
You can evaluate social media with a spreadsheet if you record costs, define useful outcomes, and admit what the data cannot show. The aim is to make a better operating decision, not to give every sale a perfectly certain origin.
Start with one campaign or one month. Choose a business outcome, keep the measurement window consistent, and separate observed revenue from revenue you can credibly attribute. If the sales cycle is long, begin with qualified opportunities and return to revenue later.
A small honest report is more useful than a dashboard that quietly turns every interaction into money.
Define the decision before the metrics
Write what you need to decide. “Should we keep spending four hours a week on this content series?” is more actionable than “How are our social channels doing?”
Choose the outcome that matches the work. A registration campaign can count completed registrations. A service business may track qualified inquiries and booked work. A customer-education series may reduce repeated questions, which needs a different measurement from sales ROI.
Do not force a monetary return calculation onto every useful activity. You can report a cost per completed registration or hours saved when that is the observable result. Label it accurately.
Define a qualified inquiry in advance. For example, the person needs the offered service, is in the area you serve, and asks about a concrete next step. The definition should reflect your business rather than an arbitrary point score.
Count the cost you actually consumed
Include production time, editing, review, publishing, and response handling. Add relevant tool subscriptions, contractors, and paid distribution. Decide how you allocate shared costs and use the same method each month.
If a founder spends ten hours on content, that time is not free merely because no invoice appears. Use an internal hourly estimate and state it. You can also report cash cost and time cost separately if that helps the decision.
Here is a hypothetical monthly budget:
| Cost | Amount |
|---|---|
| Eight hours of team time at $40 | $320 |
| Allocated software cost | $30 |
| Freelance editing | $100 |
| Paid distribution | $50 |
| Total campaign cost | $500 |
Do not count the full annual software bill in every monthly campaign. Do not hide editing time because the team calls it “just a quick change.”
Use contribution rather than sales totals
Revenue alone overstates the money available to cover marketing when delivery has variable costs. For a practical operating calculation, use the contribution from attributed sales: revenue less the variable costs needed to fulfill those sales.
A hypothetical campaign produces $2,000 of attributable sales, with $1,200 of associated variable costs. Contribution is $800. With $500 in campaign costs, the calculation is:
ROI = (attributed contribution - campaign cost) / campaign cost × 100
ROI = ($800 - $500) / $500 × 100 = 60%
Using $2,000 revenue in the numerator instead would report a much larger number while ignoring fulfillment cost. State which measure you use and avoid mixing revenue-based and contribution-based figures in one comparison.
The formula is simple. Establishing what is attributable is the harder part.
Give links consistent labels
Google Analytics documents UTM parameters for identifying campaign traffic. Use consistent values for source, medium, and campaign; use content to distinguish versions when useful. The values are case-sensitive, so inconsistent capitalization can split reporting.
For an invented campaign, a link could be:
https://example.com/workshop?utm_source=linkedin&utm_medium=organic_social&utm_campaign=workshop_november&utm_content=checklist
Keep a small naming sheet. Do not invent a new medium for every post. Test the final link, including any redirect and the destination form, before publishing.
Tagged links identify visits that retain the parameters. They do not prove that the click caused the sale or capture every person who later searches your brand on another device. Treat them as one source of evidence.
Add a simple question at the point of inquiry
Ask how someone found you or what helped them decide. Keep the question short and optional when appropriate to your process. Preserve the answer rather than forcing every response into the nearest tracking label.
Someone might say they saw a LinkedIn post, received a recommendation, and then searched your business name. That sequence is useful. Assigning the entire sale to the final search would lose the social contribution; assigning it entirely to the post would also simplify too much.
Keep tracked source, self-reported source, and your attribution decision as separate fields. When they disagree, retain the disagreement. You can review it later without rewriting the underlying evidence.
Do not put personal information into UTM parameters. Campaign labels should describe content and distribution, not identify individual customers.
Build a scorecard with clear definitions
Use one row per campaign and a small supporting lead log. Your main scorecard can include cost, visits, qualified inquiries, sales, attributed contribution, and observation window.
Record the date of inquiry and sale. If a November inquiry closes in January, decide whether the result belongs to the November cohort and keep that rule. Otherwise one month can appear weak while the next receives revenue from work it did not do.
A simple funnel might show 200 visits, twenty inquiries, eight qualified inquiries, and three sales. Those are illustrative numbers. Calculate visit-to-inquiry and qualified-inquiry-to-sale rates separately; each answers a different question.
When a metric is unavailable, leave it blank. A blank reach field should not become zero and trigger a report claiming infinite efficiency.
Keep network metrics in their proper role
Reach, views, saves, and replies can help explain what happened before an inquiry. They are not revenue. Different networks define and expose metrics differently, so avoid summing unlike measures into one universal engagement score.
Postbean has real metrics integrations for Facebook, Instagram, Threads, Bluesky, and YouTube. Its analytics are not a verified all-network revenue-attribution system. For other destinations, use native reports and your own inquiry records where available.
The analytics feature can support content review, while tagged links and your sales records answer different questions. The LinkedIn and Instagram pages describe publishing destinations, not a promise that every metric exists for both.
Report an attribution range when certainty is weak
You can show a conservative case based on clearly connected sales and a broader case that includes plausible assisted sales. Explain the inclusion rule and avoid presenting the broader case as proven.
Suppose conservative attributed contribution is $400 and the broader case is $800, with $500 cost. The contribution-based ROI range is −20% to 60%. The wide range is a finding: attribution uncertainty is large enough to change the decision.
Do not fill that uncertainty with invented lifetime value. If repeat purchases matter, use actual cohort records and a stated period. Until those records mature, report the current result and what remains open.
Check the destination before blaming the channel
Follow the same path as a prospective customer. Open the campaign link on a phone, read the offer, submit a test inquiry, and confirm that someone receives it. Record test entries so they do not inflate the lead count.
If visits arrive but inquiries disappear, inspect the form and the offer before concluding that the audience is wrong. If inquiries arrive but nobody responds promptly, improving follow-up may matter more than producing another post. Your scorecard should identify where evidence stops, so the next change addresses an observed problem rather than a convenient explanation.
Turn the report into one operating change
Review monthly, or after enough time for the relevant buying cycle. Decide whether to continue, change the offer or content, improve the destination page, or pause the work.
Use the calendar to connect planned content with campaign names and review dates. Keep the spreadsheet small enough that someone updates it next month.
If scheduling and reviewing network content in one place would help your routine, create a Postbean account. Keep your sales and cost evidence alongside it so the return calculation remains inspectable.
Sources
- Google Analytics: campaign URL parameters — checked November 22, 2026