Each points to a different part of the business. A report that combines them into one result can conceal the decision you need to make.
Collective’s Channel Health Framework separates the job of the channel from the performance of the page and the commercial outcome. We use that distinction to investigate the problem before choosing an intervention.
The working sequence is straightforward: confirm the data, inspect execution, review the content and message, then examine the strategy where the evidence supports it.
Describe the symptom precisely
“Marketing isn’t working” is difficult to investigate. “We received fewer suitable enquiries from the same campaign over the last completed review period” gives the team somewhere to start.
Define what changed, when it changed and how the result is measured. Identify the affected project, service, audience or channel.
Check the comparison. A launch period and a quieter ongoing campaign may have different objectives, budgets and audiences. Comparing their totals without that context may produce a misleading diagnosis.
Record known changes alongside the result. These might include a new page, different creative, a pricing change, reduced sales coverage or a tracking update. A change occurring at the same time is a clue, not proof of cause.
Check the records before interpreting the report
Trace a small sample from the source to the report. Does a genuine enquiry appear once? Is the source recorded correctly? Do the CRM stages mean the same thing across the team?
Separate test submissions and duplicate records. Check whether recent enquiries have had enough time to be contacted and assessed. A group still being worked cannot be compared fairly with one whose sales process is complete.
If a website change occurred, test the enquiry path. A broken form can reduce recorded conversions even when the audience and message remain appropriate.
Data checks should be proportionate. The aim is to establish whether the result is dependable enough to act on. Where it is not, repair the measurement and label the uncertainty rather than presenting an exact-looking conclusion.
Identify which layer needs attention
Give each part of the system a defined job and a useful signal.
| Layer | Question to investigate | Example signal |
|---|---|---|
| Distribution | Are we reaching the intended audience? | Visibility or relevant outbound response |
| Website or landing page | Can visitors understand the offer and take the intended action? | Enquiry conversion and observed form issues |
| Handover | Are genuine enquiries reaching a responsible person? | Assignment and contact within a defined window |
| Qualification | Does the enquiry fit the service or project? | Qualification rate with recorded reasons |
| Sales | Are suitable opportunities progressing? | Stage movement, win rate and cycle length |
The framework selects a primary signal for each job and uses supporting information to interpret it. Those signals remain influenced by context. A page conversion rate, for example, can change because the traffic mix changed even when the page did not.
For paid Google Search, impression share compares impressions received with estimated eligible impressions. It is an auction visibility measure, not a measure of all demand in the market. Google Ads Help: about impression share.
For organic search, a fixed group of relevant queries can provide a manageable visibility check. State which queries are included and how positions are assessed. The result describes that sample, not the entire market.
Inspect execution first
Check whether the activity is being delivered as intended. Look at targeting, placements, campaign settings, destination pages and the information passed between systems.
For an email, inspect the audience and destination link. For a landing page, inspect the form. For a lead handover, inspect assignment and overdue tasks.
A concrete execution issue gives you a testable action. Correct it, record the change and review what happens. Avoid changing several unrelated elements at once unless the situation requires it; otherwise, you may struggle to identify which change mattered.
This sequence is a starting discipline. If the offer has been withdrawn or the market has fundamentally changed, investigate that directly. There is no value in following a checklist past evidence that already identifies the problem.
Review how the content expresses the message
If execution is sound, inspect what the audience sees. Can they understand the subject quickly? Does the content answer the question that brought them there? Is the next step clear?
Separate the underlying message from its expression. A useful idea may be poorly explained, difficult to read on a phone or presented in a format that hides the relevant detail.
Then examine message alignment. Does the advertisement create the same expectation as the page? Does the enquiry process continue that promise? Are salespeople receiving prospects who understood the offer differently?
Use actual questions and feedback. If several prospects are asking for a service you do not provide, inspect the wording that may have created that expectation before blaming the entire channel.
Decide whether strategy needs review
Strategy includes the market you serve, the problem you address, the offer and the reason someone should choose you.
Revisit those decisions when the evidence points there. Suitable people may understand the offer but have little reason to act. The service may not match their requirements. Acquisition may be commercially difficult at the value and margin available.
Ask the team to state the diagnosis and its basis. “The audience is wrong” needs evidence from enquiry records or conversations. “The offer is unclear” needs examples of misunderstanding.
Where several explanations remain plausible, choose a practical test that distinguishes them. That might be a small messaging change, a review of lost opportunities or a conversation with recent prospects.
Work through one example
Illustrative scenario. Suppose a business reports that qualified enquiries have fallen after a new campaign launch.
The first check shows that total genuine enquiries are similar, but fewer have been assessed. The handover review then finds that a routing change left one service category unassigned.
The immediate action is to restore ownership and work the outstanding enquiries. The team still needs to examine whether the new campaign attracts suitable people, but the qualification total alone cannot answer that while records remain unassessed.
The diagnostic table would contain the observed issue, the evidence, the proposed action and the person responsible. It should also record when the team will review the result and what would indicate that the explanation was incomplete.
Turn the diagnosis into a decision
End the review with an action that someone can complete. Define the responsible person, the expected observable change and the next review date.
Keep the commercial outcome in view. Better channel response is useful only in the context of the job you need that channel to perform. Continue following the enquiry into qualification and sales rather than treating an upstream improvement as proof of revenue.
Collective helps businesses connect marketing, lead handling and sales into a measurable system. If you need to understand where performance is breaking, start with your current constraint.