The Race to Own
the Subscriber Portfolio

For years, subscription software companies competed to help publishers convert more readers, grow subscription revenue and reduce churn. Now several are assembling something much bigger. The question is whether they recognize what they're assembling—and what it would mean to lead with it.

There is a tempting version of this story in which the entire market has overlooked a valuable opening and a single company, cleverly positioned, is ready to claim it.

That isn't what the evidence says anymore.

Zuora, Recurly, Mather Economics, Piano and Subsets are not sitting still in separate product categories. Acquisitions, integrations and AI-native products are bringing them closer together. Each has a different starting point and important missing pieces, but the direction is visible. A company that still treats them as isolated billing, paywall or retention tools would misunderstand the competitive field.

What has not become comparably clear in the public material is who is making the changing economics of the subscriber base the primary management problem it promises to help executives solve. That is a claim about the conversation vendors lead, not an assertion that their products cannot already supply relevant answers.

Which is why I think the race is interesting.

Don't just count subscribers. Manage them like a portfolio.

The words are easy. The management discipline behind them is not. Neither is building a software company that executives would choose for that reason.

Publishers didn't need a paywall.

At the dawn of the digital subscription business, publishers had a much more consequential challenge: they needed a sustainable direct relationship with readers that could fund their journalism as print economics and advertising changed.

A paywall was one response. And an important one. The New York Times' metered model became a reference point, and paid access moved rapidly from exception to convention. The American Press Institute found that just six of the 98 large US newspapers in its sample used a digital subscription model in 2010, compared with 77 in 2015. ^[1]

Once the answer took hold, the software-buying question narrowed: How do we build a better paywall? Vendors competed over entitlements, meters, offers, conversion testing and checkout. Later they moved to engagement, propensity and churn reduction. Those were real advances, not mistakes.

But the original business challenge was never merely to make a payment possible. It was to build an economically durable direct-to-consumer revenue business.

That distinction changes what should be managed. A paywall is a transaction mechanism. A subscriber relationship is an asset with a history: how it was acquired, what it costs to sustain, what the customer pays over time, how the relationship responds to price or product changes, and whether the economics improve.

It is possible to win every local optimization contest—conversion rate, renewal rate, reactivation rate—and still build a less attractive portfolio of customers.

The questions accumulate rather than replace one another. No period had a monopoly on business-oriented thinking; software vendors and sophisticated operators have long considered lifetime value and cohort economics.

What is changing isn't that executives suddenly discovered lifetime value. It's that acquisition, pricing, content investment and retention increasingly demand a joined view of the same subscriber base. The separate questions need to become one set of management decisions.

The urgency isn't confined to software vendors. In Disney's February 2023 earnings call, Bob Iger recalled a global arms race for subscribers, questioned aggressive promotions, and said the business needed loyal, higher-quality subscribers it could price effectively. That doesn't establish what Disney could or couldn't measure internally. It does show an executive shifting attention from the subscriber total to the durability of the relationships underneath it. ^[13]

Even the suppliers of subscription infrastructure describe the shift. Recurly CEO Joe Rohrlich argued in January 2025 that retention had become the growth driver and that pursuing subscriber lifetime value was essential. Recurly's acquisitions later that year showed how seriously the company was expanding beyond billing.^[2][3]

A well-funded publisher may already have an internal data team capable of this analysis. The strategic opportunity is not proving that no one can calculate it. It's making the discipline reliable, repeatable and commercially valuable beyond the handful of organizations willing to build and maintain it themselves.

Good numbers can conceal a weakening business.

Imagine two publishers reporting identical subscriber growth, similar churn and improving average revenue per subscriber. Their board packs look reassuring.

At the first, the new subscribers are steadily becoming more durable. They renew at full price, develop habits, and require relatively little paid intervention to retain.

At the second, headline growth increasingly comes from expensive acquisition channels and promotional offers. Discount-dependent subscribers leave at the first renewal, and retaining them requires progressively more incentives. The blended numbers may still look respectable because older cohorts are carrying the newer ones.

