Key Metrics Overview
Why Serving All Tobacco Content Readers Reduces Overall Profit? — User Segmentation and Prioritization
At around 11 PM on June 12, 2024, I finished reviewing the month's private messages in a co-working cubicle in Yuhang, Hangzhou. Notion stats: 312 valid DMs received in 28 days (excluding ads and arguments). I factored in the reply time — each thoughtful reply took 6–8 minutes on average, consuming about 35–40 hours on DMs alone. Transactions for the same period: 19 experience packs, 4 21-day check-in coaching programs, 1 one-on-one plan. By rough gross profit calculation for the month, each hour of DM work contributed less than ¥45 in gross profit — lower than my hourly rate for writing a reusable long-form article locally.
More striking was the breakdown: of the 312 DMs, only about 60 actually came with quit timelines, symptom milestones, or budget questions; the rest were "Does vaping count as quitting," "Checking for my boyfriend," "Just curious what you sell," "Can you guide me for free" — half of which went read after I gave thorough replies.
That day I wrote just one line in my memo:
It's not that the service attitude is bad — it's that the service targets were wrong.
This article explains why, in tobacco health/quit smoking content businesses, "serving every reader well" systematically reduces profit, and how to distinguish high-intent quitting users from casual browsers.
1. Let's Settle the Numbers: Profit Is Not Eaten by "Can't Write" but by "Can't Refuse"
In the content arena, there's often a moral sentiment: if someone reads you or asks you, you should be patient, comprehensive, and treat everyone equally. That works for public science platforms; for small teams that rely on manual conversion to make money, this is suicidal etiquette.
In business, there's a rough rule validated countless times (Pareto principle): roughly 80% of results come from about 20% of customers or leads. In email marketing and content conversion, it's also common to see a minority of assets and users contributing most leads and sales. You don't need to worship the exact 80/20 ratio, just accept that distribution is extremely skewed — profit is highly concentrated in a small cluster of "ready to act" people.
The quitting smoking track further amplifies this skew:
1. Intent itself is layered. In clinical smoking cessation, it's long been emphasized: quit intention is the foundation of successful intervention; the path differs fundamentally between those who are ready and those who aren't (motivational enhancement vs direct assistance). Content business operators who treat "passing by curiosity" and "going to battle next week" as the same service package are using the highest-cost labor for the lowest-conversion-probability work.
2. The decision window is extremely short. Many people search during a few days triggered by abnormal checkups, gum bleeding, childbirth, or an ultimatum from their partner. The window is typically 3–14 days. If you spend time on endless popular science to bystanders, high-intent leads go cold in the queue.
3. Labor costs are non linear. The first DM takes 5 minutes; by the 50th one explaining "Does nasal dryness count as withdrawal," marginal quality has already collapsed. Your hourly wage is flattened to the average value of all questioners — and that average is dragged down by a large number of zero-conversion users.
My view:
"Serving all tobacco content readers" sounds like brand vision, but in practice it's usually: using the response speed and solution depth that high-intent users deserve, subsidizing those who won't pay and aren't ready to act. The result is a lose-lose — paying users find you slow and shallow; you find yourself busy and underpaid.
2. Two Types of Readers: Both "View Tobacco Content," But Needs Differ by an Order of Magnitude
The table below is a rough segmentation I used in private channels in the second half of 2024 — not academically perfect, but executable.
| Dimension | High-Intent Quitting Users | Casual Browsers |
|---|---|---|
| Typical Trigger | Checkup, oral/respiratory symptoms, family pressure, Nth attempt after relapse | Scrolled past, curiosity, pastime, forwarding for someone, category comparison |
| Time Sense | Often carries "this week / next month / already on day X" | "Later" / "Just looking" |
| Question Type | Specific scenarios + constraints (years smoking, frequency, budget, failure history) | Open-ended, argumentative, asking for full summary |
| Attitude to Price | Asks "What's included, how soon results, refund policy" | First asks "Is there a free version" / "Can you send the full set" |
| Conversion Path | Materials → Small purchase/trial → Coaching or plan | Like, bookmark, occasional comment, mostly stay in public domain |
| Contribution to Your Hourly Rate | High (measurable LTV) | Near zero (ad or brand contribution at most) |
| Suitable Service Depth | Priority, structured, followable | Self-service content, template replies, no one-on-one |
High-Intent "Signals" Are More Reliable Than "Enthusiasm"
After I categorized DMs into four tiers (S/A/B/C) in September 2024, I found that true high intent rarely relies on slogans like "I must quit," but on verifiable details:
- Time anchor: "Starting next Monday" / "Insomnia on day 4 of quitting"
- Failure history: "Last year I lasted 11 days on willpower"
- Scenario list: Specific triggers like social events, overtime, after meals, driving
- Payment-related questions: Duration, service scope, whether relapse rescue is included — not just asking for the lowest price
- Execution actions: Already downloaded the check-in sheet, filled out the questionnaire, measured dependence level per your article's steps
Conversely, these I later default-downgraded (not blocked, but not occupying prime human hours):
- Repeatedly asking "Is there a cigarette with fewer side effects?"
