If your firm cannot explain which inquiries its marketing produced or why the report disagrees with intake, a focused audit should give you a concrete place to start.
Thirty days is enough to produce a useful, scoped marketing operations audit if the firm has access to the relevant systems and people. It is not a promise to repair every problem or prove a revenue improvement within a month.
The audit's purpose is to explain how campaigns are administered, how information moves, whether reports can be reconciled, and which changes deserve attention first. The deliverable is a documented set of findings with evidence, owners, and a practical action plan.
Choose a manageable scope. A firm might audit one practice and its main acquisition channels before expanding to every office, vendor, and historical record.
Days 1–3: Establish the questions and access
Begin with the decisions leadership cannot make confidently. Examples include whether campaign costs are comparable, whether inquiry counts reconcile, or whether the current stack can support another practice.
Write the audit scope, systems, date range, participants, and exclusions. Name a sponsor who can resolve access and ownership questions. Identify the marketing, intake, technical, and finance contacts needed to explain the records.
Use appropriate read-only access where it serves the review. Follow the firm's approved handling process for prospective-client information and limit extracts to what the audit needs. A reviewer should not collect a full legal narrative merely to verify an inquiry identifier or status.
Save the current reports and relevant configuration records before making changes. The audit must be able to distinguish what it observed from what was later corrected.
Days 4–7: List the tools, owners, and manual work
List the tools involved in campaign execution, website management, tracking, inquiry handling, communication, integration, and reporting. For each, identify the account owner, administrator, purpose, cost boundary, and dependent processes.
Then map one common journey from campaign to inquiry to engagement. Include the manual steps. Ask staff to demonstrate their actual work rather than explain the ideal process.
Use the technology-stack guide to investigate overlapping tools and transfers. The inventory should reveal responsibilities as well as subscriptions.
By the end of the first week, you should have a system map, access gaps, a list of important definitions, and a small set of journeys to inspect more closely.
Days 8–12: Inspect campaign setup and measurement
Review a bounded sample of active and recently completed campaigns. Include different channels and any campaign associated with a known reporting discrepancy. State how the sample was chosen; a targeted sample is useful for finding problems but does not establish their firm-wide prevalence.
Check the brief, campaign identifier, destination, creative version, dates, budget, and approvals. Follow the public journey using controlled test activity where appropriate. Verify that the destination and next step match the campaign's promise.
Inspect the measured events. Determine whether a reported conversion means a click, a successful submission, a call event, or something else. Record the definition before evaluating the number.
Do not automatically change collection rules during the audit. For example, Google's internal-traffic filter guidance distinguishes testing from active filtering and warns that active exclusion permanently prevents the excluded data from being processed. A proposed correction should have its own reviewed implementation step.
Days 13–17: Trace records and exceptions
Select representative inquiry records with varied outcomes: engaged, declined, unresolved, duplicate, and transferred for review. Include a record associated with an integration failure if one is available.
Trace the source evidence, received time, owner, status history, next action, and engagement connection where applicable. Compare the sending and receiving systems. Identify missing records, overwritten fields, and definitions that differ between users.
Check the exception path. Who notices a failed transfer? Who handles an absent owner? What happens when an inquiry is merged or a person changes communication preferences? A successful ordinary path does not answer those questions.
The marketing analytics guide explains why the journey matters to measurement. During the audit, retain enough evidence to explain each finding without exposing unnecessary client information in the report.
Days 18–21: Check the report against the actual records
Choose a small number of figures that materially affect decisions. These might be distinct inquiries, engagements from a defined cohort, and included acquisition cost per engagement.
Reproduce each from the source records. Document the unit, period, exclusions, duplicate treatment, pending outcomes, and cost categories. Explain discrepancies rather than adding an unsupported adjustment to make totals match.
Keep observed errors separate from unresolved questions. “Six valid records are excluded by this status filter” is a finding if the records and rule have been verified. “The report may be missing referrals” is a question until evidence establishes what is missing and why.
Review financial definitions with the responsible owner. The profitability metrics guide provides context, but the audit should use the firm's actual treatment of fees, costs, and collections.
Days 22–25: Validate findings with the people involved
Share the evidence with the system and process owners. They may explain an intentional exclusion, a known timing delay, or a workaround the initial review missed. Update the finding when the evidence changes.
When a finding concerns phone handling, review a bounded, authorized sample of calls instead of inferring the experience from status fields. Record the sample rule, access and privacy limits, the promise and next step you observed, and any handoff that needs verification. Treat a targeted sample as diagnostic evidence, not a firm-wide prevalence estimate.
Use a consistent finding record: observation, supporting evidence, business consequence, likely cause, confidence, recommended action, owner, and acceptance condition. Avoid labeling every imperfection “critical.”