The businesses do not deserve the same strategic answer—or necessarily the same valuation. Yet averages alone won't reliably tell management which one it owns.

Consider four decisions that expose the difference.

A churn model can help with one part of these decisions. A paywall engine can improve another. A billing platform can record the payment history. But the executive question is about the economic composition of the subscriber base across those actions.

That is what I mean by managing subscribers like a portfolio. Not that subscribers are financial securities, and not that every reader must be assigned a perfect present value. A portfolio manager needs to understand the different positions, the reasons their performance changes, and the tradeoffs behind allocating the next dollar. Subscription leaders need a comparable operating view.

Blue Apron published both stories in the same filing.

Blue Apron is not a publisher, and its 2017 public offering should not be reduced to a story about subscriber composition. Its reporting is still an unusually clear illustration of the difference between a growing aggregate and customer-level averages underneath it.

Its June 2017 final prospectus showed net quarterly revenue rising from $172.1 million to $244.8 million over five quarters, from March 2016 through March 2017. Over those same quarter endpoints, orders per customer declined from 4.5 to 4.1 and average revenue per customer from $265 to $236. The series was not monotonic, but the endpoints diverged.^[11]

The company disclosed the numbers itself. There was no missing secret metric. Yet the increasing revenue and deteriorating per-customer averages invited very different questions about the growth model.

Blue Apron’s June 19 preliminary prospectus estimated an IPO price range of $15–$17 per share; the June 28 final prospectus set the price at $10. Those filings establish the price change. They do not establish how much subscriber economics, competition or wider market conditions contributed to it.^{[11] [12]}

The larger lesson is not that investors uncovered a hidden truth management couldn't see. It's that aggregate growth and economic composition can diverge, and the distinction becomes especially consequential when a company must defend its economics to outside capital.

Better to understand the quality of growth while there are choices available—acquisition channels to redirect, offers to retire, prices to test, customer experiences to improve—than to discover the weakness only when someone else is setting the price.

Five credible paths toward the same bigger question.

When I first looked at this market, it was tempting to divide it into behavioral analytics companies that lacked subscription economics and billing companies that lacked behavioral context. That map is now too simple.

The lines are blurring quickly—and the movement is observable, not speculative.

There are other routes. Twilio Segment offers identity-resolved customer profiles and predictions, including purchase propensity and 120-day predicted spending; its native churn prediction is framed around the absence of a selected event, which differs from subscription contracts, renewal prices and intervention costs. Amplitude, BlueConic and other analytics/CDP environments may have relevant behavioral foundations. Chargebee and specialized platforms have subscription-native starting points. None should be dismissed by a two-word category label. ^[10]

The map is deliberately approximate. It depicts published starting points, not independently audited product depth, actual contract economics or the winner of a race. A customer could already integrate some of these systems and create the management view proposed here. Some vendors may be doing more in private customer work than their public positioning reveals.

That doesn't make the race imaginary. It makes the claim narrower and the competitive stakes higher. Several companies can now reach much of the same technical territory. The differentiator may increasingly be the question they persuade executives to organize their businesses around—and whether their products can support the answer without heroic integration work.

Not another dashboard. A Subscriber Portfolio Statement.

The missing conversation becomes tangible only when it changes management practice. Imagine a recurring statement, reviewed by the CEO, CFO and growth leadership, that answers four questions: what is the subscriber base economically, how has its composition changed, why, and where should the next investment go?

It would not replace financial reporting, billing analytics or retention models. It would join them around recurring decisions.

Design sketch. The fields describe the proposed decision instrument; they do not imply that any named vendor ships this exact statement, or that every cost can be attributed precisely.

To produce it responsibly, a provider would need to connect identity and behavior before and after conversion; acquisition source, offer and full-price renewal history; actual payments and discounts; retention interventions; and the cost of acquiring and serving different groups. Its findings must beat the decisions currently being made with existing churn scores, finance reports and analyses.

