You are tracking social media leads when you record every inquiry by source, confirm the customer need, document the outcome, and separate discovery channel from attribution. Start by logging the contact point—a direct inquiry such as a comment requesting service, a message asking a question, a completed form, or an inbound call—with a unique inquiry identifier, the customer request, the team member who owns the response, and whether the conversation moved to a qualified lead, booking, order or payment.
Define qualification against the work you actually offer. A useful record can distinguish HVAC inquiry types, urgent versus scheduled plumbing requests, roofing estimate readiness or junk-removal scope and access questions. These are examples of different qualification rules, not reasons to count every contact as a booked job.
What lead tracking measures
Lead tracking follows a prospect from initial contact through each state until the business outcome is known. An inquiry is any customer-initiated question or request about your service. A qualified lead meets criteria you define—service area, timing, budget or stated intent. A booking is a confirmed appointment. An order is a placed, accepted commitment. Payment received closes the transaction. Each state requires evidence, and one person's journey may stop at any point.
Social media produces contact events—profile visits, post clicks, comments, direct messages—but the business result depends on what happens after the platform registers the event. Counting events overstates leads when the same person messages twice about the same job, clicks multiple links in one session, or never asks for a quote. Counting platform reach or engagement as leads confuses audience size with customer intent.
| State | Evidence | System holding the record |
|---|---|---|
| Inquiry received | Comment text, message thread, tagged post or call from profile number | Platform inbox, CRM notes, phone log |
| Qualified under stated rule | Service area confirmed, timing/budget noted, explicit request for quote or booking | Inquiry log, CRM status field |
| Booking confirmed | Calendar entry with customer name, agreed date/time, service description | Calendar, work-order system |
| Order placed | Accepted proposal, signed agreement, purchase-order number or cart checkout | Order ledger, e-commerce back end |
| Payment received | Bank deposit, card settlement, cash receipt or cleared check, net of refunds | Accounting ledger, payment processor |
Define qualification before you count qualified leads. A flower-delivery business may require delivery inside a ten-mile radius and at least 24 hours' notice; an automotive-accessories installer may need the year, make and model before quoting; a building-material supplier may set a minimum order value. Record the rule, apply it consistently, and mark inquiries that do not meet it as unqualified rather than forcing them into the qualified column.
Assign one inquiry identifier per customer need. When the same person sends three messages about the same deck repair, log one inquiry with three message timestamps under the same ID. When a past customer returns six months later with a new project, create a new inquiry ID under your stated rule—most businesses treat a returning customer's separate need as a new inquiry for tracking and attribution purposes.
Capture the contact point and preserve the source claim
Record where the inquiry arrived: Facebook Page message, Instagram post comment, profile link click that led to a website form, or phone call where the customer said they found the number on your Facebook profile. The contact point is observable. The true discovery source—where they first learned about the business—often is not.
Separate four pieces of information:
- Contact channel today: Instagram DM, website form, phone call.
- Tagged session source (if the website visit carried UTM parameters or referrer data and your analytics captured it):
utm_source=facebook,utm_medium=social, or browser referrer. - Customer self-report at intake: their answer to "How did you hear about us?" using clearly labeled options.
- Assigned attribution under your chosen model: which channel receives credit in your report.
UTM parameters label external links—profile URLs and post links you control—so analytics software can group referred visits by campaign when the destination page collects those parameters. UTMs do not install analytics, join a click to a phone call or message, or follow the visitor across devices. A click on a tagged link produces a session record only if the site implementation, consent and redirect chain preserve the parameters and the visitor grants tracking permission.
Platform analytics count actions inside that platform. Platform-reported action counts are not automatically verified unique business inquiries. Google Analytics assigns credit to important actions under its model and settings; this does not establish that every session is reported under the same attribution model. None of these systems automatically deduplicate the same person across tools. Summing Facebook messages, Instagram profile clicks and website form submissions produces a total event count, not a unique-person count.
When the true source is unknown, mark it unknown. Forcing every inquiry into a single known channel creates false attribution. A website form submission with no UTM, no referrer and no session history may have come from an organic search, a saved bookmark, a typed-in domain, an untagged email link, or word-of-mouth that led to a search. Record what you observed—"website form, no tagged source"—and code it as Unknown web or Direct/unknown in your report rather than guessing or assigning it to the last campaign that ran.
