Outbound clicks, destination sessions and leads: what each number proves

A practical checklist for interpreting portfolio referrals: define the event, verify the destination, compare consistent periods and separate a click from an arrival or enquiry.

A click from a group website to a company website shows an attempted handoff. It does not, by itself, prove that the destination loaded, that a new session was recorded, or that somebody became a lead. Treating these as separate observations makes a portfolio dashboard more useful and prevents a healthy-looking click count from hiding a broken destination.

This operating note proposes a measurement checklist for independent companies. It describes a reporting method, not Oryvelon's conversion results or an assertion that every company has implemented it.

Define the three observations

Outbound click: the originating site recorded someone activating a link to another domain. Google's GA4 outbound-click documentation explains that enhanced measurement can detect these clicks. It also notes an important exception: links to domains configured for cross-domain measurement do not trigger outbound-click events through that mechanism.

Destination session: the receiving website recorded a visit under its own measurement setup. The origin's click event cannot establish this. A redirect, an unavailable page, a parked domain, or a visitor leaving before measurement runs can separate a click from an observed arrival.

Lead: a defined expression of interest, such as a successfully accepted enquiry. Google lists generate_lead among its recommended GA4 events. A click on a portfolio card should not acquire that meaning simply because it is convenient to count. Record the business action that actually occurred.

Write a small measurement contract

Before comparing totals, document what each number means. A practical contract can fit into six fields:

• Question: which decision will this report support?
• Trigger: what exact user action produces the event?
• Destination: which company and public landing page are intended?
• Counting unit: events, users, sessions, or accepted enquiries?
• Window: which completed dates and reporting time zone are included?
• Known gaps: what cannot be observed reliably?

For example, a hypothetical group may use a custom event named company_outbound for portfolio links. The name alone proves nothing about its implementation. Confirm whether it fires once per activation, which links qualify, and whether another tag records the same action. Do not add both event totals together and call the result unique visitors.

Keep the contract independent of a dashboard. When a tag changes, a future reviewer should be able to see why the series before that date is not directly comparable with the series after it.

Check the destination before interpreting demand

Open the actual linked URL, follow its redirects, and confirm that the final page belongs to the intended company. A successful HTTP response can still be a registrar parking page. Record the destination's visible status, not merely whether a server answered.

Next, check whether campaign parameters survive any redirect. A naming convention helps the destination interpret where an arrival came from; it cannot make an unavailable product available. The existing guide to UTM standards across brands covers link naming. This checklist addresses the separate question of what can be concluded after those links are used.

If the destination is not ready, correct misleading link wording or publish a truthful status on the relevant company page. Do not explain away the discrepancy as weak demand until the handoff itself works.

Compare like with like

Use completed periods of equal length and record the date on which each report was checked. Separate event counts from user counts: one person can activate the same link more than once. Avoid a ratio whose numerator is clicks from one property and whose denominator is sessions from another unless its limitations are explicit.

Consider this hypothetical example: an origin records 20 outbound events while the destination reports 12 attributed sessions. It is tempting to label the difference as eight lost customers. The observations do not support that conclusion. Repeated clicks, attribution differences, reporting delay and collection gaps are possible explanations. Investigate them before assigning a cause.

A missing event row also needs care. Write “not observed in the selected report” until the event's configuration, collection and reporting availability have been checked. That wording preserves the difference between no measured events and a verified absence of the underlying behaviour.

Turn the result into a decision

Use the smallest action that addresses the evidence:

1. If the destination is wrong or parked, fix the destination or its public status before changing marketing copy.
2. If event definitions differ, align the report's definitions before comparing performance.
3. If arrivals are observed but enquiries are rare, review the receiving page's promise, next step and form behaviour.
4. If the sample is too small, record the baseline and revisit it rather than declaring a winner.

For a group of independent companies, aggregate reports can answer these questions without constructing a person-level journey across brands. Share the observation, its definition and the decision it supports. Keep an outbound click as evidence of a handoff attempt, a session as an observed arrival, and a lead as the business action it was defined to represent.

Published by Oryvelon · About the group · Our companies