I. Setting It Straight: Profit Final Review, Not Data Final Review

# How to Determine Whether a Tobacco-Themed Campaign Truly Generates Profit

# How to Determine Whether a Tobacco-Themed Campaign Truly Generates Profit

In mid-March 2024, at a small desk in a shared coworking space in Binjiang, Hangzhou, I reviewed what initially appeared to be a "very successful" tobacco health information-feed campaign. The channel was short-video information feed + landing page form, centered around "oral changes after quitting smoking / nicotine dependence" as a educational funnel. The 7-day media spend was approximately 8,600 yuan. The backend numbers looked respectable: over 400,000 impressions, 2.1% CTR, 187 form submissions — at about 46 yuan CPL (cost per lead). At the time, someone on the team said: "We've got volume, we can increase the budget."

Two weeks later, after factoring in private-domain connection rate, effective consultations, actual transactions, and refunds, the picture turned cold: only 41 effective conversations, 6 orders for related services/products (average ticket 299–599 yuan), 1 refund, and verified gross profit of about 980 yuan. If only the media spend of 8,600 yuan is considered, this campaign wasn't "almost breaking even" — it was clearly a loss. Worse, someone had included 3 orders from concurrent organic search into the campaign's results — this was the kind of false prosperity I became most alert to afterward: it wasn't a calculation error, it was the metrics being unintentionally inflated.

Tobacco-themed content (especially quit-smoking education, health risks, and cessation tool consulting) has a bad temper: long conversion chains, prolonged hesitation, and compliance sensitivity. Getting impressive exposure is easy; landing actual profit is hard. The method below is the "final review standard" I later forced myself and partner accounts to follow — ignore the noise, only look at whether the money truly increased.

---

I. Setting It Straight: Profit Final Review, Not Data Final Review

Many people mix up ROI, ROAS, CPA, and CPL in one messy pot. My usage is blunt:

The simplified marketing ROI often written in industry is:

ROI = (Marketing Revenue − Marketing Cost) ÷ Marketing Cost × 100%

In tobacco-themed campaigns, I almost never use the "revenue version"; I enforce the gross-profit version:

ROI = (Attributed Gross Profit − Total Cost) ÷ Total Cost × 100%

The reason is simple: many peripheral products, course packages, and consulting packages have impressive surface GMV, but once gross margin is calculated and returns deducted, even ROAS 2.5 can be a loss. ROAS is a tactical dashboard; gross-profit ROI is the verdict.

There's another layer: a more rigorous calculation deducts "natural growth that would have happened without the campaign." Beginners who can't set up a control group should at least do one thing — campaign-attributed orders must be traceable to clicks/forms/promo codes/dedicated customer service links. Don't mix organic volume in to give yourself a medal.

---

II. A Simple Algorithm You Can Copy (With Cost Checklist and Time Windows)

1. Total Cost: Miss One Item, and ROI Is Lying to You

I require the campaign table to list at least these rows (fill 0 if absent, don't pretend they don't exist):

1. **Media Spend**: platform consumption for information feed, search, boost, promotion, etc.
2. **KOL/Distribution Fees**: rate cards, pure commission, sample costs
3. **Creative Costs**: filming, editing, copywriting outsourcing; self-operated count as labor hours (don't say "I did it myself so it doesn't count")
4. **Landing Page & Tools**: site building, forms, SMS, CRM, editing subscriptions — amortized over the period
5. **Fulfillment Costs**: if physical goods — product cost, packaging, shipping, damage; if services — delivery hours cost
6. **Platform Fees & Payment Processing**
7. **Refunds & After-Service** (actual or estimated based on historical refund rate)
8. **Compliance & Review Costs** (repeated creative changes, lawyer reviews, takedown and relaunch window losses — estimate as possible)

"Counting only media spend" is the most common self-anesthesia among tobacco content accounts. The 8,600 yuan spend looks like total investment, but adding two days of editing outsourcing at 1,200, form and SMS at 180, and customer service overtime at 600, the real investment is close to 10,600 yuan.

2. Revenue: Prioritize Verifiable Gross Profit

• **Physical/Standard Products**: Gross profit ≈ actual receipt − product cost − shipping − platform fees − refund losses
• **Consulting/Courses/Communities**: Gross profit ≈ actual receipt − delivery cost (your time has a price too) − refunds
• **Leads Only, Not Yet Monetized**: Don't claim ROI victory yet, at most report "acquisition cost and subsequent conversion assumptions"; **no payment means no profit ROI**

3. Time Windows: T+0, T+7, T+30

Quit-smoking related decisions are often not impulse purchases. I set three fixed windows:

• **T+0 (Same Night)**: Only look at spend, CTR, landing page arrival, form quality sampling — **forbidden** to declare profitability
• **T+7**: Connection rate, effective consultations, first orders, preliminary gross-profit ROI
• **T+30**: Supplementary orders, repeat purchases, refund-adjusted "near-final" ROI

In November 2023 in Shenzhen, I helped an oral health account review search ads. The same-day CPA looked great, T+7 had 4 orders, but T+30 had 2 refunds and a negative review costing 6 hours of customer service — the period's profit went from "small gain" to "break-even." Without a 30-day window, the adrenaline of the first 48 hours will trick you into increasing the budget.