Set up a simple inquiry log
Build a tracking system that your team will actually use. A spreadsheet inquiry log works when one person handles all inquiries or the team shares a single file with clear row ownership. A shared customer-relationship-management system works when multiple people respond and need live status visibility. Either way, the log must record:
- Unique inquiry ID
- Date and time of first contact
- Contact channel (Facebook message, Instagram comment, website form, phone call from profile link)
- Customer request or question in a few words
- Qualified status under your stated rule (Yes, No, Pending information)
- Team member responsible for follow-up
- Outcome: quoted, booked, ordered, paid, declined, no response
A ready inquiry-log template provides column headers, dropdown options and a sample row. Customize the qualification rule, outcome states and channel list to match your business.
Preserve the raw customer answer to "How did you hear about us?" in one field and your coded attribution assignment in a separate field. The raw answer reveals language customers use, mentions you did not track, and combined sources—"I saw your Facebook post, then searched your name." The coded field groups similar answers for reporting: Facebook, Instagram, Google search, Referral, Repeat customer, Unknown. Allow multiple answers and an explicit Other/multiple code rather than forcing every inquiry into one bucket.
Check that the log does not contain personally identifiable information in column headers, formula text or fields that sync to analytics platforms. Customer names, phone numbers, email addresses and street addresses belong in your private business system under appropriate access controls, not in Analytics URL parameters or event labels. The inquiry ID, contact channel, qualification status and outcome are safe to report; the customer's identity details are not.
Move the inquiry through each state with evidence
An inquiry becomes a lead when you confirm the customer need and your ability to fulfill it. A lead becomes a booking when the customer commits to a specific date, time or delivery window and you confirm availability. A booking becomes an order when both parties agree to scope and price. An order becomes received payment when funds clear, net of any refunds.
Moving a social inquiry to a confirmed outcome requires clarity about what was agreed, who owns the next step, and what evidence proves the state change. "I'll call you Monday" is an intent, not a booking, unless the customer confirmed the time and you created a calendar entry. "Sounds good, send me a quote" is interest, not an order. "I'll pay when you deliver" is an agreement about payment terms, not received payment.
Update the inquiry log when the state changes. If the same inquiry progresses from qualified lead on March 10 to booked on March 15 to paid on March 22, record all three dates under the same inquiry ID. If the customer cancels, mark the outcome Cancelled and note the last completed state. If the customer never responds after the initial question, mark it No response after your documented follow-up window—most small businesses wait three to five business days.
Review inquiries stuck in Quoted or Booked status for more than two weeks. Either the customer is still deciding, the job was completed but the log was not updated, or the inquiry should be closed. Stale inquiries distort your qualified-lead count and obscure actual conversion patterns.
Run reports that match the business question
A tracking system answers questions only when the report matches the question. Counting inquiries by contact channel shows where people ask questions. Counting qualified leads by attributed source shows which channels produce inquiries that meet your criteria under your chosen attribution model. Counting completed paid orders by the source assigned to the original inquiry estimates the return from each channel, assuming your attribution model and cost allocation are documented.
Attribution assigns credit to a marketing touchpoint under a rule. It does not prove that the channel caused incremental sales or that stopping the channel would eliminate those customers. A customer may have found you through organic search, clicked a Facebook ad two weeks later, then called the number on your Google profile before booking—your attribution model decides which touchpoint receives credit. The report reflects the model, not an observed causal path.
Track these measures separately:
- Total inquiries by contact channel: observable, no attribution model required.
- Qualified leads by assigned source: depends on your qualification rule and attribution model.
- Bookings by assigned source: includes only confirmed appointments, counted once per customer need.
- Completed paid orders by assigned source: net of cancellations and refunds.
- Unknown source inquiries: the count of inquiries where neither UTM, referrer, nor self-report provided a known source.
A high unknown-source count is diagnostic information, not proof that tracking failed. It may reflect offline word-of-mouth, repeat customers who no longer remember the original source, or visitors who found your website through an untagged path. Acknowledge the unknown portion in your report rather than hiding it or redistributing it into known channels without evidence.
Compare inquiry counts month-to-month, but state alternative explanations for changes. A March increase in Facebook inquiries may reflect your posting frequency, a local event that increased group activity, seasonal demand, or a single popular post. A decline in Instagram inquiries during a week you posted daily does not prove that posting more hurts performance—the audience, offer clarity, contact path and follow-up speed all affect outcomes. When posting produces no inquiries, check each gap before concluding the channel does not work.
Prepare the report structure before you need it
Decide now what your monthly report will include so the inquiry log captures the necessary details. A useful monthly report leads with business outcomes—inquiries received, qualified under your rule, booked, completed and paid—then shows activity that supported those outcomes: posts published, comments answered, messages replied to, groups monitored.
