Delight 不是撒彩纸：同时满足功能和情感，才叫好产品

Host Lenny Rachitsky frames the episode around two kinds of product teams: those who see delight as the way to win, and those buried in features, deals and infrastructure who see it as a distraction. Changuel, who worked on Skype, Spotify, Google Chrome and Google Meet and later served as a dedicated "delight PM" at Google, answers with a simple opening question she asks in every workshop: what is your favorite product? The answers vary widely, because people connect with products for productivity, joy or comfort, often without being able to explain why.

She defines delight as the ability to build products that serve emotional and functional needs together. Even well-functioning products can struggle to gain traction if they never let users feel something. Spotify has many competitors, she says, but features like Wrapped, mood-based recommendations and Discover Weekly make users feel seen, heard and connected.

For Changuel, delight is no longer a luxury because markets are crowded, and it works as a differentiator that drives loyalty, word of mouth, growth and revenue. Lenny adds his own examples: he has stopped opening Instagram because it doesn't feel good, while Linear is beating Jira and Asana largely through thoughtfulness about every part of the experience.

She notes that designers talk about emotional design and marketers about emotional ads, but in the business and product world the topic is rarely discussed. That gap is why she wanted a concrete, actionable framework, since every founder says they want to delight users but few know how.

Changuel explains that, in theory, delight is a combination of joy and surprise, drawn from the wheel of emotion created by professor Plutchik. To put it into practice she names three pillars. The first is removing friction, which means finding the low moments in a journey where users feel anxiety or stress and reducing them.

Her example is an Uber ride she booked on the way to a train from Paris to London. The driver cancelled, she grabbed a taxi, and later found she had been charged for a trip she never took. Instead of the long complaint she expected to write, she needed only two clicks to get a refund, which removed the stress and made her more willing to use the service again.

The second pillar, anticipating needs, means not waiting for users to tell you what they want. When her family landed in Singapore facing steep roaming fees, her husband opened Revolut, found an eSIM tab and bought one for €7. Revolut had guessed that many of its users travel, and she cites Superhuman's Rahul Vora, who says a loved product has to set the bar higher than its users do.

The third pillar is exceeding expectations. Changuel describes watching her husband about to pay €120 for a coffee machine when Edge offered a coupon through autofill that cut 15% off the price. Nobody asked for the coupon, and that gap between what was expected and what happened is the surprise. Teams do not have to apply all three pillars at once, but each should be considered in every build.

Changuel is blunt that decoration for its own sake misses the point. After a workshop attendee showed her a feature where shaking the phone makes snowflakes fall, she coined the "confetti effect" for features that add flair but no value. If confetti brings nothing, she says, don't ship it, though she is not against it when it carries meaning.

Her example of confetti done right is Airbnb. As a host she cares about keeping Superhost status, which Airbnb reassesses every three months. When the status is renewed, the app turns into confetti, which tells her that Airbnb recognizes her effort and celebrates the moment with her.

Lenny, who helped build the Superhost program at Airbnb, shares some history. It came from veteran hotelier Chip Conley rather than from a metric to move, and the team feared it would hurt search conversion. It made hosts very happy but, to the team's surprise, did not move metrics at the time. Lenny also notes that the framework is really a general way to build great experiences, with delight as part of the result.

Lenny raises the elephant in the room: enterprise giants like Workday, SAP and Salesforce did well while being far from delightful. His view is that in a green-field market with a big pain point the experience matters less, but as competition grows, delight is how a product stands out. Changuel agrees and adds that as long as humans use a product, their emotions need to be honored.

Her career had been mostly consumer, so for the book she interviewed people at B2B companies such as GitHub, Atlassian and Snowflake. They value emotional connection too, under different names: Dropbox has a product principle called cupcake, and Snowflake has one called superhero. She sums it up as "business to human", noting that experiences from consumer products raise expectations for enterprise software.

One delighter she favors is humanization, which asks how the experience would improve if the product were a person. At Google Meet, which she joined as the pandemic began, the team compared the product not to Zoom or Teams but to an actual in-person conversation. A Dyson product head told her something similar: they compare their vacuum not to other vacuums but to a hired cleaner, which led them to add room-by-room recommendations.

She also points to a Buffer product leader who emailed the roughly 2% of customers who had been inactive for months and offered refunds. It meant losing revenue, but it built trust, and some customers decided to stay. Lenny adds Slack's playful animations and quotes, and Revolut's standout experience in a crowded banking market, as proof that the more competition there is, the more delight matters.

Not every company can afford a dedicated delight PM, as Google did for Chrome and Meet, so Changuel built the Delight Model to make the practice actionable without changing a team's existing ways of working. It has four steps. The first is identifying user motivators, and she argues for a third kind of segmentation beyond demographic and behavioral: motivational, based on why people use the product.

At Spotify, she says, some people want to find a specific track, while others want inspiration, to feel less lonely, productive or connected. Lenny likens this to jobs to be done with an added layer of how you want to feel. Emotional motivators split into personal ones (how I want to feel using the product) and social ones (how I want others to feel about me), which explains why people share their Wrapped.

The second step converts those motivators into product opportunities, shifting the discussion from a problem space to an opportunity space. Changuel says she is not strict about the method: how-might-we questions or any other framework will do, so long as the strategy starts from motivators. Emotional examples vary by product, from Uber users wanting to feel secure to Miro users wanting to feel like a better facilitator.

