Frictionless Is a Trap

For fifteen years, the entire discipline has been organized around a single instruction: remove the friction. Cut the steps. Kill the fields. Get them to value in one click, then get out of the way. It became the closest thing marketing and product have to a religion, and like most religion it stopped being examined. The frictionless funnel is now a reflex, not a decision.

It also has a blind spot large enough to lose a business through. A frictionless experience produces usage. It does not produce attachment. And the mechanism that turns a user into an owner — the thing that makes someone defend a product, renew without shopping, and forgive a bad week — runs on exactly the effort the industry has spent a decade deleting. This is where the IKEA effect stops being a curiosity and becomes a strategy. The brands winning the retention fight in 2026 are not the ones with the fewest steps. They are the ones who learned which steps to keep.

What the industry optimized away

Frictionlessness is easy to sell internally because it is easy to measure. Conversion rate goes up when you remove a form field. Time-to-first-action goes down when you strip an onboarding screen. Every one of those wins is real, and every one of them is recorded on a dashboard the day it happens.

What does not show up on that dashboard is what you traded for it. When a product asks nothing of a person, the person invests nothing in it, and an uninvested customer is a customer with no reason to stay. They arrived through the path of least resistance, and the path of least resistance runs both directions. The competitor who is one click cheaper is also one click away. You built a door so wide that people walk out of it as easily as they walked in.

The error is treating all friction as the same substance. It is not. There is friction that taxes the customer and returns nothing — confusing navigation, dead ends, forms that ask for what you already know. That friction is pure loss, and removing it is correct. Then there is friction that is investment — effort the customer puts in and gets something back for, effort that produces a result the customer can see is theirs. That friction is the raw material of ownership. Deleting it does not make the experience better. It makes it forgettable.

Labor is what turns a product into yours

The naming comes from flat-pack furniture, but the finding is older and stranger than the joke. In a set of controlled studies, Michael Norton, Daniel Mochon, and Dan Ariely had people assemble IKEA boxes, fold origami, and build with Lego, then measured what those people would pay for their own amateur output versus an expert's. They paid more for their own. They valued the wonky, self-made version as if it were the professional one, and expected everyone else to see the value they saw. The researchers named it the IKEA effect: when labor leads to love.

The reason this matters to anyone selling anything is that valuation is not a property of the object. It is a property of the relationship, and labor manufactures relationship. The customer who configured the workspace, named the project, imported the data, and set the preferences is not looking at your software. They are looking at something they made. That is a different psychological object entirely, and it does not get abandoned at the price the frictionless version does.

Norton's work also carries the boundary condition that many people copying the tactic miss. Labor increased valuation only when the labor succeeded. When participants built something and then had to take it apart, or failed to finish, the effect vanished. Effort that ends in a completed, kept result creates ownership. Effort that ends in nothing creates resentment. The distinction is the whole discipline in one sentence.

The head start that makes people finish

If labor builds value, the obvious objection is that people avoid labor — they will not start. The correction is one of the most quietly useful findings in the field, and it is twenty years old.

Joseph Nunes and Xavier Drèze ran a field experiment at a car wash. Customers were given a loyalty card requiring eight purchases for a free wash. One group got a card with eight blank slots. A second group got a card with ten slots and two already stamped. Both groups needed the same eight purchases. The only difference was the framing of where they stood. The group that had been handed two free stamps completed the card at nearly twice the rate — roughly 34 percent against 19 percent — and they moved faster as they got closer to the reward. Nunes and Drèze called it the endowed progress effect, and it sits alongside Ran Kivetz's work on the goal-gradient hypothesis: motivation intensifies the nearer the goal appears.

Put the two findings together and you have the actual playbook. People value what they build, but they finish what they have already started. So you do not ask for effort from a standing start. You give the customer a head start they did not earn, place the finish line in view, and let the goal gradient pull them through the work. The effort still happens. The investment still lands. You have simply removed the one form of friction that was pure loss — the cold, zero-progress beginning — while keeping every ounce of the friction that creates ownership.

Two kinds of friction, and the discipline of telling them apart

This is the operative skill, and it is not a slogan about "good friction." It is a specific test you can run on any step in any flow. Ask what the customer gets back for the effort. If the answer is nothing — the step exists for your convenience, your data model, your legacy process — remove it without ceremony. If the answer is a result the customer can see, keep, and recognize as theirs, that step is not a cost. It is the product.

A required credit-card field before value has been delivered is a tax. A five-minute setup that produces a dashboard configured to the customer's own accounts is an investment. They feel identical to a conversion-rate spreadsheet and they are opposites in the mind of the person doing them. The frictionless orthodoxy cannot tell them apart because it measures only the cost, never the return. That is why it keeps deleting the wrong things — it strips the setup that would have created an owner and leaves the dead form that creates a defector.

What it looks like when it is engineered

The strongest onboarding flows of the past two years are not the shortest. They are the ones that demand real work early and are unafraid to lose the tourists. Analytics teams now describe this openly; Amplitude's own guidance frames intentional setup effort as the IKEA effect applied to retention — asking users to invest in configuration precisely because people protect what they have configured. The counterintuitive move is that a step which raises early drop-off can raise long-run retention, because the people who complete it are no longer using your product. They are using theirs.

The same mechanism runs the most durable engagement features in consumer tech. A Duolingo streak is manufactured investment with a visible finish line renewed every day; the reporting credits its gamified, effort-anchored loops with cutting churn in major markets from the high forties into the twenties. Once a person identifies as someone with a two-hundred-day streak, the streak is no longer a feature of the app. It is a fact about them, and defending it costs less than losing it. That is the IKEA effect wearing a progress bar.

Spotify Wrapped is the same idea at campaign scale. The customer does the labor across an entire year without noticing — every listen is a keystroke in a self-portrait — and the product simply hands it back, assembled, in December. It works because it is unmistakably yours; Fortune reported over 200 million engaged users inside the first day of the 2025 release. Nobody shares a generic ad. Everybody shares the thing they made. Wrapped turned twelve months of passive behavior into an object the listener feels authorship over, and authorship is what gets posted.

Where the tactic turns on you

None of this is a license to add work for its own sake, and the failure modes are as reliable as the effect. Norton already named the first one: effort that does not complete destroys value instead of building it. A setup flow that is genuinely too long, or that breaks before it produces a result, is worse than no setup at all, because you have collected the cost and delivered none of the return. If you are going to ask for investment, you are responsible for making sure it finishes.

The second failure mode is coercion. Effort creates ownership only when it feels chosen. The moment the customer senses the work is being extracted for your benefit rather than invested for theirs, the same psychology that would have produced attachment produces reactance — the reflexive push against a freedom being taken. Designed friction and manipulative friction use the same materials and end in opposite places. The difference the customer feels is whether the result was theirs to keep.

The third is impatience with the measurement. Investment shows up in retention and defense, not in first-session conversion, and if you judge a setup flow on the metric it is supposed to trade against, you will kill it on schedule. The value of what a customer builds is not visible on the day they build it. It is visible on the day a cheaper competitor fails to move them.

Frictionless is a decision to be forgettable

The frictionless mandate was never wrong so much as incomplete. Remove the friction that taxes people and returns nothing — always. But understand what the reflex is doing when it keeps going. Past that line, every step you strip is a step of ownership you refused to let the customer earn, and a customer who owns nothing you made owes you nothing when the next option appears.

Influence is engineered, and this is one of the places the engineering is least obvious and most decisive. The competitor optimizing purely for the shortest path is building the widest exit in the category. The brand that knows which effort to keep is building something its customers made themselves — and people do not walk away from what they built. Be subtle about the work you ask for. Be certain it is theirs.