Skip to content

MODULE 6 · LESSON 33 · Practical business analytics

Website analytics

Interpret session, event and conversion metrics without confusing activity with people or causal impact.

By Laurence Svekis · Self-paced · All worked data is synthetic unless you supply your own source.

Try the practice ↓

PredictInspectCalculateCheckExplainTransfer

Your starting point

Prerequisites: Use the ideas and vocabulary from lesson 32; revisit its worked examples if needed.

Definition: Website analytics records interactions under a tracking system’s definitions. A session groups activity by a rule; an event is a recorded action. A conversion is a designated outcome. These terms vary by implementation, so inspect the actual metric dictionary.

Read the implementation, not just the label

A field called users may represent browser identifiers rather than unique people. Consent choices, blockers, device changes and tracking failures can affect what is observed. A conversion count may allow several events per session or may count at most one. The academy’s synthetic traffic data explicitly defines conversions as sessions with one target event, avoiding an unstated vendor assumption.

Compare rates with compatible opportunities

Conversion-per-session requires conversions and sessions from the same scope and period. Comparing channel totals can reveal volume, while a rate describes efficiency under the chosen definition. Neither alone proves the channel caused the outcome. Returning visitors, audience intent and attribution rules can differ across channels.

Treat behavior metrics as evidence of activity

A long visit can indicate engagement, confusion or an idle tab. A high page count can reflect useful exploration or difficulty finding information. Link a metric to a concrete user task and supplement it with appropriate evidence before interpreting quality. Do not send learner answers or uploaded dataset contents into analytics merely to make a richer dashboard.

Two worked examples

Volume and conversion rate

Search records 200 sessions and 20 converting sessions, while Email records 50 and 10. Search produces more conversions, but its rate is 10% versus Email’s 20%. A decision about total reach differs from a decision about conversion efficiency, and audience differences remain possible.

Counting events

Suppose one session triggers three download events. An event-count field would show three, while a sessions-with-download field would show one. Dividing raw events by sessions can exceed 100%. That may be valid as events per session, but it is not the percentage of sessions that converted.

Where this becomes useful

A learning website wants to improve resource discovery. Search use and tool opens can indicate navigation activity, but they do not establish understanding. A task-based usability check or independent learner artifact answers a different question.

Make an attempt

Load Website traffic in the Dashboard Builder. Group sessions and conversions by channel, calculate a pooled conversion rate, and state the synthetic conversion definition. Add a limitation about what the activity metrics cannot show about learning.

Open the accompanying interactive lab →

Misconception check

Claim: More activity on a learning site proves more learning.

Inspect the correction

Activity can support investigation, but independent performance or other aligned evidence is needed to examine learning.

Independent application

Create a small anonymous website summary with sessions, target events and a clear conversion rule. Calculate volume and rate by channel, write two distinct findings, and propose a non-invasive check of whether users actually accomplished their task.

Completion records practice and self-review. It is not certification of statistical competence.