In step three, teams place candidate solutions on what Changuel calls the delight grid, a matrix of functional and emotional motivators. This sorts features into three kinds. Surface delight serves only an emotional motivator, like Spotify's Wrapped or the birthday balloons on her Apple Watch. Low delight serves only a functional one.

The third and most powerful kind is deep delight, where one feature serves a functional need and an emotional one together. She gives Spotify's Discover Weekly and Jam as examples: users find new music in a personalized way that makes them feel understood by the app.

Step four is validation with a delight checklist, which makes sure delight has user impact and business impact rather than being an excuse for aesthetics. Among other items, she stresses feasibility, familiarity and especially inclusion, since what makes one person happy may not please another, or even the same person tomorrow.

Her cautionary tale is a French delivery app that sent a Mother's Day notification styled as a missed call from Mom. Some users found it joyful, but for others it triggered grief, and the feature drew heavy criticism in the press. The intention was positive, she says, but it was not inclusive enough.

Asked for other checklist items, Changuel returns to familiarity: people say they love surprise, but they do not want to be surprised too much. She tells how Discover Weekly was originally designed to contain only tracks the listener had never heard. Soon after launch, engineers found a bug that sometimes injected songs the user already knew.

When they fixed the bug, all the feature's success metrics fell. Users, it turned out, valued the accidental familiarity, so the team restored it, and she jokes that the Discover Weekly people enjoy today is the buggy version. She worked close to that team on transcoding the tracks. Lenny wonders how many other products miss out because they never make such mistakes by accident.

Earlier in the discussion Lenny asked about a taxonomy of emotional needs. Changuel says it depends on the product and industry, with security, or wanting to feel a better version of yourself, as common examples. The wheel of emotion can serve as a starting point, but not every emotion applies to every product.

Changuel's first case study is Chrome's tab management on iOS. People leave tabs open as reminders, to-dos or simply because they forget them, and some have hundreds. Interviews showed her that users have a real relationship with their tabs: some were embarrassed and apologized for how many they had, and for others closing tabs automatically was out of the question.

The resulting feature moves tabs untouched for 21 days into an inactive tabs folder. The tab grid gets cleaner, people feel less stress, and trust in Chrome stays intact because nothing is closed. A purely functional fix, like compressing thumbnails, was possible, but this approach solved the functional problem and the emotional one together. Lenny says a related approach in the Arc browser, which deletes tabs after some days, was why he switched.

Her second case is Google Meet, which she joined about a month before Covid hit Europe, just as usage exploded. Research found three patterns: boredom, low interaction and what became known as Zoom fatigue. She notes that identifying demotivators can be easier than identifying motivators, because people find it simpler to say what frustrates them.

A Stanford study suggested that seeing your own video is a major cause of fatigue, like walking down the street with someone holding a mirror. The team made it possible to minimize self view while still broadcasting. To tackle low interaction they added reaction emojis, so people can show agreement without unmuting and interrupting the speaker.

For leaders who say no to delight, Changuel's first advice is not to try to convince them. She distinguishes perception, how you see delight, from perspective, how others see it, and notes that leaders likely see it as a cherry on top. Instead, find out what they value and show how delight helps reach that goal.

She illustrates this with a founder of a startup that helps musicians find curators. He came to her wanting OKRs and strategy, but when she asked whether his users were proud to use the product, he realized they weren't, because needing help made them feel small. Two weeks later he asked to rebuild the strategy around making users proud, which would drive word of mouth, and the company has since grown into the US. Lenny's takeaway is that you may not need to say the word delight at all.

On prioritization, she says to stop balancing delight against functionality and move to delight in functionality. Her 50-40-10 guideline says that 50% of the roadmap should be functional-only, because a product has to work, 40% deep delight, and only 10% surface delight, perhaps two features a year.

She gives the example of Spotify's roadmap, which mixed search improvements (functional only), video podcasts, and Canvas, the short looping video clips on tracks. Success, she says, comes from a bouquet of all three kinds rather than from picking one.

To make delight part of an organization, Changuel says it must be a mindset and a culture. At Google, delight was a permanent product pillar in the strategy, which served as a standing reminder to build features aligned with it. She also recommends making it a routine, like Spotify's quarterly squad health check or its monthly hack days and yearly hack week, with leaders showing they care.

She now helps companies run delight days. At Migros, the largest supermarket chain in Switzerland, teams brainstormed delightful features, demoed them, and chose a winner to implement. Lenny observes that many of the fastest-growing products today, such as Cursor, Lovable, ChatGPT and Linear, feel delightful in ways that look like more than a coincidence.

Changuel warns that surprise fades through the habituation effect, so a delight checklist should include a plan to maintain it. At Google Meet, background replacement began with blur, then static images, video backgrounds, immersive backgrounds and now AI-generated ones. Lenny compares this with Snapchat's constant stream of new lenses.

She also cites the Apple reactions feature, where hand gestures trigger fireworks, as delight gone wrong. Meet users complained they couldn't disable it, and she recalls a person showing a hurt finger to a therapist while fireworks went off. Her lesson is that delight needs to work for every situation. She closes with an overlooked benefit raised by a product director at Doctolib: building delightful features strongly motivates the product managers themselves.