In September 2023, I spread out my spreadsheets from the previous 8 months in a shared workspace in Binjiang, Hangzhou, and for the first time I clearly understood the "customer acquisition cost." The public account had 21,000 followers, Douyin had accumulated over a million views, and I had added more than 1,800 friends on WeChat — it sounded decent. But after factoring in the costs of outsourcing writing, editing, news feed testing, tool subscriptions, and my own time valued at market rates, only 47 people had actually paid for smoking cessation-related services. The monthly service gross profit was about 18,000 yuan, while the book expenditure on the acquisition side plus hidden labor costs exceeded 24,000 yuan. Readership was growing, but the accounts were losing money.
From that day on, I stopped asking "did this post go viral?" and started asking only three questions: How much does it cost to bring in one paying user? How much gross profit can his first payment contribute? How long will it take to earn back that acquisition cost? The tobacco science track survived not by pushing out content harder, but by squeezing CAC down to a level that profit could cover.
1. First, Get the Numbers Straight: What Does "Covered by Profit" Mean?
The formula is blunt:
CAC = Total acquisition-related costs in a period / New paying customers in the same period
Don't just count advertising costs. I now regularly include these items:
| Cost Item | How I Calculate It |
|---|---|
| Content production | Writing/editing/proofreading billed by hour or by piece; video includes shooting and editing |
| Distribution and traffic | News feeds, search ads, platform boosting, Zhihu+ etc. |
| Tools | SEO tools, forms, WeCom SCRM, editing memberships |
| Private domain operations | Community manager hourly wage, welcome message materials, activity prizes |
| Labor imputation | My own time logged at replaceable hourly rate (otherwise you will always think it is "free") |
Covered by profit, I use two hard thresholds:
- Gross profit per transaction >= 1.5 x CAC (leave room for after-sales, refunds, and failed repurchases)
- CAC payback period <= 45 days (tobacco science customer unit prices are usually low; dragging into a quarter's cash flow will kill a small team)
Here is a reproducible example (the numbers are from my 2024 Q1 "28-day smoking cessation check-in accompaniment" type, not industry averages):
- Accompanying program unit price 399 yuan, fulfillment and tool cost about 90 yuan, per-customer gross profit about 309 yuan
- If CAC is compressed to within 180 yuan, gross profit/CAC ≈ 1.7, basically recovered upon the transaction completion
- If CAC reaches 350 yuan, one transaction cannot cover it; must rely on repurchase or higher unit price — and a large portion of quitting users "do not buy again once they have quit," so repurchase cannot be a default assumption
On the industry side, there are references: public materials roughly estimate content marketing/SEO customer acquisition at about 50–150 USD/customer, paid search often in the 200–500 USD/customer range; content is commonly said to be cheaper and generate more leads than traditional methods. These numbers come from overseas SaaS/general content marketing contexts and cannot be directly used as a price list for Chinese tobacco science, but they remind you of one thing: paid clicks are rising, compounding content is worth its own ledger. My approach: use my own gross profit to set the "affordable CAC ceiling"; industry ranges are only used to judge direction, not for self-comfort.
Personal view:
The easiest mistake in tobacco science is replacing "paying customer CAC" with "follower CAC." Adding a WeChat contact for 8 yuan looks great, but if only 1 out of 100 friends converts, the real CAC is 800. Since October 2023, my spreadsheet only keeps the "paying customer CAC" column; friend count is downgraded to a process indicator.
2. Channel Comparison: I Rank by "Intent x Compounding x Risk," Not by Platform Hype
The following is not a pros-and-cons list; it is my trade-offs after repeated testing from winter 2023 to spring 2025.
1. News Feed / Short Video Advertising: Can Drive Volume, But Also Easiest to Blow Through Margins
In March 2024, I tested two sets of content on the Massive Engine platform for science short videos about "body changes in the two weeks after quitting smoking," targeting ages 25–45 with interests in health/mother and baby, daily budget 300 yuan, running for 11 consecutive days.
- The first 4 days had acceptable click costs; the landing page was a long article with a WeChat QR code at the bottom
- Problem 1: Lots of onlookers, few who actually had plans to quit; the cost to add WeChat once reached 42 yuan/person, but the 7-day payment rate was under 2%, converted to a paying CAC over 2000 yuan
- Problem 2: Auditing and content compliance boundaries were tight; tobacco-related expressions had to be revised repeatedly, content iteration could not keep up with spend
- On day 8 I cut off the "general health curious" targeting and switched to behavior-based audiences closer to "searched for quitting/nicotine"; the cost to add WeChat rose to 55 yuan, but the 7-day payment rate reached about 6%, bringing paying CAC back to just over 900 yuan — still higher than the gross profit of a 399-yuan product, the advertising itself was still losing money
My conclusion:
News feed is suitable for validating copy and content, not suitable as the main staple when the customer unit price is below 800 and there is no strong repurchase. Emotional clicks in tobacco science are cheap, intent is expensive. Unless you are selling high-unit-price consulting or entourage-style long-term services, do not use news feed to force conversions.
