A campaign report can contain accurate numbers and still create confusion. Marketing may count every submitted form as a lead. Sales may count only conversations about a supported job. Operations may care about appointments that can be scheduled. If those definitions remain implicit, each team is evaluating a different outcome.
Agree on the terms before deciding whether a campaign is working. With one shared definition, the meeting is about what to change, and nobody argues over labels.
Begin with observable qualification criteria
- Relevant visit
- A person reaches the page.
- Inquiry
- The person contacts the business.
- Qualified lead
- The inquiry meets the agreed rules.
- Booked job
- The customer chooses to proceed.
Define the minimum facts that make an inquiry relevant to the business. These may include a supported service, a serviceable location, a reachable contact, and enough context to choose the next step. Add criteria only when they reflect the operation and can be applied consistently.
Avoid subjective labels such as “good customer” or “serious buyer” unless the team can explain what evidence supports them. A customer may be relevant but early in planning. Another may be ready immediately but need work the company does not provide. Ready to buy and a good fit are two separate questions.
Separate the stages instead of compressing them
Contacts, reviewed inquiries, qualified requests, appointments, estimates, and accepted work should remain distinguishable. The exact stages depend on the business, but each should represent something that has actually happened. A phone button click is not automatically a completed call, and a requested appointment is not a confirmed visit.
| Label | Example definition |
|---|---|
| New inquiry | A distinct contact asking about possible new work. |
| Reviewed inquiry | A request assessed against the agreed fit criteria. |
| Qualified request | A reviewed inquiry that meets those criteria. |
| Completed next step | An actual conversation, visit, or assessment occurred. |
| Accepted work | The business recorded the customer's acceptance through its process. |
Adapt these example definitions to your business; they aren't industry standards. Use the terms your staff can reliably record. We hold our own network to this. Our Indiana roofing and restoration records count calls and form fills and leave missed calls out. We report those as inquiries, because a call isn't proof of a booked job. A small, consistent system is more useful than a complex funnel that only the dashboard builder understands.
Write the denominator beside the metric
A qualification rate is incomplete without stating what it divides by. Qualified requests divided by all contacts differs from qualified requests divided by reviewed new inquiries. Both can be calculated, but they answer different questions and should not share an unexplained label.
As a hypothetical example, suppose a group contains 60 new inquiries, 50 reviewed inquiries, and 30 qualified requests. The rate among reviewed inquiries is 60 percent. The share of all new inquiries already qualified is 50 percent. These are arithmetic illustrations only, and they aren't performance benchmarks. The ten unreviewed requests need a visible status rather than an assumed outcome.
Keep open opportunities visible
A project may remain open while the customer gathers information, schedules an assessment, or reviews an estimate. Do not classify it as lost simply because the reporting month ended. Preserve its current stage and the date it entered the process.
When reporting acceptance, explain whether the denominator includes open estimates or only estimates with a decision. Show pending work separately so the reader can assess the limit of the current evidence. This is especially useful when comparing services with different decision timelines.
Define duplicates and existing customer contacts
A customer may submit a form, call, and later send a correction. Decide how those contacts connect to one inquiry where your records allow it. Counting each touch as a separate new lead can exaggerate acquisition and hide the actual customer journey.
Existing customer requests should remain identifiable. They may represent retention, additional work, or ordinary service communication. The business can choose to report those outcomes, but they should not silently inflate a new customer acquisition measure.
Use known reasons for mismatch and loss
Create a manageable reason list that staff can apply. Unsupported scope, outside territory, timing unavailable, duplicate, customer postponed, and contact not reached are different situations. Add an unknown category for cases where the team lacks evidence.
Do not automatically interpret nonresponse as a price objection or a competitor win. Those guesses can lead the business to change the wrong part of the offer. A report should separate what happened from what the team suspects may have happened.
Connect source information without overstating attribution
- Search → guide
- First discovery and useful reading.
- Return → inquiry
- A later visit can lead to contact.
Keep the recorded source and relevant page context with the inquiry as it progresses. A later booking should remain connected to the original request where the system supports it. If source information is missing, identify the gap instead of distributing the outcome among channels by guesswork.
A customer's decision can involve search, referrals, prior familiarity, and several visits. The report should explain the attribution method it uses. A recorded first source or last tracked visit can be useful, but neither should be described as a complete account of everything that influenced the sale.
Use a glossary the office can maintain
For each stage, record its meaning, who updates it, and what evidence confirms it. For each rate, record numerator, denominator, time basis, and treatment of unresolved items. Keep the glossary short enough to use during an ordinary review.
If the inquiry process includes forms, clear submission feedback helps distinguish a received request from later stages. W3C addresses feedback about form outcomes. The language customers see and the labels the office records should describe the same event.
Review the report against actual examples
Before judging a campaign, walk through a few real records with marketing and the receiving team. Confirm that everyone would apply the same stage and reason. Resolve disagreements in the definition rather than quietly changing the numbers.
Then ask what decision the evidence supports. A mismatch pattern may justify clearer targeting. Qualified inquiries that do not reach appointments may need a handoff review. Open estimates may simply need more time. A shared qualified lead definition makes those distinctions possible and gives the next marketing decision a stronger foundation.
Should a definition change to improve a reported rate?
No. Change it when the business's actual scope or evidence justifies a revision, then explain the effect on comparisons. Reclassifying inquiries to make a chart look better undermines the report's purpose.
What if some records are incomplete?
Show the gap and limit the conclusion. A reviewed subset can still support a useful discussion when its scope is clear, but it should not be presented as the complete pipeline.
Continue the work
For the broader approach, read Lead Qualification for Service Businesses: A Practical Guide, How to Review Lost Inquiries Without Blaming the Channel, Use Service Boundaries to Improve Inquiry Quality.
Bring your inquiry criteria so we can discuss what belongs in campaign reporting.
Related services: paid campaign services and workflow automation.
