The Investment Phase Design Pattern
Get users to store value so they cannot leave without losing something
The Investment Phase Design Pattern leverages three psychological tendencies to increase user retention: we irrationally value our own efforts (the IKEA effect), we seek to be consistent with our past behaviors, and we change our preferences to avoid cognitive dissonance. By designing products that accumulate stored value through user effort, companies create powerful switching costs that keep users engaged even when alternatives emerge.
Stored value comes in five forms: Content (songs in iTunes, posts on Facebook), Data (LinkedIn profiles, Mint.com financial data), Followers (Twitter audiences), Reputation (eBay seller ratings, Airbnb reviews), and Skill (Photoshop expertise, keyboard proficiency). Each investment increases the product's utility and makes leaving increasingly costly.
Critically, the investment phase differs from the action phase in two ways: it concerns anticipation of future rewards rather than immediate gratification, and it intentionally increases friction. This works because investment is requested after the variable reward, when users are primed to reciprocate. Research shows that even computers that are 'helpful' receive more effort from humans in return, suggesting reciprocity extends beyond human relationships to product interactions.
- People irrationally value things they have put effort into (the IKEA effect), making invested-in products feel more valuable than objectively equivalent alternatives.
- We seek to be consistent with past behaviors; small investments lead to larger future commitments through escalation of commitment.
- Cognitive dissonance drives us to change our preferences to justify past investments, creating a self-reinforcing cycle of increasing attachment.
- Investment should be requested after the variable reward, leveraging the principle of reciprocity when users are in a positive state.
- Stored value (content, data, followers, reputation, skill) creates switching costs that compound with each use, building a moat against competitors.
- Identify Five Forms of Stored ValueAudit your product for opportunities to accumulate stored value in each category: Content (user-created or curated items), Data (personal information that improves the service), Followers (social connections), Reputation (trust scores and reviews), and Skill (learned expertise with the product). Map which forms your product already utilizes and which are untapped.Pro tipLinkedIn discovered that getting users to enter 'just a little information' dramatically increased return visits. Even small data investments create powerful hooks.
- Design Progressive Investment RequestsStage investments from small to large, starting with easy tasks during onboarding and building up to harder tasks during successive Hook cycles. Do not ask for too much too early. The first investment should feel almost effortless.Pro tipAny.do teaches its task management app through simple instructions, then asks users to connect their calendar (investment). This loads future triggers by enabling post-meeting notifications.WarningIf users are not making the investments you designed, you are probably asking for too much too early.
- Connect Investment to Next TriggerDesign each investment action to load the next external trigger. On Tinder, every swipe creates potential match notifications. On Snapchat, every sent photo is an implicit prompt to respond. On Pinterest, every pin grants permission to notify when others interact with the content.Pro tipThe self-destructing messages on Snapchat encourage immediate responses, creating a rapid back-and-forth relay that loads triggers with each message sent.
- Minimize Time Between Investment and TriggerReduce the delay between a user's investment and the next trigger firing. The shorter this gap, the faster users cycle through the Hook Model and the more quickly habits form. Audit your notification and engagement systems for unnecessary delays.Pro tipReal-time notifications are more effective than daily digests for habit formation because they minimize the gap between investment and the next trigger.
- Make Stored Value VisibleShow users the value they have accumulated. Display their content libraries, profile completeness meters, follower counts, reputation scores, and skill levels. Making stored value visible reinforces the IKEA effect and increases perceived switching costs.Pro tipeBay prominently displays seller reputation scores. LinkedIn shows profile strength meters. These visible indicators of accumulated investment remind users of what they would lose by leaving.
Twitter's investment mechanism works on both sides: followers curate lists of people to follow (improving content quality), while content creators invest effort in producing tweets to attract followers. Each follower gained represents stored value. App.net built a technically superior Twitter clone but failed because it could not replicate the years of follower investment users had accumulated on Twitter.
Dan Ariely's origami study showed that people who assembled their own paper cranes valued them five times more than non-assemblers did, and nearly as much as expert-made origami. IKEA's self-assembly furniture model accidentally leverages this psychological principle: customers who build their own furniture develop irrational attachment to it. The same mechanism operates digitally when users invest effort in customizing profiles, curating content, or building social graphs.
Eyal drew on Dan Ariely and Michael Norton's IKEA effect research (2011) showing that people value their own origami creations five times higher than others valued them, and nearly as high as expert-made versions. Combined with classic psychology research on consistency (the yard sign study showing 76% compliance after a small prior commitment versus 17% without) and cognitive dissonance (Aesop's fox and the sour grapes), Eyal identified investment as the critical fourth phase that closes the habit loop.
The framework was further informed by Stanford research showing that people perform twice as much work for computers that were previously helpful to them, demonstrating that reciprocity extends to human-machine interactions.