Case Study 37.1 — Hollow Knight: A Kickstarter Promise Triples in Scope

Game: Hollow Knight Studio: Team Cherry Founders / Designers: Ari Gibson, William Pellen (and programmer David Kazi joining later) Platforms: Windows/Mac/Linux (2017), Nintendo Switch/PlayStation 4/Xbox One (2018) Kickstarter: November 2014 — $57,138 AUD raised (roughly $45,000 USD at the time) Why it matters: Hollow Knight is one of the best-documented cases of a scope expansion that worked. Team Cherry started with a modest Kickstarter scope — a small, focused Metroidvania — and across roughly three years of development delivered a game approximately three times larger than promised, without the scope collapse that kills most expansion projects. The postmortem gives designers a rare template for deliberate scope growth, and, equally, a rare case where the growth could have destroyed the studio but didn't.

The Kickstarter Promise

In November 2014, Team Cherry — at that point two designers in Adelaide, Australia, Ari Gibson and William Pellen — launched a Kickstarter for Hollow Knight. The campaign was modest by Kickstarter-era game standards. They asked for $35,000 AUD. They raised $57,138 AUD from about 2,158 backers. That is small. It would not fund the game they eventually shipped. It would not even fund the game they originally promised, at the rates a small studio needs to pay itself over any reasonable timeline. It was, by any honest accounting, seed funding for a project whose total budget would need to come from elsewhere — savings, side work, delayed compensation, and eventually the game's own revenue.

The original scope, as written in the Kickstarter campaign and the stretch-goal structure, was:

• Approximately 12 bosses
• Roughly 4 major areas plus several sub-areas
• A hand-drawn 2D aesthetic at a consistent quality level
• An estimated 10-15 hours of gameplay for the core campaign
• An expected release mid-2015 (roughly 6-8 months after the Kickstarter)

This was already an ambitious scope for two people. Even at the modest funding level, the promise was aggressive. Kickstarter in 2014 was full of ambitious promises that failed. Team Cherry's campaign did not look, on paper, significantly different from many that would later become cautionary tales.

What Team Cherry had, that most campaigns didn't, was discipline. And they would need all of it.

The Decision to Grow

Development began in earnest in late 2014. Within months, it became clear that the initial schedule was fantasy. The estimated mid-2015 release slipped, first quietly, then publicly. By the end of 2015, Team Cherry was saying "we need more time." By the end of 2016, they had added a third team member (programmer David Kazi), pushed the release to early 2017, and started communicating that the game itself had grown in scope.

This is the inflection point that kills most projects. At this moment, a team has three honest choices: cut the scope back to the original promise and ship on a later-than-planned but still-reasonable date; double down on the promised scope and try to finish what was promised; or expand the scope and push further out. The first choice is usually correct; the second is the safest; the third is the most dangerous.

Team Cherry chose the third. They decided, during development, that Hollow Knight would be significantly bigger than they had promised. The reasons, as Gibson and Pellen have described in interviews and postmortems, were specific and worth listing:

The engine and pipeline had matured. By mid-2016, the team had stable workflows for hand-drawing new areas, building combat, and composing music. Adding content was still expensive, but the per-area cost had dropped significantly from the first areas they built. What had taken eight weeks in 2015 was taking four weeks in 2016. They could get more area for the same time.

The world kept revealing more. As they built the initial areas, each area hinted at what lay beyond. Dirtmouth and the Forgotten Crossroads invited Greenpath. Greenpath invited the Fungal Wastes and the City of Tears. Every area was a question that the next area could answer. Cutting felt like leaving the world incomplete. Gibson has described this as a creative pressure rather than a production pressure — they could see the rest of the world, and not building it felt dishonest to the game.

The audience was patient. Because they had communicated transparently about the delay, and because the alpha footage they released looked striking, the Kickstarter backers were not revolting. The early Kickstarter promises had not been broken so much as deferred with explanation. The team interpreted this, correctly, as permission to take more time — and, more cautiously, as permission to do more with it.

They were willing to bear the personal cost. This is the one that most scope-expansion decisions gloss over. Gibson and Pellen had not raised enough money to pay themselves for three years. Extending development meant continuing to work on reduced income, using personal savings, taking on some side work. The decision to expand was also a decision to extend financial precarity. They chose to accept that cost because they were artistically committed to the larger game.

All four of these conditions were specific to Hollow Knight. The team could check each one. If any had been false — if the engine hadn't matured, if the world hadn't compelled further building, if the audience had been hostile, if the team had not been able to eat during the extended development — the expansion decision would have been the wrong one. Scope expansion only works when the specific conditions support it.

The Final Scope

Hollow Knight shipped on PC on February 24, 2017, roughly two years after the original Kickstarter target. The final scope, measured against the original promise:

• Approximately 30+ bosses (more than double the Kickstarter promise, depending on how optional and alternate bosses are counted; closer to 40 if DLC is included).
• 10+ major areas, each with distinct aesthetics, enemies, and mechanics: Forgotten Crossroads, Greenpath, Fungal Wastes, Crystal Peak, City of Tears, Deepnest, Royal Waterways, Resting Grounds, Kingdom's Edge, and Ancient Basin, plus the late-game areas.
• Estimated 25-40 hours to complete the core campaign, with completionist playthroughs pushing 60+ hours.
• Four free DLC expansions released over the following two years: Hidden Dreams, The Grimm Troupe, Lifeblood, and Godmaster. These added roughly 15+ more bosses, new endings, and new side content.

By almost any measure, the final shipped game was three to four times the original Kickstarter promise. For a small indie team working largely on savings, this was an enormous over-delivery.