2. Search Advertising: Expensive, But Intent is Clean
Around the same time I tested keywords like "what are the ways to quit smoking" and "how long does coughing last after quitting" on the search side. The cost per click was significantly higher than news feed, and some terms jumped to over 10 yuan per click. The advantage was that people entering the private domain had clear questions: many started with "I have smoked for 15 years, I would like to ask if I can join the program."
The paying CAC once reached about 260–320 yuan (depending on whether the landing page clearly stated the price and target audience). Still tight for a 399-yuan product, but manageable for a 1280-yuan one-on-one consultation.
My conclusion:
Search advertising is an accelerator for "products with profit margins," not a lifeline for "no product, only reads." If you do not have a gross profit >= 3 times the click cost, do not launch a search campaign.
3. Organic Search (SEO) and On-Site Long-Form Content: Slow, But the Only Compounding Asset I am Willing to Go Heavy On
From November 2023 to June 2024, I shifted my main effort from chasing trending short videos to "problem-keyword long-form articles + on-site columns." The topic selection avoided vague "smoking is harmful to health" topics and focused on specific decision scenarios, for example:
- How to deal with coughing, bad breath, and insomnia 1–4 weeks after quitting smoking respectively
- The applicable boundaries of nicotine replacement vs. complete cold turkey (only providing information, not making prohibited efficacy promises)
- Clearing secondhand smoke and communication in family settings (for spouses/children)
Operationally I fixed four steps:
- Keyword tiering: Curiosity keywords (why smoking yellows teeth) -> Problem keywords (which day of quitting is hardest) -> Solution keywords (how to make a 28-day quit plan) -> Transaction keywords (is quit accompaniment reliable). The first two layers build trust, the last two drive conversion.
- One pillar per cluster: Each pillar article 3000–5000 words, surrounded by 5–8 satellite articles interlinking with each other, avoiding dozens of isolated articles.
- Titles write "scenario + result/boundary", avoid clickbait. Search users want answers, not goosebumps.
- Review ranking and conversion every 90 days: update data, update Q&A, update "next steps" (add WeChat to get a checklist/book a 15-minute assessment), rather than just changing titles to trick clicks.
By August 2024, the paying customers coming through organic search-related paths had a comprehensive CAC (including writing time) of about 70–140 yuan. From publication to stable inquiries, a single article typically takes 8–16 weeks. Some say it is too slow — slow is a fact, but once you pass the inflection point, the marginal cost approaches "the time to update an old article."
It is often mentioned in public discussions that quality content can continue to generate leads for over 12 months after publication, completely different from paid traffic which "stops when you stop paying." My own spreadsheets support this: an article written in February 2024 about "the timeline of oral changes after quitting smoking" was still contributing WeChat additions by early 2025.
My conclusion:
If you plan to work in this vertical for more than a year, SEO is not a "do I have the energy" option — it is the only base that can structurally squeeze CAC down. Short videos can be the distribution spearhead, but a spearhead without a handle will cut your hand.
4. Zhihu / Public Account / Xiaohongshu: Platform Traffic Requires "Relocation Cost"
- Zhihu: Friendly for high-intent Q&A. In January 2024 I wrote a structured long answer under "most common reasons for quitting failure." After two weeks of accumulated upvotes and favorites, the homepage visits and the "get the withdrawal checklist" at the end of the article brought in a batch of good-quality friends. The problem is that external link and traffic diversion rules change frequently; compliance relocation is more draining than writing the answer itself.
- Public Account: Suitable for deepening trust before private domain and secondary contact with existing friends, but cold-start acquisition efficiency is average. I use it as a "deep page + text-based CRM for email/community," not as the main acquisition engine.
- Xiaohongshu: Covers and pain-point titles can easily drive volume; notes about "skin improvement after quitting" can get great reads, but the conversion path is long, and consultations are mixed with a large number of non-target people. I only use it for brand and agnostic reach, strictly controlling invested hours.
Personal view:
Platforms are about feel, not sentiment. Tobacco science has compliance red lines on every platform; what can be accumulated to your own reachable list (WeCom/email/owned site) counts as acquisition completion, otherwise you are just planting crops on someone else's land.
