Google Ads reports more inquiries. Intake says the inquiries are poor. Admissions remain below expectations, and the next meeting turns into a discussion about reducing the advertising budget.
Before making that decision, identify what happened between the first inquiry and the admission decision. Were people reached? Were they seeking a service the program provides? Could the program offer appropriate care when they needed it? Did the source attached to the inquiry survive the intake process?
Each answer points to a different response. A campaign attracting requests for an unavailable service needs different attention from a campaign whose appropriate inquiries reach voicemail. A source-reporting error requires another kind of correction altogether.
The useful report connects advertising to the complete lead journey and makes those differences visible. Here is how to build and use it.
Key Takeaways
- Follow the same group of inquiries through intake, allowing comparable time for outcomes. This month's admissions may have begun with earlier inquiries.
- Find the first stage where results diverge, then inspect the records. Poor contact rates call for a different investigation from a mismatch between demand and available care.
- Preserve captured source evidence separately from later staff updates. A changed source label should not erase the inquiry's history.
- Keep patient-linked analysis inside approved systems. An internal connection between advertising and an admission does not authorize sending that outcome to an ad platform.
Compare the same group of inquiries through intake
Dividing this month's admissions by this month's inquiries can produce a misleading rate. Some admissions may have started with inquiries from an earlier month. Some of this month's inquiries may still be waiting for an assessment or a decision.
For campaign evaluation, group inquiries by when they first arrived and follow what happened to that group. Analysts call this a cohort. Keep a separate view of admissions occurring during each month for operational planning. Both reports are useful, but they answer different questions.
Choose the follow-up period from the provider's actual intake timing. If many inquiries remain unresolved when a report is prepared, show the unresolved count and date of the review. Compare groups after similar amounts of follow-up; otherwise, an older group has had more opportunity to progress.
Also agree on what counts as a new inquiry. Three calls and a form from the same person about the same care need should not automatically become four separate opportunities. A family member may contact the provider on someone else's behalf. A repeat caller may be asking about an existing appointment. Intake needs a consistent way to connect related contacts while preserving their history.
That does not mean Google Ads and the intake system must display identical totals. They may count different events and assign them to different dates. Label those differences before interpreting a gap as lost demand.
Build an internal view of the complete lead journey
Use the provider's approved systems to connect each inquiry with its later disposition: the recorded result or current status of intake. Clinical details should remain available only to the people authorized to use them. Marketing needs an approved summary sufficient to investigate the campaign.
The following structure is a starting point to adapt to the actual care process:
| Stage | What the record should establish | A distinction worth keeping |
|---|---|---|
| New inquiry | When the inquiry arrived and which related contacts belong to it | A new inquiry versus a repeat contact |
| Contact established | Whether a substantive intake conversation occurred, and when | An answered call versus a completed conversation |
| Next step determined | Whether assessment or another action was appropriate | Not yet evaluated versus evaluated and unsuitable for this program |
| Assessment | Whether it was scheduled and completed | Waiting, cancellation and nonattendance are different outcomes |
| Admission or other disposition | What happened, when, and the approved reason if the person did not proceed | Unavailable care versus an unresolved inquiry or a decision to go elsewhere |
Avoid making “unqualified” the default destination for everything that did not become an admission. A person nobody reached has not necessarily been evaluated. Someone who needs a service the provider does not offer is different from someone who could be served but could not obtain an appointment. If both receive the same label, the report hides the action required.
Preserve the time each stage occurred, including the first response attempt. A callback timestamp alone does not establish that a conversation happened. When staff correct a status, retain the correction history so a later review can explain why the totals changed.
Find the first stage where results diverge
Consider this hypothetical example, created to illustrate the analysis. These are not client results or industry benchmarks. Two groups each contain 100 distinct inquiries associated with Google Ads, measured after the same follow-up period using the same definitions. For simplicity, all admissions in this example followed a completed assessment.
| Outcome | Earlier group | Later group |
|---|---|---|
| Distinct inquiries | 100 | 100 |
| Completed intake conversations | 70 | 40 |
| Completed assessments | 35 | 20 |
| Admissions | 14 | 8 |
The overall inquiry-to-admission rate fell from 14% to 8%. That is a real difference in the example, but it does not locate the problem.
Half of completed conversations led to completed assessments in both groups: 35 of 70 and 20 of 40. Admissions followed 40% of completed assessments in both groups: 14 of 35 and 8 of 20. The first visible deterioration was the share of inquiries that reached a completed conversation, which fell from 70% to 40%.
Start the investigation there. Check whether the later group contained more unanswered calls, invalid contact details or inquiries that arrived when intake coverage was limited. Also verify that staff recorded completed conversations consistently.
Those numbers do not establish that intake caused the decline. A change in the kinds of inquiries the campaign attracted could also affect contact rates. The comparison identifies the first place to investigate; the underlying records must establish the explanation.
Keep the limits of small groups in mind. One or two outcomes can move a rate substantially. A percentage change should prompt inspection of the counts and circumstances before it supports a budget decision.
Test the explanation before changing the campaign
Review a manageable selection of records with the person responsible for the relevant stage. Include inquiries that progressed as well as those that stopped. Looking only at unsuccessful inquiries makes it harder to identify what differs between the two groups.