An illustrative finding register might look like this:
Scroll sideways to review every column.Each row is shown as a labeled card.
| Finding | Evidence | Recommended action | Acceptance condition |
|---|---|---|---|
| Campaign name variants split reporting | Verified variants map to one campaign identifier | Add a controlled mapping and naming check | Original values retained; consolidated view reconciles |
| Valid inquiries omitted from a report | Reviewed records fail an outdated status filter | Update and test the reporting rule | Corrected count traces to the eligible source records |
| Failed transfers lack ownership | A documented error has no assigned recovery task | Create an exception owner and escalation | A controlled failure is detected, assigned, and recovered |
| Acquisition measure omits agreed costs | Finance records contain excluded cost categories | Reconcile cost boundaries and restate the measure | Report names included costs and reproduces the calculation |
These are example findings, not assumptions about your firm. Replace them with what the audit actually establishes.
Days 26–30: Deliver a prioritized action plan
Rank findings by their effect on the prospect experience, reliability of material decisions, recurrence, and effort to resolve. Address lost or misrouted inquiries before cosmetic dashboard improvements. Consider dependencies: a report cannot be repaired fully until its underlying status definitions are settled.

For each priority, name the responsible person, the next action, necessary approval or dependency, and a realistic review date. Separate immediate corrections from projects requiring configuration, training, or vendor work.
The final package should include the scope and limitations, system map, definitions reviewed, evidence register, reconciliations, findings, and action plan. Preserve the baseline and record any changes made during the audit.
If access or history was unavailable, list the affected questions explicitly. A completed 30-day audit can have bounded coverage; it should not claim to have verified systems it could not inspect.
Completed example: one practice, two channels, four priorities
A fictional six-attorney personal-injury firm scopes its audit to one practice, paid search and referral-partner campaigns, and the prior 90-day inquiry cohort. Leadership’s blocked decision is whether to increase paid-search spend. The source report shows 148 “leads” and 11 “clients.”
By day 17, the audit has reconstructed the path:
- 148 contact events;
- 22 additional contacts belonging to existing inquiries;
- ten mutually exclusive tests or spam records;
- 116 distinct eligible inquiries;
- 111 records initially received by intake because five transfers failed; and
- 102 records in the dashboard because nine valid intake statuses were excluded.
After recovery and filter correction, the 116 inquiries receive mutually exclusive current dispositions: 58 outside the wanted matter profile, 22 outside the served market, 15 unresolved, and 21 qualified. The sum is 58 + 22 + 15 + 21 = 116. Of the 21 qualified inquiries, 14 complete consultations, five do not schedule, and two remain scheduled pending. Of the 14 completed consultations, eight sign and six do not. Seven matters open, and one signed engagement remains pending opening.
Finance also finds that the report’s $60,000 media numerator excludes $12,000 in campaign management and landing-page work included in leadership’s chosen acquisition-cost definition. Included cost is therefore $72,000. The report’s $60,000 / 11 “clients” = $5,454.55 is not a usable acquisition measure because the denominator does not match a verified stage. The audit can state $72,000 / 21 = $3,428.57 per qualified inquiry, $72,000 / 8 = $9,000 per signed engagement, and $72,000 / 7 = $10,285.71 per opened matter, each descriptive and limited to the defined cohort.
The day-25 validation produces this finding register:
Scroll sideways to review every column.Each row is shown as a labeled card.
| Priority | Finding and consequence | Action | Acceptance evidence |
|---|---|---|---|
| 1 | Five eligible inquiries failed transfer and had no owner | Add visible exception queue and coverage owner; recover records | Controlled failure appears once, is accepted, and replays without duplicate |
| 2 | “Client” mixes an unsupported report label with several business stages | Replace with inquiry, qualified, consultation, signed, opened, and pending definitions | Sample records classify consistently across intake and report |
| 3 | Nine valid statuses disappear from the dashboard | Repair filter while preserving original report and change date | Corrected 116 total traces to source records |
| 4 | $12,000 is missing from the chosen cost boundary | Add finance reconciliation and restate the period | Numerator reproduces from the approved cost schedule |
The firm does not approve a spend increase on day 30. It authorizes priorities one through four, assigns owners and 14-day acceptance reviews, and schedules the investment decision after the 15 unresolved inquiries and pending signed engagement mature. The audit is complete because it produced a defensible order and testable next actions, even though the budget question remains deliberately open.
Assign the fixes and check that they work
Schedule a review of the highest-priority actions after the responsible owners have had time to implement them. Check that the reported problem is resolved in practice; a closed ticket alone does not show that the fix worked.
Repeat the relevant checks when forms, channels, integrations, or reporting definitions change. A full audit does not need to happen every time, but its evidence should help the firm know which dependencies deserve attention.
The value of the month is a clearer account of how the system works and a defensible order of improvements. JurisOS is the current Juris Digital page associated with this work, but it does not publicly define a standard 30-day audit or repair package. If outside coordination may help, bring the scope, baseline exports, evidence register, reconciliation, priorities, and acceptance tests to a contextual JurisOS discussion and ask what can be defined in a written engagement.