High Five Studio

November 2026

Two Progress Bars Beat One at Onboarding, 31% Finish More Steps

Splitting onboarding progress into two distinct tracks can lift step completion by around 30%, offering a simple way to keep users moving forward

Two Progress Bars Beat One at Onboarding, 31% Finish More Steps

Onboarding flows have a quiet, unglamorous problem: the further a user gets, the less motivating the remaining distance becomes. A single progress bar tells someone they are 40% done, then 60%, then 80%, and each increment delivers roughly the same small signal. The question worth sitting with is whether that flat signal is the best we can do, or whether splitting progress into two distinct tracks — say, setup tasks and first-value actions — changes how many steps people actually complete. Product teams that have tested dual-track progress reporting report completion lifts in the range of 30% on the later, harder steps, and the behavioral mechanics behind that number are more interesting than the number itself.

Why a single progress bar loses its grip after the first third

A progress bar is a goal-gradient device. It leans on the well-documented tendency for effort to intensify as a perceived goal gets closer — the effect that makes rats run faster near the end of a runway and coffee-card customers buy more frequently as their free drink approaches. In onboarding, the goal-gradient works beautifully for the first few steps. The bar moves visibly, the end is imaginable, and the user keeps going.

Then the middle happens. The bar is at 50%, the remaining work is a vague list of things the user does not yet care about, and the marginal value of one more step collapses. This is not a motivation failure specific to your product; it is what happens when a single aggregate metric tries to represent two different kinds of work. Some onboarding steps are obligations — verify email, set a password, connect a data source. Others are investments — invite a teammate, create a first project, configure a template. Obligations feel like tolls. Investments feel like building something. Collapsing both into one denominator makes the tolls look expensive and the investments look like more tolls.

There is a second problem: a single bar communicates a single remaining quantity, and human beings are poor at valuing a large, undifferentiated remainder. Kahneman and Tversky's work on how people evaluate prospects showed that losses and unfinished obligations loom larger than equivalent gains. A bar at 60% reads, to many users, as "40% of chores left," not "60% of value unlocked." The framing is doing damage that no amount of visual polish fixes.

The case for two bars: separating obligation from investment

Two progress bars work because they let you attach different psychological contracts to different kinds of steps. The first bar — call it Setup — is short, finite, and framed as a gate: three or four steps, each clearly bounded, each with a visible payoff. The second bar — Activation — is longer, open-ended, and framed as accumulation: the user is building capability, not clearing a checklist.

The split does three things at once.

It shortens the perceived horizon. A four-step setup bar is completable in one sitting. Users who finish it get a genuine closure signal, which matters because the Zeigarnik effect cuts both ways: unfinished tasks nag at attention, but finished ones release it. Giving someone a clean finish on the obligation track frees cognitive room for the investment track.

It changes what the second bar measures. If Activation tracks first-value actions rather than configuration, the bar moves when the user experiences the product working. That is a far better correlate of retention than "steps completed," and it means the bar is rewarding the behavior you actually want.

It creates a second goal-gradient. This is the underappreciated part. A single bar gives you one approach to the finish line. Two bars give you two, and users can be near the end of one while early in the other. The motivational pull of "almost done" gets applied twice, at different moments, without the user ever feeling like they are grinding through an endless middle.

The reported 31% lift in steps completed is consistent with what you would expect from stacking these effects. It is not magic; it is the same goal-gradient lever pulled at two points instead of one.

A concrete reference point

The clearest published evidence for splitting progress into sub-goals comes from consumer loyalty research rather than software. In a well-known field experiment on coffee cards, customers given a card requiring twelve stamps — but with two already filled in — completed purchases faster than customers given a ten-stamp card with none filled. Same effective work, different framing of progress, measurably different behavior. The lesson transfers directly: what a progress display implies about remaining effort matters as much as the literal count. Two bars are a structured way of controlling that implication, step by step, so that no single stretch of the flow ever reads as a long undifferentiated remainder.

Designing the split so it does not become two chores

The failure mode of dual progress is obvious and common: teams split the bar, then populate both halves with obligations, and the user now faces two checklists instead of one. The split only pays off if the two tracks are genuinely different in character. A few design rules that hold up in practice.

Keep the obligation bar brutally short

Three to five steps, maximum. Every step on this bar should be something the user cannot proceed without, and each should have a one-line justification visible at the point of action. If a step is nice-to-have, it belongs on the other bar or nowhere. The obligation bar's job is to end quickly and cleanly so the user gets a completion event early in the session.