- Only comparing e-cigarette/heated tobacco flavors, not discussing quitting goals
- "Helping my partner quit, the partner is not in this conversation" and refusing to involve them
- Asking for a voice call to "chat first" but refusing any structured questionnaire
My view:
High intent is not "the person who tells the saddest story," but "the person willing to be bound by a process." Those willing to fill out 8 questions, set a quit date, and accept "We don't guarantee success rate but we guarantee follow-up rhythm" — they deserve a spot on your calendar.
3. How Serving Everyone Specifically Drains Profit
Explained with a substitutable business model (you can replace the numbers with your own backend data, but keep the structure).
Assumption: 80K Monthly Unique Readers, Limited DM Capacity
| Metric | "Equal Treatment" Model | "Segmented Prioritization" Model |
|---|---|---|
| Monthly DM volume | 400 | 400 (same inflow) |
| Manual deep-chat ratio | 80% (320 DMs) | 25% (100 DMs, S/A priority) |
| Deep-chat time | 320×7 min ≈ 37 hours | 100×10 min ≈ 17 hours (deeper per DM) |
| Template/self-service | 80 DMs | 300 DMs |
| Deep-chat→conversion rate | 4% (mixed crowd) | 12% (screened) |
| Paid orders | 320×4%≈13 | 100×12%=12 |
| Average gross profit/order | ¥280 (diluted by low-ticket) | ¥520 (better small-order + coaching mix) |
| Monthly DM-related gross profit | ≈¥3,640 | ≈¥6,240 |
| DM hourly wage (gross profit/hr) | ≈¥98 | ≈¥367 |
Note: paid order counts can be similar, or even slightly lower after segmentation, but the hourly wage and profit structure diverge significantly. This table doesn't yet include: referrals from fast-responded high-intent users, repurchase consolidation packs, and the long tail from the 20 hours you saved to write reusable content.
In January 2025, I reviewed a small project (oral health + quitting care) using similar rules: after changing from "must reply manually within 24 hours" to "S/A within 2 hours, B workday template + info pack, C public FAQ only," monthly manual hours dropped about 40%, coaching high-ticket orders went from 3 to 7. Readership didn't surge — what changed was who occupied your time.
Three Hidden Costs Not Fully Captured in the Table
1. Emotional tax. Unmotivated bystanders are more likely to argue, question your motives, and demand you "prove you're not just selling." You get dragged into pointless debates to "maintain your persona," and your writing efficiency tanks the next day.
2. Product distortion. To please everyone, you turn your course into an encyclopedia that touches everything but makes no one feel "this was designed for me." Average order value stagnates while refund queries multiply.
3. Ad misjudgment. If your optimization targets are "comment count, completion rate, DM volume" rather than "questionnaire completion, 48-hour activity after adding contact, paid intent score," you'll buy more C-tier users and end up working for the platform with your profit.
4. Differentiation Strategy: Not "Discriminating Readers" but "Matching Costs"
1) Public Content: Open Information to All, Not Your Calendar
- For browsing layer: Mechanism explanations, timelines, risk popular science, self-check lists — let them self-screen. Place a "complete 3-step self-test before DMing" threshold at the end of articles — harsher but cleaner than "feel free to ask."
- For high-intent: Under the same topic, create a separate "action page": quit date setting, environment modification checklist, relapse contingency plan, service boundaries and price range. Don't hide purchase information for readers to guess.
In clinical work, those not ready receive motivational enhancement while those ready receive direct assistance — content business can analogize: public domain handles awakening and education; private domain seats are reserved for those who've crossed the willingness threshold. You're not a hospital, but you can borrow the "split-path" thinking rather than copying treatment protocols.