4. Two Worked Examples (Real Structure, Easy to Adapt)

Case A: Strong Forms, Poor Profit

• Total Cost: Media 8,000 + Creative 1,000 + Tools/Labor 800 = **9,800 yuan**
• 200 forms, 50 connected, 5 orders closed, ticket 399, gross margin ~45% (after product/delivery), 0 refunds
• Attributed gross profit: 5 × 399 × 0.45 ≈ **897 yuan**
• ROI: (897 − 9,800) / 9,800 ≈ **−91%**

A CPL of 40 yuan in the backend would be praised as "stable" — that's false stability. The real problem is lead quality and conversion rate, not exposure.

Case B: Small Volume, But Calculable

• Total Cost: **5,200 yuan**
• 18 orders closed, ticket 299, comprehensive gross margin 50%, 1 refund (gross profit loss ~150)
• Attributed gross profit: 18 × 299 × 0.5 − 150 ≈ **2,541 yuan**
• ROI: (2,541 − 5,200) / 5,200 ≈ **−51%** (still losing, but the path is clear)

To push ROI above 0, either reduce acquisition cost, increase ticket/gross margin, or improve form-to-transaction rate — nail at least two of three before scaling budget. Throwing money at it is just buying the "illusion of busyness."

People in the industry often cite "5:1 revenue-to-cost ratio as decent" as an experience benchmark. My stance: use it as a rough reference, but tobacco compliance and conversion volatility is high — use your own gross-profit accounting, don't repeat others' slogans.

---

III. Identifying False Prosperity: Appearances, Tricks, Verification

Each item below is something I've repeatedly halted in campaign review meetings.

1. High Impressions, High Engagement, Orders Meh

Appearance: Hundreds of thousands of plays, comments saying "you're right."

Why it tricks you: Engagement comes from bystanders and trolls, not necessarily high-intent quitting smokers.

Verification: Check landing page arrival rate, form rate, consultation initiation rate; sample 20 DMs — how many ask about price/solutions?

2. High CTR, Low Conversion

Appearance: CTR looks great, optimizers are excited.

Why it tricks you: Clickbait headlines, fear marketing, exaggerated before-and-after — people click for entertainment.

Verification: Compare landing page bounce and dwell time; if clicks are expensive but forms are very low, kill the creative first, don't add budget.

3. Many Forms, Few Connections, Even Fewer Payments

Appearance: CPL meets target.

Why it tricks you: Competitors, students, people filling for fun, even dead numbers. Tobacco health topics especially attract "just want free advice" traffic.

Verification: Manual call-back within T+1; record connection rate, effective conversation rate, appointment rate. My experience line is strict: form campaigns with connection rate persistently below 25%, stop first, don't scale first.

4. ROAS Looks Good, Gross Profit Looks Bad

Appearance: Ad backend ROAS 3.0.

Why it tricks you: Calculated on GMV; doesn't deduct product cost, KOL commissions, big coupons, free shipping, returns.

Verification: Export orders and recalculate gross profit; recompute "Gross-Profit ROAS / Gross-Profit ROI." Post-coupon actual receipt is your real blood.

5. Coupons and Freebies Creating Fake Order Spikes

Appearance: Order volume doubles on promotion day.

Why it tricks you: You're pre-spending future profit for sales volume; may attract gift-chasers with zero repurchase.

Verification: Calculate "post-coupon gross-profit ROI"; separately track promo-order 30-day repurchase. If repurchase doesn't rise, it's a one-time firework.

6. Organic Volume Disguised as Campaign Volume

Appearance: Campaign launches, whole store/account sales rise.

Why it tricks you: Seasonality, trending topics, other videos, repeat customers are "credited" to ads.

Verification: Dedicated links, promo codes, landing page parameters, customer service first-touch tags; compare against non-campaign baseline days.

7. Last-Click Attribution's "Credit to Heaven"

Appearance: One channel's report shows all conversions.

Why it tricks you: A user may have watched 3 educational videos, then clicked a search ad to buy — all credit goes to search.

Verification: In small-budget phases, "turn-off channel exclusion testing" is more practical than modeling — turn it off for a week and see how much total profit drops.

8. Small Sample Treated as Trend

Appearance: 2 orders, ROI 400%, ready to go all-in.

Why it tricks you: Random noise.

Verification: Set a minimum decision sample (e.g., at least 30 valid leads or 15 orders depending on ticket) before talking about "model validated."

9. "Consultation Intent" Written as "Profit"

Appearance: 20 high-intent leads today.

Why it tricks you: Intent is not cash received.

Verification: Only payment success or contract signing enters the revenue column; intent enters "pipeline," not the ROI numerator.

