some customers who are more or less lifetime subscribers, some who are intermittent ones, some who will only sign up for specific IP, etc. At maturity, growth in users comes from some combination of population growth and whatever churn reduction approach makes those intermittent subscribers and IP-specific ones likely to go one month longer. But this has a big impact; take monthly churn from 3% to 2.9% and you take steady-state revenue up 3.4%, and the marginal cost of that incremental usage, for a fully-scaled product, is probably pretty low. 

This actually leads to two kinds of maturity-driven steady states. In one, the company basically runs things on autopilot, accepts that over time their steady-state userbase will erode, and slowly curtails investment in their platform. Sometimes you see this in products that have gone so long without a refresh that users would experience future shock if the product even caught up halfway with how design standards have changed since then. GoodReads, for example, passed an event horizon long ago where, even though they're a critical marketing channel for some genres, updating their design sensibility from 2011-chic would be net harmful.

For other companies, especially the ones that make multiple fixed investments but get variable returns from them (think restaurants and retail, but also conglomerates) the steady-state is a slow decline. As they mature, their growth moves to same-store/same-unit sales rather than to additional units. But general economic growth also increases the ceiling on the number of potential units. There are more intersections that can support a Starbucks today than there were in 2000, and more suburbs that justify a Costco, too. Some of these companies have a long twilight where they can still reinvest for growth, but they’re not constrained by anything other than waiting for general economic growth to open up slightly more opportunities. For these companies, there isn’t really a final end-state of maturity—if they actually stop growing, it implies that someone else found a way to grow into the opportunities they would have targeted first, and that’s a good sign that they’ve flipped to being in decline.

We’ve looked at the corporate lifecycle in many Diff pieces:

• Big tech was too cheap in the 2010s because investors were overly fixated on their current business rather than their track record of revamping it.

• A good way to model growth is to look at all the reasons it stops ($).

• You can get high returns from above-average growth persisting for a surprisingly long time ($).

• One reason to be skeptical of indefinite growth is that it creates a valuation paradox ($).

• There are some surprising beneficiaries to a general increase in GDP growth ($).

Join the discussion!