It is also, notably, the kind of over-delivery that destroys studios when it happens badly. The developer's promise to "deliver more than we said" sounds generous but often masks scope chaos. Many projects that over-deliver end up shipping broken, or shipping late enough that the studio collapses first. Hollow Knight did not.

Why the Expansion Worked

Four disciplines, visible in retrospect, prevented the expansion from becoming collapse.

Discipline 1: The expansion was always driven by the game, not the team's ambition. Gibson and Pellen were not adding features because they wanted the game to be more impressive. They were adding features because the existing features implied them. Every new area was connected to an existing area. Every new boss filled a narrative or mechanical role that had become visible during play. The expansion had the character of completion rather than addition. This distinction is subtle but consequential: a team that expands to complete an implied whole can stop when the whole is implied to stop, while a team that expands to add ambition has no stopping condition.

Discipline 2: The team kept the core mechanics stable. The combat system, the movement system, the charm system — the core mechanical language of the game — was largely set by mid-2015. What changed during the expansion was content: new areas, new enemies, new bosses, new charms that fit within the existing system. They did not add new fundamental systems during the scope expansion. This kept the integration cost manageable, because each new area was exercising existing machinery rather than requiring new machinery to be built. Many expansion failures add not just content but new systems, and the system proliferation is what actually kills them.

Discipline 3: The polish bar stayed consistent. Every new area was polished to the same standard as the earliest areas. New enemies had fully-realized animation. New bosses had the same level of fight-choreography. New music was composed by Christopher Larkin with the same care. The team did not accept quality regression in order to fit more content. This is perhaps the most striking discipline. Many expansions produce a game where the early content is polished and the later content is rough, because the team ran out of polish-time. Hollow Knight does not feel that way. The last areas are as polished as the first.

Discipline 4: They were willing to delay. Every time the scope grew, they let the schedule grow with it. They did not try to ship more content in the same time. They communicated new delays to backers; they took the reputational hit of "it's taking longer." They prioritized shipping a finished game over shipping on time. This is the scope-triangle choice (scope, quality, time — pick two) made explicitly. They picked scope and quality. They paid in time.

Together, these disciplines form a template. Scope expansion can work when:

• Each new feature completes an implied whole rather than reaches for something new.
• The core systems stay fixed; only content grows.
• The polish bar does not regress; new content matches old.
• The team is willing to delay to protect quality.

If any one of these conditions breaks, the expansion tends to fail. Hollow Knight held all four.

The Cost

The Hollow Knight expansion was not free. It had costs, and the team has been honest about them.

Personal financial cost. Gibson and Pellen have described the final year of development as lean. The Kickstarter money was long gone. Revenue from the game would not arrive until after launch. They worked at reduced compensation, dipped into savings, and took on financial risk that a more risk-averse team could not have borne.

Delayed DLC commitments. Because they had promised DLC during the Kickstarter ("Hidden Dreams" and others were stretch goals), extending the base-game scope meant delaying the DLC. The DLCs did ship — all four, all free — but they shipped over the following two years rather than shortly after launch. Backers waited.

The cost of the next project. The Silksong sequel, announced in 2019, did not ship in 2020 as many fans expected, nor in 2021, nor 2022, and has been through an extended development cycle. There are many possible causes, but one that applies to any team coming off a large-expanded project is fatigue. A team that has just spent three years expanding one project has accumulated an exhaustion that carries into the next one. The cost of Hollow Knight's scope expansion is partly visible in Silksong's timeline.

None of these costs undo the success of Hollow Knight. The game shipped to near-universal acclaim. It has sold over 2.8 million copies (per later sales reports). It has become one of the foundational games of the 2010s Metroidvania revival. The four disciplines paid off. But they paid off at real cost, and any designer looking at Hollow Knight as a model should look at the costs as well as the success. The expansion was a bet. The bet paid. The bet could have failed.

Lessons for Your Scope

Hollow Knight teaches three lessons that apply to your own project, even if your project is much smaller.

First: Scope can grow during development, but only under specific conditions. The expansion must be driven by the game's own logic, not by the team's ambition. Core systems must be stable. The polish bar must hold. The schedule must be flexible. If any of these conditions fail, do not expand — cut. Most first-time designers will be tempted to expand when they should be cutting. Team Cherry expanded because all four conditions were true. Yours probably are not. Default to cutting.

Second: Communicate honestly. Team Cherry's willingness to tell their Kickstarter backers, repeatedly and without euphemism, "it will take longer than we said" is part of what preserved the audience's patience. Designers who hide delays lose their audience's trust. Designers who own delays keep the audience invested. The honesty itself is part of the production discipline.

Third: Scope decisions are personal decisions. Team Cherry's expansion was, in effect, a decision to accept extended financial precarity in exchange for the game they wanted to ship. That is a choice individual people make about their lives. It is not a production-process insight you can import without the personal cost. If your life cannot afford the delay that scope expansion requires, you cannot afford the expansion. This is not a failure of design ambition; it is an honest recognition that scope is ultimately a life decision, not just a project decision.

Further Reference

• Team Cherry's Kickstarter page (archived): the original scope as promised.
• Hollow Knight developer diaries and YouTube interviews with Gibson and Pellen, 2015-2018.
• Silksong announcement and subsequent updates as an illustration of the long-tail cost.
• Noclip's Hollow Knight documentary (if produced; see also general indie-development documentaries covering the project).
• Christopher Larkin's interviews on the scoring process for additional perspective on the polish discipline across the expanded content.