Telos by Shrikant Kumar
Latest
  1. The Library in the Air
  2. A Factory for Other People’s Ideas
  3. The Blue Dot
  4. Paying a Stranger
  5. A Computer of One’s Own

Decide Case 9 2013

The Software in the Box

You run a software company whose design and imaging applications are the professional standard. You sell them as boxed suites that customers buy once, for up to $2,599, and upgrade every year or two. A year ago you also began offering them by monthly subscription. Now you must decide what happens to the box.

Do you stop releasing new versions of the software people can buy and own, and sell it only by subscription?

The situation

It is the spring of 2013. Your company makes the applications that most professional designers, photographers, illustrators and video editors use every day. For a decade you have sold them in bundles, called suites, released roughly every eighteen months to two years. The largest suite lists at $2,599; smaller ones at $1,299 to $1,899, with upgrades for existing owners at a few hundred dollars. A customer who buys a suite owns a licence to that version forever.

That model has made you very profitable, and it has a shape you know well. Revenue surges when a new suite ships and sags in the months before the next one, as customers wait. Your engineers work towards a single large release, so a feature finished just after one launch reaches customers more than a year later. Many customers skip versions, running software three or four years old. And in many countries the boxed software is copied illegally on a large scale; your own filings call piracy a persistent problem.

In November 2011 you told investors you would reorganise around digital media and online marketing, cut about 750 jobs, and move your creative software towards a subscription service. Your share price fell by the most in three months. In April 2012 you launched the subscription alongside the newest boxed suite: $49 a month on an annual plan, or $79 month to month, for every application you make, plus online storage and services. The boxed suite remains on sale. For a year customers have been able to choose.

What you know

Your finance team has watched the boxed business closely. The number of suites you sell each year has been roughly flat, at about three million units, for a long time, even as the number of people using your software has grown. Growth has come from price rises and upgrades, not from new customers.

The subscription has done better than you planned. Your internal projections assumed about 5,000 new members a week. By the end of the summer of 2012 you had more than 200,000 paid members, and by the end of the year about 300,000.

Running both models has costs. Every feature has to be built and tested twice, once for the boxed product and once for the subscription. Your sales teams and resellers send mixed messages. And as long as the box is on sale, many of your most loyal customers will keep buying it, because they prefer to own.

You have spent the past eighteen months preparing investors. You have explained that moving from one-off sales to monthly payments means revenue will fall before it rises, because a customer who paid $2,599 at once now pays $600 a year. You have started reporting the number of paid subscriptions and the annualised value of recurring revenue, so that the growth of the new business is visible while the old one shrinks.

What you don’t know

You do not know how many of your three million yearly buyers will follow you if the box disappears, or how many will keep using old versions for years. You do not know how many people who never paid the boxed price, students, hobbyists, small studios and those using illegal copies, will pay a monthly fee. You do not know whether a competitor will use the moment to offer professional-grade software that customers can own. And you do not know how long investors will tolerate falling revenue before they lose patience.

The voices in the room

Some argue the hybrid is working: the subscription is growing on its own, so why provoke the professionals who built the company’s reputation? Others argue that the hybrid is a trap. As long as the box exists, the company will build two products, split its customers, and stay tied to the eighteen-month release cycle; the transition will take a decade and the valley in revenue will be long and shallow instead of short and deep. Still others worry less about investors than about customers. Designers keep years of client work in your file formats. If they stop paying, can they still open their own files? Asking them to rent what they used to own, they say, will be taken as a breach of trust.

What happened

The company was Adobe, and the product was its Creative Suite, which included Photoshop, Illustrator and InDesign. On 6 May 2013, at its MAX conference in Los Angeles, Adobe announced that Creative Suite 6 would be the last version sold as a perpetual licence. It would remain on sale with bug fixes, but every new feature would go only to Creative Cloud. The first subscription-only versions shipped on 17 June 2013.

