01. A director wants to hear about a sharp fall in weekly signups without opening a project to check.
What meets that requirement?
a) Anomaly detection applied to the signup trend in a project
b) Anomaly detection applied to the signup trend in a project
c) A scheduled delivery of the signup project each week
d) An alert configured on the signup metric
02. A stakeholder assumes that switching on Virtual Analyst will improve the quality of the data the site collects.
What correction should the analyst offer?
a) It analyzes data already collected and changes nothing about collection
b) It repopulates historical variables and backfills the gaps once a baseline exists
c) It enforces the naming standards agreed for components
d) It applies a correction factor to metrics that show unusual variance
03. An analyst is choosing components for a metric measuring completed purchases.
Which kind of component supplies that count?
a) A conversion variable, which persists a value for attribution
b) A traffic variable, which captures a value on the hit without persisting
c) A success event, which counts an action the implementation records
d) A classification, which adds descriptive attributes to collected values
04. An analyst needs video starts expressed relative to the traffic that could have produced them, so campaigns of different sizes compare fairly.
Which formula expresses that?
a) Video starts divided by video completions
b) Video starts divided by visits
c) Video starts divided by unique visitors
d) Visits divided by the number of video starts
05. An analyst shares a project containing a metric they built, and a colleague reports that the column shows nothing.
What most likely needs to happen?
a) The metric itself needs sharing as a component
b) The project needs re-sharing after the metric is approved
c) The metric needs converting into a segment before it can travel
d) The colleague needs to rebuild the metric, since the description records the formula
06. An organization marks certain calculated metrics as approved.
Which two outcomes does that produce?
(Choose two.)
a) The metric is recalculated using a more precise attribution model
b) Analysts can identify the sanctioned version among near-duplicates
c) Teams quoting the KPI are quoting one agreed formula
d) The metric becomes available in every report suite automatically
07. Anomaly detection flags an unusual spike in orders on a single day, and a stakeholder concludes that the new campaign caused it.
What is wrong with that conclusion?
a) The expected range is built from the campaign period rather than from history
b) Detection reports only that the value departed from what was expected
c) Contribution analysis has to run before any value can be flagged
d) A spike above the expected range is reported as an alert rather than an anomaly
08. Conversion rate has held steady for a month, but the team suspects it is masking a rise in one customer group offset by a fall in another.
Which capability is best suited to testing that?
a) Segment comparison across the two customer groups
b) An alert set on the conversion rate metric
c) Contribution analysis on the current month
d) Anomaly detection on the overall conversion rate
09. Orders dropped sharply on one day and the drop has already been flagged as anomalous.
What does running contribution analysis on it add?
a) It confirms whether the point really lies outside the expected range
b) It removes the anomalous day so the trend line reads more smoothly
c) It ranks the dimension values that account for most of the departure
d) It projects how the metric is likely to move over the following days
10. Two analysts build the same revenue-per-campaign metric and get different numbers from the same report suite and date range.
Which difference inside the definitions best explains it?
a) The attribution model on the campaign component differs
b) One metric was saved and the other was left unsaved in the table
c) One metric carries a description and the other does not
d) One metric was approved and the other was not