Some fields will be ambiguous. Some retention actions will not have control groups. Cohorts can differ in age, seasonality and acquisition intent. A good system has to show those limitations rather than manufacture precise forecasts.

This is also where the opportunity differs from a prettier data warehouse query. The statement could become the occasion for recurring management action: challenge an acquisition budget, test an offer, redirect product investment, review a save program—and then measure the result.

The question a buyer would ask is no longer How well does your software predict churn? It becomes How will we know whether we are building a more valuable subscriber business?

The company that owns the decision might become a different kind of supplier.

There are two strategic moves here. They are related, but not interchangeable.

First: reframe the executive challenge.

Move the conversation from producing more subscriptions and reducing churn toward managing the economics of the subscriber portfolio. This changes what gets evaluated: not only point-solution performance, but whether the provider helps leadership understand growth quality and choose where to invest.

Second: build an environment that improves with use.

AI agents can increasingly assemble calculations, reports and campaign workflows from other tools. A vendor whose whole proposition is a replaceable sequence of steps may face growing pressure on differentiation and price.

What is harder to reconstruct is the reliable historical record that explains which acquisition choices, offers, behaviors, renewals and interventions led to durable outcomes within a particular business. If that record becomes the foundation for continuing decisions—and if it is governed, trusted and operationally useful—the software can become harder to replace than the workflow alone.

This isn't a claim that a platform owns its customers' data or that switching costs automatically confer value. Identity matching, data rights, historical coverage and customer portability all matter. Nor does an accumulating dataset guarantee improved judgment. The environment earns its place only if repeated decisions are better because it exists.

But consider the commercial difference if both moves succeed.

That could expand the relationship beyond a departmental technology budget. It could make customer expansion more natural as leadership relies on the statement across quarters. And it might justify different contract economics if the provider demonstrates durable value.

But a bigger-sounding buyer problem doesn't automatically mean a bigger deal. The decision maker must really own the problem, the analysis must change choices, and the product must deliver without an uneconomic services burden. The commercial model has to be proved—not inferred from executive language.

Having the pieces is not the same as leading the conversation.

Zuora might use the combination of billing relationships and Zephr to create a portfolio practice with financial credibility. Recurly might turn its lifecycle acquisitions into an executive decision system. Mather might translate its pricing economics and Activation Platform into a more scalable recurring product. Subsets might broaden AI experimentation from retention actions into cross-functional investment choices. Piano might recombine its audience and subscription capabilities into a more coherent economic view.

Each path begins with something real. None is a prediction of the winner.

Nor is it certain that the market will crown one supplier. A large publisher might assemble its own system. Mather's services-plus-software model might remain an advantage. The best solution might be a recombination through partnership or acquisition. Different segments may reward different approaches.

That is why the next evidence I would seek isn't another capability comparison. It is a buying account. Has a CEO or CFO engaged one of these companies to make recurring decisions about the economic composition of the subscriber base? What did they replace? What decision changed? What did the customer pay for—and did the new relationship expand or endure?

The first company to lead that conversation credibly may gain a commercial advantage even if a rival can match its software. The first to connect that argument to a trusted, compounding history of decisions may gain a more durable one.

For founders and investors, that is the real race. It's not simply who has the better retention model or the most complete platform. It's who changes what customers believe they need to manage, then becomes indispensable to managing it.

The clock is running. The position is not yet settled.

Subscription software providers have spent decades improving the answers to increasingly sophisticated operating questions. It helped publishers attract audiences, convert readers, manage subscriptions and keep customers. Now its technological boundaries are collapsing.

That makes this moment different from the discovery of an empty space. The ingredients are in motion, and several credible companies have a path to the prize. It is precisely because more companies are close that the strategic choice matters.

Who will make subscriber portfolio economics the reason a CEO wants to work with them? Who will show that the discipline changes what gets funded, saved, priced or stopped? And who will build enough product and historical advantage that the relationship becomes more valuable over time rather than easier to copy?

I cannot answer those questions from public positioning alone. But I think the basis of competition is available to be contested now.