Include these sections:
Inquiries and outcomes
- Total inquiries received this period, by contact channel
- Qualified leads by attributed source
- Bookings confirmed, orders placed, payments received
- Unknown source inquiries as a count and percentage
Activity completed
- Posts published by platform
- Comments and messages responded to, with median response time
- Groups monitored, recommendation requests answered
- Profile and post link clicks from platform insights
Diagnostics
- Inquiries with no follow-up response after three days
- Qualified leads that did not convert to bookings, with documented reasons when known
- Orders booked but not yet paid, aged by days since booking
- Proposed changes to qualification rule, contact paths, or follow-up process based on observed patterns
Avoid plotting attributed revenue by source when the total known cost, cost allocation method, and attribution model are not documented. If you calculate return, follow the defined-cost method and state every assumption.
Tracking checklist
- [ ] Define what qualifies as a lead for this business: service area, timing, budget, or explicit request
- [ ] Assign one unique inquiry ID per customer need; count repeat messages about the same job under the same ID
- [ ] Record contact channel, date, request, owner, and outcome in a shared log
- [ ] Capture raw self-reported source and maintain a separate coded attribution field
- [ ] Mark unknown sources as unknown rather than forcing them into a known channel
- [ ] Update inquiry status when it moves to qualified, booked, ordered, or paid with supporting evidence
- [ ] Exclude customer names, email, phone, and street addresses from analytics parameters and event labels
- [ ] Count events and unique people separately; do not sum platform counts as unique qualified people
- [ ] Review month-to-month changes with alternative explanations before concluding a single cause
- [ ] Prepare the monthly report structure now so the log captures what the report needs
Frequently asked questions
How do I count the same person who messages on Facebook and then calls?
Log one inquiry with the first message timestamp and Facebook as the contact channel. Note the phone call in the inquiry record as additional contact under the same ID. If your attribution model gives credit to first touch, assign Facebook. If it gives credit to last touch before booking, assign the phone call. State the rule and apply it consistently.
What if the customer does not remember where they heard about us?
Mark the source Unknown (customer does not recall) in the raw answer field and code it Unknown in attribution. Do not guess, assign it to the most recent campaign, or exclude it from the inquiry count. Unknown sources are real inquiries; your report should show how many fall into that category.
Can I use Instagram link clicks as my qualified-lead count?
Not directly. A link click is one observable event. The visitor may have clicked, looked at the site for five seconds, and closed the tab. They may have clicked twice in the same session. Or they may have filled out a contact form. Count link clicks as a platform activity metric. Count inquiries and qualified leads separately, using the lead definitions and evidence documented in your log.
How long should I wait before marking an inquiry as no response?
Most small businesses follow up once within 24 hours, again after three business days if no reply, then mark the inquiry No response after five to seven days. Document your follow-up rule and apply it uniformly. If a customer replies after you have closed the inquiry, reopen it with the new timestamp rather than creating a duplicate ID.
Do I need different tracking for delivery vs appointment-based services?
The states differ slightly. A flower-delivery or building-material business moves from inquiry to qualified lead to placed order to payment received; the booking state may not apply unless delivery requires a scheduled time window. An automotive-accessories installer or contractor moves from inquiry to qualified to booked appointment to completed work to payment received. Choose state names that match your fulfillment process and define the evidence that proves each transition.
In a barbershop, counting paid visits per barber adds separate ledger lines for what happens around a booked appointment: no-shows, late cancels, slots refilled from walk-ins or a waitlist, and clients rebooked before leaving.
Should I track inquiries that come through customer service channels separately?
Yes, if the inquiry is about an existing order, billing question, or complaint rather than a new service request. Use a separate log or a distinct inquiry type code so customer-service volume does not inflate new-lead counts. If a support conversation leads to a new upsell or repeat order, log that as a new inquiry with Existing customer support as the source.
What to do next
Review your inquiry records from the past three months. Identify the contact channels that produced the most inquiries, the percentage of inquiries where the source is unknown, and the conversion rate from inquiry to paid order for each assigned source. Document your qualification rule and the evidence required for each state change. Prepare the log structure and train every team member who will update it so the system produces consistent, usable data from day one.
If social posting has not produced measurable inquiries, a free preliminary SEO, GEO and SMM strategy can outline examples of potential customers for your business.
Free SEO, GEO and SMM strategy, explained with examples of potential customers.