---

IV. Extra Hurdles Specific to Tobacco Campaigns

1. **Compliance & Platform Policy**: Creative approval is difficult, sudden traffic restrictions, keyword pool narrowing. Spend fluctuations are often misread as "creative decay" when they're actually audit and targeting cuts. Flag "policy interference days" separately in reviews.
2. **Naturally Longer Funnel**: From "knowing smoking is harmful" to "paying for a service/tool" there's a trust gap. Pushing high-ticket items directly with performance ads invites resistance. I more commonly split campaigns into: front-end low-friction content touch → mid-end trust assets (case studies, mechanism explanations) → back-end clear offer. Front-end can tolerate "no direct ROI" but must define a **mid-to-back-end recovery timeline**, otherwise front-end becomes a money pit.
3. **Returns & Cooling-off Period**: Health-anxiety-driven orders have noticeably higher return rates than impulse snacks. If you don't calculate at T+30, you haven't calculated at all.
4. **Homogenization**: Oral ulcers, yellow teeth, withdrawal anxiety — when topics cluster, cheap CPM can still bring garbage conversions. Cheap traffic ≠ quality business.

I don't advocate using borderline tactics to promote tobacco consumption itself. If you're doing cessation and health-oriented content, you should define profit as money that is sustainable, verifiable, and doesn't cross red lines. An ROI built on non-compliant copy will be clawed back with interest when it gets taken down.

---

V. Actionable Campaign SOP (I Follow This Now)

Before Launch

• Write down: **Target Gross-Profit ROI** (e.g., ≥0% as trial success line, ≥30% as scale-up line — adjust per your cash flow)
• Write down: **Target CPA / Acceptable CPL**, and back-calculate:

`Acceptable CPA ≈ Per-unit Gross Profit × Safety Factor on Target Conversion Path`

• Set **Daily Spend Circuit Breaker**: e.g., 2 consecutive days below a conversion threshold and CPA exceeding target by 30% → auto-reduce budget or pause plan
• Attribution method documented: exclusive codes / parameters /客服标记, no post-hoc verbal credit splitting

During Launch (Track Only a Few Metrics Daily)

I only force-check: spend, CPA or CPL, landing page conversion rate, effective connection rate, confirmed same-day payments. Likes, follower growth — review weekly, don't enter daily budget decisions.

Post-Launch (Mandatory Accounting Meeting on Day 7)

Agenda is fixed to four lines:

1. Total cost summary
2. Attributed gross profit (after refunds)
3. Gross-profit ROI
4. False volume investigation (organic, coupons, dead numbers, attribution)

The accounting meeting can be 25 minutes, but without a table, don't discuss "feeling okay."

---

VI. My Clear View: What "Truly Generates Profit" Means

One-sentence standard:

Within the agreed time window (I recommend at least T+7, robust at T+30), attributed gross profit from traceable orders, after covering all related costs, still has surplus — and that surplus is not from one-time coupon abuse or non-compliant content — that's what truly generates profit.

Conversely, I judge the following as "not yet proven profitable":

• Only impressions, engagement, forms — no stable payments
• ROAS looks pretty but the gross-profit table is red
• Only flips positive by including organic volume or repeat customer purchases
• Sample too small to survive a one-week pause

My conditions for scaling up: gross-profit ROI consistently meets your scale-up line + lead quality metrics not deteriorating + compliant creatives are replicable. Missing any one — don't scale.

Conditions for stopping are simpler: persistently negative under total-cost accounting, and you've tried at least two of creative, targeting, and landing page structural changes with no improvement — at that point, continuing to burn is not persistence, it's sunk cost holding you hostage.

After that Hangzhou review in 2024, I removed "form volume" from the weekly report front page and replaced it with three big lines: Total Cost, Attributed Gross Profit, Gross-Profit ROI. The team went quiet, order volume dropped a notch, but the money in the account stopped contradicting the backend screenshots. Tobacco-themed campaigns can be lively, and they can be profitable, but rarely on the same vanity metric. If you have to choose, pick the one that goes into the bank statement.

---

Appendix: One-Page Accounting Template (Field List)

Copy into a spreadsheet:

• Campaign Name / Channel / Start-End Dates
• Media Spend, KOL Fee, Creative Cost, Tool Amortization, Labor, Fulfillment Cost, Platform Fees, Refund Loss → **Total Cost**
• Impressions, CTR, Forms, CPL, Connections, Connection Rate, Effective Consultations, Orders Closed, Ticket, Actual Receipt, Gross Margin
• **Attributed Gross Profit**, **Gross-Profit ROI**, ROAS (reference only)
• Includes Organic Volume (Y/N), Coupon Impact Amount, Compliance Issue Notes
• T+0 / T+7 / T+30 Conclusion Lines: Scale Up / Maintain / Stop
• Next Action: Revise Creative / Revise Landing Page / Revise Offer / Shut Down

Before the numbers are clear, I recommend blocking any "let's add another ten thousand to test" proposal. Money is limited; self-congratulatory data is infinite.