The reaction was immediate. A photography student started a petition on Change.org asking Adobe to keep perpetual licences; it gathered more than 4,400 signatures in four days and, by Bloomberg’s later count, about 50,000 in all. Critics objected to paying for ever for software they could no longer keep, and to losing access to their own work if they stopped. Copies of the new subscription software appeared on file-sharing sites within days, so the move did not end piracy, though it changed its economics.

The valley came as planned. Adobe’s revenue fell from $4.40 billion in fiscal 2012 to $4.06 billion in fiscal 2013, a drop of 8 percent, and was roughly flat the year after. Underneath, the new business grew. Paid Creative Cloud subscriptions rose from 0.3 million to 1.4 million during fiscal 2013, and the annualised value of creative subscriptions from $155 million to $768 million. By the end of fiscal 2015 Adobe had 6.2 million paid subscriptions, roughly double the number of suites it used to sell in a year. In fiscal 2025 its revenue was $23.8 billion, 96 percent of it from subscriptions, and its net income $7.1 billion.

The cost to trust lingered. In June 2024 the US Federal Trade Commission, through the Justice Department, sued Adobe and two executives, alleging that the company hid a large early termination fee on its most popular plan, an annual commitment paid monthly, and made cancelling deliberately difficult. Adobe disputed the claims. In March 2026 it agreed to a proposed settlement of $75 million in civil penalties and $75 million in free services, without admitting wrongdoing; the same month the UK’s competition regulator opened its own investigation into the fees.

Judging the decision apart from the outcome

The process was unusually strong. Adobe did not switch overnight. It ran both models for a year, which gave it evidence that demand for the subscription was real and running ahead of its own projections. It told investors years in advance that revenue would fall, and gave them new measures, subscriptions and recurring revenue, to judge the transition by. It understood the irreversibility of the step and chose to make it once, cleanly, rather than prolong a hybrid in which the old model kept pulling customers and engineers back.

The decision also rested on a clear diagnosis. Flat unit sales meant the boxed model had stopped finding new customers; the subscription was cheaper to start, which opened the software to people who had never paid $2,599, and it ended the release cycle that had slowed product development. Those were not hindsight arguments. They were visible in the company’s own numbers in 2012.

Where the process was weaker was in its treatment of the customers who did not want to rent. Adobe’s own feedback, it said, favoured subscriptions; the petition showed that a significant group felt otherwise, and for most of them there was little real alternative. That combination, high loss aversion among loyal users and weak exit options, made the switch commercially safe and morally fraught. The later regulatory action concerned practices adopted after 2013, not the decision itself, but it shows how a subscription business, once established, can be tempted to keep customers by making leaving hard rather than by making staying worth it, a question of choice architecture as much as of pricing.

What a good process would have weighed

Each alternative had a real case. The hybrid avoided provoking loyal customers, but the evidence suggested it would keep the company building two products for years. A gradual approach protected earnings, but Adobe had already prepared investors for the fall and would have had to prepare them again later. Cheaper boxed software addressed the entry price but not the release cycle or the flat unit sales. The all-in choice accepted a short, deep valley and an angry minority in exchange for a single product, continuous development and recurring revenue. A good process would have added one more element: a credible commitment to customers about what they would keep if they stopped paying, such as continued access to their files and a simple way to leave, which would have reduced the sense of loss that the move created.

Sources and limitations

The financial figures come from Adobe’s annual reports. The figures on flat unit sales and the preparation of investors are the chief financial officer’s own account, given after the transition had succeeded, and should be read as a participant’s recollection. The figure of 200,000 subscribers is from Adobe’s results as reported by a technology writer; the petition counts are from contemporary press and a later Bloomberg article. The FTC’s claims are allegations that Adobe disputed, and the settlement was proposed, without admission of wrongdoing, and awaiting the court’s approval at the time of writing. The suggested commitment to customers is the author’s judgement.

↑ ↓ to move↵ to open⌘K to toggle