These contrasting findings would justify different responses:
| What the review finds | What to verify next | A response the evidence could support |
|---|---|---|
| Unanswered calls cluster when intake coverage is limited | Routing, staffing and whether callbacks reached the callers | Repair routing or coverage; consider advertising schedules if coverage cannot be provided |
| Completed conversations reveal demand for a service the program does not offer | Relevant search terms, ad promises and landing-page language | Correct the mismatch and review targeting or exclusions |
| Appropriate inquiries cannot obtain timely care | Actual availability by program and the options intake offered | Align promotion with available services and review the handoff to other care options |
| Assessments are scheduled but frequently not completed | Appointment availability, communication and recorded reasons | Address the supported intake issue before treating every missed assessment as an advertising failure |
| Admissions move from paid search to “referral” in the source report | Original inquiry evidence and subsequent source changes | Correct reporting rules and preserve the source history |
For the hypothetical decline above, suppose the record review finds that many additional unanswered calls arrived after the staffed intake period. That finding would support examining coverage and callback handling. Restricting ad hours might be appropriate if the provider cannot serve those inquiries, but first check whether there is a workable response process. Cutting exposure could also remove opportunities to reach people who later connect successfully.
If the completed conversations instead reveal that callers consistently expect a different level of care, inspect how the advertising set that expectation. A search term is one clue. Read the ad and landing page as well: the misunderstanding may come from the provider's own wording. Care staff determine suitability; marketing uses approved summaries to improve how the service is represented.
Keep “could not serve” distinct from “should not have attracted.” A program with temporarily limited capacity may be receiving appropriate demand. That can justify a temporary campaign adjustment without establishing that the advertising strategy is fundamentally wrong.
Preserve the source history when intake updates a record
A source dispute can change the apparent diagnosis. An inquiry may begin with an ad click, while a family member later mentions a clinician's recommendation. Intake then selects “referral,” and the admissions report loses the recorded paid-search connection.
Both contacts may belong to the journey. Keep the original captured evidence, the person's reported source and later staff corrections as separate entries. A single editable dropdown cannot adequately represent that history.
With Google Ads auto-tagging, an available Google Click Identifier, or GCLID, provides evidence of an ad click. Save that exact identifier with the corresponding inquiry and carry it through later intake handoffs and record merges. Google's offline-conversion setup guidance specifically describes retaining it with the prospect's information. It must retain its original characters and capitalization.
Every inquiry record should support storing the available attribution evidence. Some inquiries have no GCLID. Preserve that gap honestly, along with whatever source evidence exists. A manually selected “Google Ads” label cannot recreate a missing click identifier for a GCLID-based import.
Test the entire handoff with an authorized test inquiry. Confirm that evidence captured at the website remains attached to the right inquiry after intake updates its source and the records are merged. Seeing the identifier in the landing-page address is only the beginning of that check.
For phone inquiries, verify what the call-tracking setup actually connects. Attribution to a campaign does not guarantee identification of every keyword or every later care outcome. Related calls and website forms also need reconciliation within the approved intake process.
A preserved GCLID helps investigate misattribution and can support an eligible offline conversion import. It does not establish that the ad deserves all the credit or that the person would never have contacted the provider otherwise. Reports should distinguish an observed connection from a claim that advertising caused the admission.
Keep the analysis within approved healthcare data boundaries
The operational report described here belongs in the provider's approved environment. An internal link between an inquiry and an admission does not authorize sending that event to Google Ads or another vendor.
HIPAA protects individually identifiable health information held or transmitted by covered entities and their business associates. Its scope includes information about receiving healthcare and paying for it, as well as health status. A call report can therefore contain protected health information even when it is being used for a marketing analysis. HHS's Privacy Rule summary explains that scope.
For example, an AI service analyzing identifiable intake calls alongside campaign performance may be processing PHI on the provider's behalf. When the vendor acts as a business associate, a business associate agreement must cover the actual service being used, alongside the other applicable safeguards. Opting out of model training does not replace that agreement. Have the provider's legal and compliance team review the arrangement before transferring the reports. HHS's cloud guidance explains the underlying vendor obligations.
Google imposes its own product restrictions. Its customer-data policy prohibits health-related conversion information in enhanced conversions and store-sales uploads. Hashing identifiers does not override that restriction. Other proposed measurement methods still require their own assessment of the exact event and data flow; a different matching method is not automatic permission.
The separate guide to healthcare marketing data, HIPAA and sharing with vendors covers that review in more detail. Even internal summaries need appropriate access controls, particularly where small counts or narrow categories could identify people.
Turn the diagnosis into an operating decision
A useful review ends with a specific finding, the evidence supporting it and someone responsible for the next step.
In the hypothetical example, that might be: “The decline first appears at completed intake conversations. The later group has more unanswered calls during uncovered hours. Intake will check routing and response coverage, and marketing will evaluate the next inquiry group after the same follow-up period.” If the evidence instead shows a service mismatch, the action belongs with the campaign and landing-page review. If the source history is wrong, repair the report before reallocating budget based on it.
Track the stage you intend to improve alongside later outcomes. More answered calls are useful, but the provider still needs to know whether appropriate conversations and assessments follow. Record other changes, such as staffing or program availability, that could explain the result.
There is also a bidding consequence. Smart Bidding optimizes toward the conversions or conversion value available through the account's measurement and goal settings. An outcome retained only in an internal report does not, by itself, return that outcome to Google. Where a later event is permitted to be shared, it must be correctly recorded and used in the relevant bidding setup. Our guide to choosing and verifying Smart Bidding conversion goals explains that process.
For the provider, the first decision is where the evidence says attention belongs. A reliable view of the complete lead journey lets the organization direct its effort to that point and then check whether the response helped.
If your campaign reports and intake reports lead to conflicting conclusions, contact Morpheus Consulting Inc. to discuss the measurement and marketing questions behind the discrepancy.