Make the second bar about outcomes, not configuration

Activation steps should be phrased as things the user gets, not things the user sets. "Send your first invoice" beats "configure billing preferences." "Share a dashboard with a colleague" beats "set notification rules." This is where the variable-ratio idea deserves a caution: reward schedules that pay out unpredictably are powerful, but in onboarding they read as slot-machine design and erode trust fast. The reward here should be predictable and earned — the user does the action, the product visibly works, the bar advances. Predictability is the point. You are building a habit of competence, not a compulsion loop.

Let the bars move at different rhythms

If both bars advance in lockstep, users will perceive them as one bar with extra decoration. Deliberately stage them: the setup bar completes in the first session, the activation bar advances over the first week. This gives you two separate celebration moments and keeps the product present in the user's mind across the fragile early-retention window without nagging.

Show the second bar only after the first is done

There is a real argument for sequencing rather than showing both from the start. A user facing two incomplete bars at signup sees more unfinished work than a user facing one. Reveal the activation bar at the moment the setup bar completes — the transition itself becomes the reward, and the second bar arrives as an invitation rather than an additional debt.

What this changes about how you instrument onboarding

If you adopt dual-track progress, your analytics need to follow. The single "onboarding completion rate" metric becomes two metrics with different targets, and the interesting number is the gap between them — users who clear setup but stall on activation are a distinct population with a distinct problem, and lumping them in with users who never verified an email hides the diagnosis.

Three measurements worth adding:

  • Setup-to-activation handoff rate. What share of users who finish the obligation bar take at least one activation step within 24 hours? A low number means the handoff moment is not landing.
  • Step-level drop-off on the activation bar, indexed by position. The goal-gradient predicts drop-off should be lowest near the end of the bar and highest in the middle. If your data shows a flat or inverted curve, the bar is not doing its job and the steps are probably misordered.
  • Time-to-first-value, not time-to-completion. The activation bar should be optimized to shorten the interval between signup and the user experiencing the product's core promise. Steps that lengthen that interval without deepening the experience are candidates for removal, regardless of how good they look on a checklist.

One more instrumentation note specific to the Croatian market and similar smaller-language markets: activation steps that depend on network effects — inviting a teammate, sharing with a colleague — behave differently when the addressable user base is smaller and professional networks are denser. In a market like Croatia, a single well-placed invitation can cascade through a company faster than in a larger, more diffuse market, but the pool of plausible invitees per user is also smaller. That argues for weighting activation steps toward depth (configure a real project, run a real transaction) over breadth (invite many people), and for measuring the cascade separately from the individual action.

Risk, uncertainty, and the psychology of the second bar

There is a decision-making dimension here that is easy to miss. Onboarding is, from the user's side, a series of small risk decisions: each step asks them to invest time and sometimes data before they know whether the product will pay off. Under uncertainty, people are loss-averse — the pain of a wasted setup effort weighs more than the pleasure of a speculative gain. A single progress bar makes this worse by front-loading obligations: the user pays the tolls before seeing any return, and the longer the toll road, the more the sunk-cost anxiety builds.

Two bars restructure that risk profile. The short obligation bar caps the initial exposure — the user knows exactly how much they are putting in before the first payoff, and it is small. The activation bar then lets them invest incrementally, each step producing a visible return before the next is requested. This is closer to how people actually make repeated decisions under uncertainty: small stakes, fast feedback, adjust as you go. It also respects the finding that people evaluate sequences of outcomes, not just totals — a flow that delivers a quick win, then a bigger one, feels better than one that delivers the same total value in reverse order, even when the endpoint is identical.

Competitive play adds a wrinkle worth naming. Users arriving from a competing product carry a reference point: they know what "done" felt like elsewhere. A single bar invites comparison on a single axis — how many steps until I am set up? Two bars let you compete on a different axis entirely, framing your onboarding as shorter to first value rather than shorter in total. That is a more defensible claim and a more honest one.

Where to take this next

The practical next move is not a redesign. It is a measurement. Take your current single-bar flow, identify which steps are genuine obligations and which are first-value actions, and instrument the two groups separately — even if the UI still shows one bar. You will almost certainly find that the obligation steps complete at a high rate and the value steps do not, and that the drop-off is concentrated in the stretch where the single bar sits between roughly 40% and 70%. That stretch is your evidence.

From there, the sequencing question is the one worth testing first: does revealing the activation track only after setup completes outperform showing both bars from the start? The answer will depend on how heavy your obligation steps are and how quickly your product can deliver a first real result. Run it as a genuine experiment, with the handoff rate as your primary metric rather than raw completion, and let the goal-gradient do the rest — twice, at two different finish lines, instead of once at a line most users never reach.