When the Product Disappears Into the Work

Just because your product got adopted doesn't mean it’s going to stay that way.

A team can switch to a new tool, use it for months, and never cross the line from "something we chose" to "something we'd fight to keep." The trial ends, the annual renewal comes up, and someone asks "do we actually need this?" — and nobody has a strong answer.

Continued Use — the final stage in the Jobs-to-be-Done adoption timeline — is what happens when a product crosses that line. When it stops being a choice and starts being a habit. It's not automatic. It requires specific things to go right.

Continued use isn’t automatic

It's the stage where a product has become the new status quo. The thing the person would now need to be pushed away from — the same way they were pushed away from whatever came before it."

The transition from First Use to Continued Use isn't automatic. It's the hardest conversion in the adoption timeline precisely because it doesn't have a moment. There's no equivalent of the pricing page, the demo, the migration decision. There's just a gradual deepening of dependency — or a gradual drift toward abandonment.

The investment loop

In 2014, Nir Eyal published "Hooked: How to Build Habit-Forming Products," which described what he called the Hook Model — a four-stage cycle of Trigger, Action, Variable Reward, and Investment that, when repeated, builds the habit loop that characterizes Continued Use.

The Investment stage is the one most directly relevant here. Every time a user adds data, customizes a workflow, builds a template, or teaches the product something about their preferences, they're investing — storing value in the product that they'd lose if they left.

This is the mechanism by which Continued Use builds its own gravity.

Spotify's Discover Weekly, launched in July 2015, is a clear illustration of how this works at scale. The playlist — generated weekly and personalized to each user's listening history — gets better the more a user engages with Spotify. Every song played, skipped, or saved refines the algorithm's model of that user's taste.

By 2015, Spotify had 75 million monthly active users. The year Discover Weekly launched, that number grew toward 100 million. The feature worked because it made continued use the prerequisite for the product's best version of itself.

A user who moved to Apple Music lost more than their playlist. They lost the model — the years of listening data that made Discover Weekly specific to them. They'd be starting from scratch. The more someone uses Spotify, the worse every alternative gets by comparison.

It should be noted that this same loop can work against users. When the investment creates genuine value — better recommendations, a more useful workspace — Continued Use benefits everyone.

When the investment mainly creates switching costs without improving the experience, the product isn't earning retention. It's taxing it. Eyal has acknowledged this tension directly — his follow-up book, Indistractable, was partly a response to criticism that the Hook Model gave companies a blueprint for manipulation as much as for value creation."

Accumulated value vs. accumulated habit

There's an important distinction between two things that both show up in Continued Use: accumulated value and accumulated habit.

Accumulated value is the stored investment — the data, history, configurations, and customizations that make the product increasingly useful over time and increasingly costly to leave. Spotify's listening history. Gmail's years of searchable archive. A Notion workspace built around how a team actually thinks.

Accumulated habit is behavioral — the muscle memory, the keyboard shortcuts, the reflexive reach for a specific tool when a specific job arises. The person who has used Excel for six years has a way of thinking about data that's organized around Excel's model.

Both create retention. But they create it differently.

Accumulated value creates rational switching costs. The person can articulate what they'd lose: "we have four years of project history in here" or "my recommendations won't be this good on any other platform."

Accumulated habit creates emotional switching costs. The person feels the loss before they can name it. The new tool that requires them to relearn where everything is doesn't just cost time — it costs the fluency that made them feel competent.

The Slack workspace nobody wants to rebuild

A Slack workspace that's been running for three years is a different product than a Slack workspace that was set up last month.

Three years of channels have been named and organized around how the team actually communicates. Integrations have been built to the specific tools that matter. Norms have developed — which channels are for announcements, which are for decisions, which are for informal conversation.

None of this was designed by Slack. It was built by the users, in the product, through Continued Use.

This is what makes switching from Slack genuinely disruptive in a way that the pre-adoption evaluation never fully captures. The competitor might match every feature. What it can't replicate is the three years of accumulated context that the workspace has become.

The product becomes a container for organizational memory that has no clean export path. Not because the data can't be moved — Slack supports data export — but because the meaning of that data is embedded in the system of relationships, norms, and context that grew up around it.

When things fall apart

Continued Use can last indefinitely — but it's never guaranteed. When it does end, it usually follows one of two patterns.

The first is gradual drift. The product stops delivering enough value to justify continued investment. A better alternative arrives and the accumulated switching cost, real as it is, becomes lower than the accumulated frustration of staying.

This is the pattern Evernote users described: years of continued use, followed by a long period of declining quality, followed by the eventual decision that the migration was worth it.

The second is a sharp break — a specific event that resets the calculation. A price increase large enough to make the switching cost feel worth bearing. A product failure that makes the accumulated investment feel like a liability. A company acquisition that changes the product's trajectory in ways the user can't accept.

Both endings have something in common: by the time they happen, the decision was already in progress. The user had been running their own evaluation of whether the continued investment was still worth it.

How continued use can last

The mistake most products make is treating Continued Use as a passive state — something that happens automatically once the user has committed.

It isn't. Continued Use requires ongoing value delivery. The product that coasts on switching costs is accumulating resentment, not loyalty. Users who stay because leaving is expensive are waiting for the switching cost to drop low enough to justify moving.

What characterizes products that sustain genuine Continued Use is progressive value — the experience that gets meaningfully better the more someone uses it. Spotify's recommendations improve with use. A well-structured Notion workspace becomes more useful as it grows.

There's a second requirement: the product has to surface the accumulated value regularly. Users who don't see what they've built in a product are less aware of what they'd lose by leaving.

Spotify Wrapped — the annual summary of a user's listening habits, launched in its current form in 2016 — is partly a marketing tool and partly a reminder of accumulated investment. It makes visible the data the user has stored in the platform over the year. Looking at twelve months of listening history makes the switching cost feel real in a way that abstract personalization doesn't.

The asymmetry that defines the future of use

There's an asymmetry at the heart of Continued Use that product teams rarely think about explicitly.

Everything the user builds in the product during Continued Use — the templates, the history, the configurations, the muscle memory — was built by the user, not by the product. The product provided the container. The user provided the value.

This means that the switching cost is, in a real sense, the user's own work held hostage.

Products that understand this design with some humility about it. They make data portable. They make export easy. They make it clear that the user owns what they've built.

Paradoxically, this tends to increase trust and continued use rather than reduce it — because users who know they can leave without losing everything feel less trapped, and users who feel less trapped are more willing to keep investing.

The product that makes leaving easy earns the kind of continued use that doesn't look for exits.

The product that makes leaving hard earns continued use that's always calculating the cost of walking out the door.