Static ETA creates distrust
A fixed estimate stops reflecting reality once traffic, preparation or rider conditions change.
03
Localia
Product Specialist · Experimentation
Designing a more predictable wait before building the technology behind it.
A Wizard of Oz experiment to test whether contextual, dynamic ETA updates could reduce uncertainty and support a confident product investment decision.
🍕
🥑
🥡
🥦
ExperimentDynamic ETA · Wizard of Oz
Role focusProduct Specialist
MethodWizard of Oz
Experiment1 week
Primary target−25% complaints
Product decisionValidate before build
01 · PRODUCT OPPORTUNITY
The problem was not only lateness.
It was uncertainty.
The product question was whether we could reduce perceived uncertainty without changing delivery speed — simply by making estimated arrival time more transparent and responsive to reality.
Building too early
A real-time ETA engine requires technical investment before the team knows whether the experience creates enough value.
Validate communication first
Simulate the proposition manually and let evidence determine whether the next investment is justified.
HYPOTHESIS
We believe dynamically updating ETA during the journey will reduce user uncertainty. We will know we are successful if delay-related complaints decrease by 25%.
02 · EXPERIENCE CONTEXT
The experiment lived inside
a credible delivery journey.
Before the ETA could change, the user needed a believable end-to-end experience. The prototype established context from order confirmation through rider assignment and pickup.
Order confirmed
The journey starts with a clear confirmation and a direct path to tracking.
Preparing the order
The user can understand progress before the rider begins the delivery.
Rider assigned
Introducing the rider makes the transition into active tracking feel real and trustworthy.
03 · WIZARD OF OZ STRATEGY
Test the value proposition
before the technology.
A background operator monitored active orders and manually changed the ETA when real-world conditions made the original promise unrealistic — simulating automation without first building it.
Start live tracking
Once the rider has the order, the initial live ETA becomes the baseline for the experiment.
Positive deviation
A faster route reduces the ETA and gives the user a proactive explanation.
Negative deviation
When congestion increases the ETA, the experience explains why the promise changed.
04 · DYNAMIC ETA EXPERIENCE
The ETA became a
communication system.
The strongest product insight was that the experience could not be reduced to a changing number. Context made the change understandable.
Faster route found
Communicate positive changes too. This reinforces that the system is actively following the journey.
Congestion detected
When the estimate worsens, explain the cause instead of silently extending the wait.
Traffic improved
Recalculate again when conditions recover, showing that the ETA remains dynamic rather than fixed.
05 · INTERACTIVE PROTOTYPE
Experience the experiment
as the user would.
The navigable prototype lets you move through the flow and see how the ETA changes are communicated in context.
If the embedded prototype is blocked by the browser, open it directly here ↗.
06 · CLOSING THE LOOP
Measure perception
after the wait is over.
The experiment continued through delivery and feedback so behavioral signals could be connected with perceived clarity, confidence and satisfaction.
Conditions improve
The user receives one more recalculation when the journey becomes more favorable.
Order delivered
The operational journey closes with a clear summary of the completed order.
Post-delivery feedback
The final survey captures clarity, reassurance and overall satisfaction after the complete experience.
07 · MEASUREMENT FRAMEWORK
Success required more than
one metric.
Complaint reduction was the primary success criterion, supported by behavioral and perception metrics that could explain why the experience worked — or failed.
Primary hypothesis threshold.
Fewer “Where is my order?” questions.
Users should better understand what is happening during the journey.
Measured after delivery.
08 · SIMULATED RESULTS
Enough signal to justify
a next product bet.
The simulated one-week experiment produced promising directional evidence. These are experiment results, not production metrics.
Exceeded the 25% hypothesis threshold.
Reduced “Where is my order?” contacts.
Participants found the information clear or very clear.
Participants felt calmer with contextual ETA explanations.
09 · PRODUCT LEARNINGS
We did not only validate an ETA.
We learned what creates trust.
Context matters more than the number alone
Messages such as “Traffic detected” or “We found a faster route” provided more reassurance than a silent numerical change.
Communication can create value before operational improvement
Transparency can improve the waiting experience even without increasing delivery speed.
A numeric update is not enough
The assumption that changing only the ETA would be sufficient was rejected; explanation was part of the product value.
Learning shaped the next roadmap
Update frequency, message type, ETA precision, personalization and repurchase impact became the next research questions.