2) DMs and Community: Response Priority Written into SOP
The priority system I currently use (directly copyable):
| Tier | Conditions (Example) | Response | Action |
|---|---|---|---|
| S | Has set quit date + completed questionnaire + wants to start within 48 hours | ≤2 hours (work hours) | Enter main service communication, give clear next steps |
| A | Has failure history + specific symptoms/scenarios + asks about duration and boundaries | Same day | Send comparison table (self-help/small order/coaching), set confirmation point |
| B | Interested but no timeline | 24–48 hour template | Info pack + one-time limited Q&A access |
| C | Pure curiosity/helping someone not present/explicitly wants free long-term companion | No manual deep-chat | FAQ, old article links, polite closure |
Trade-off:
B/C are not enemies. They contribute readership, occasional sharing, potential future conversion. The mistake is serving C at S-level cost.
3) Product Structure: Let Segmentation "Grow on the Shelf"
- Free tier: One structured long-form article + one self-test page, addressing 70% of repetitive questions.
- Low-price confirmation tier (typically ¥49–99): Check-in sheets, care checklists, audio courses — used to buy "commitment to execute," not to make big money.
- Main profit tier: Limited-slot coaching/plan, open only to those who completed prior actions.
This way segmentation doesn't rely on your on-the-spot judgment — rules reject orders for you.
In November 2024, in an internal review, I cut the "unlimited free Q&A week" and replaced it with "3 question quota within 7 days after purchasing the info pack." There were some complaints that week, but the effective rate of main service consultations rose significantly the next month — because those coming in had already voted with their money and time.
4) Advertising and Topics: Optimize "Intent Density," Not "Audience Breadth"
- Ad creatives: reduce pure scare tactics "scary but no path forward," increase "common blockers on days 3–7 + action options."
- Topic selection: oral health timelines, quitting schedules, relapse rescue tend to filter high-intent better than "tobacco fun facts" (I verified across multiple oral health articles: DM "day X" phrasing density was significantly higher).
- Conversion events: don't set as "click avatar," set as "submit questionnaire" or "get a step-by-step tool."
5. Clear "Stop List" and "Keep List"
Recommended to Stop (or Strictly Limit)
1. One-on-one long voice calls for people without a quit date or questionnaire.
2. Long-text replies in comments for questions that could be addressed by a pinned FAQ (reserve depth for segmented private-domain users).
3. Using "free long-term companionship" to earn reputation — the quitting track's reputation comes from results and rhythm, not from being 7×24 online.
4. Making your product a "universal encyclopedia for all smokers" without stages and paths.
5. Using DM count and comment heat as core team weekly KPIs (should change to: S/A lead count, questionnaire completion, paid gross profit, 48-hour response rate).
Recommended to Keep
1. Publish a high-density FAQ or live Q&A once a week to absorb B/C information needs.
2. S/A response speed written into policy, more important than writing another viral article.
3. Structured information before all deep chats (years smoking, past attempts, target date, budget range). Missing info means fill the info, don't gamble on chatting.
4. Review DMs by gross-profit hourly rate monthly; if it falls below your reusable-content writing rate, tighten the inflow threshold.
5. Give gray-zone cases one upgrade opportunity: e.g., "Please have the person fill out the questionnaire and come back" — if they do, upgrade; if not, stay at self-service tier.
6. My Prioritization Principles (Can Be Used Directly as Team Consensus)
1. Everyone has the right to read; calendar access is scarce. Content is for everyone; deep service is for those who complete screening actions.
2. Profit follows "action density," not "sympathy density." Truly suffering but unmoving people need motivation and public-domain education, not your overdrawn evenings.
3. Segmentation is respect for high intent. They paid money and are in a critical window, but they're queued with "just asking" — that's the real unfairness.
4. Prioritization loses some surface popularity. What's lost is invalid noise; what's kept is accountable cash flow and reviewable service rhythm.
If your current state is: decent readership, full DMs, thin profits — don't rush to publish more. One week doing three things is enough:
1. Label your last 50 DMs as S/A/B/C to see where deep-chat time goes.
2. Add a mandatory screening action (3–8 question self-test or quit date entry) at the end of articles.
3. Rule: those who haven't completed screening only get info pack templates, no calendar access.
Then recalculate your "DM gross-profit hourly rate." Most accounts will see clearly for the first time:
What reduces overall profit is often not "too few readers," but "you're too good to all readers."