5. Existing Customer Referral: The Lowest CAC Tier, Provided You Actually Helped People Through Their Struggle
From May 2024 onward, I started sending "dual benefits" to users who completed the accompaniment and self-reported no relapse for 30 days: existing users could gift a friend a 15-minute assessment, and both sides would receive one bonus Q&A session. No cash rebates (to reduce disputes and incentive distortion).
The comprehensive CAC for referral customers has long-term 30–80 yuan (materials and time), with low communication costs. The share slowly grew from almost 0 to about 15%–20% of new customers. It amplifies delivery reputation, but cannot save a product with poor delivery.
3. Content SEO: How to Turn "Being Found" Into "Willing to Pay"
Many tobacco science accounts' SEO stops at "ranking" — ranking is only a half-finished product. I require every article entering the solution keyword/transaction keyword layer to have the following structure:
1. Within the First 200 Characters, Answer "Is This Written for Me?"
For example, make it clear: suitable for people who "smoke 10+ cigarettes a day, plan to take action within 30 days, and are willing to track with check-ins"; not suitable for people who "only want miracle cures and do not want to change trigger scenarios at all." Screening people will reduce read completion rates, but will increase the conversion rate after adding WeChat. I tried removing the screening paragraph, and while readership increased by about 20%, 7-day payment after adding WeChat dropped by nearly half.
2. Give Executable Steps in the Middle, Not Vague Scare Tactics
Scare-type headlines work short-term, but long-term they destroy trust and make truly motivated people close the page. I prefer to write: what is common on days 3–7 of withdrawal, which symptoms need medical attention, and a home environment modification checklist. Replace "fear" with "next step."
3. Only One Main CTA for Conversion
Early on I stuffed one article with: join a group, get an ebook, book a consultation, follow the video account. The result was that everyone clicked on nothing. Now each article pushes one main action — usually "add WeCom to get the '14-Day Withdrawal Record Sheet.'" The sheet itself is useful, not a bait PDF. The WeChat add message clearly states what you will receive within 48 hours, lowering the "another marketing account" defense.
4. Use Data to Guide Content, Not Gut Feel for Topic Selection
Monthly exports:
- Top 20 search terms from search/on-site search
- WeCom friend source notes (which article/which answer)
- The last 3 URLs read by paying customers before adding WeChat
In July 2024 I discovered that paying customers highly overlapped on two themes: "weight after quitting" and "spouse secondhand smoke communication." I immediately deprioritized the planned "tobacco history" series. History articles had reads, but almost no payments.
Personal view:
The decisive factor of SEO is not the red and green of keyword tools, but whether you dare to cut topics that have traffic but no intent. Ineffective topics in the tobacco field are especially plentiful: curiosity, celebrities smoking, pure scare statistics. They inflate vanity metrics, but cannot inflate the CAC that covers profit.
4. Private Domain Conversion: Public Domain Handles Reach, Private Domain Thins Out CAC
In the health private domain space, some have achieved 40%+ WeChat addition rates after contact, and others say the industry common addition conversion is around 25% — that is the result of optimized scripts and scenarios, not something you can replicate by dropping a QR code. My own path is more rustic, but stable.
The Funnel Looks Like This
- Public domain hook: Search long articles / Zhihu answers / short video pinned comment area "keyword to get form"
- Reason to add WeChat: Specific, immediate, low-commitment (record sheet, 8-question self-assessment, local medical resource guide, no efficacy promises)
- Auto welcome message + 3-sequence (day 0/1/3): Self-assessment interpretation -> Common failure points -> Whether structured accompaniment is needed
- Tags: Smoking history, daily count, past failure count, trigger scenarios (after meals/drinking/overtime), whether there is family secondhand smoke pressure
- Conversion node: 2 fixed live Q&A sessions or text Q&A per week; send "plan comparison" (self-help materials / check-in accompaniment / one-on-one) to high-intent tags, state prices clearly once
The Pits I Have Stepped In
- Pit: Put people in a 500-member group and ignore them. The group quickly becomes an ad graveyard. Changed to 30–50 person small-cycle camps, archived upon completion — completion rates and reputation both improved.
- Pit: Hard sell immediately after adding WeChat. In late 2023 I tried "same-day pricing," and the same-day block rate rose significantly. Now I default to at least completing the assessment interpretation before discussing products — payment is a bit slower, but complaints and refunds came down.
- Pit: Private domain only has push notifications, no delivery feel. Even if a user does not buy, 7 consecutive days of useful short content will reduce blocking and increase future referral probability.
A Set of Process Indicators That Can Be Used for Calibration (My Healthy Range)
| Stage | My "Healthy" Rough Range | What to Check First If It Drops |
|---|---|---|
| End of article to adding WeChat | Depends on traffic quality, search traffic often better than news feed | Whether CTA is unique and consistent with article promise |
| Valid reply within 7 days after adding WeChat | >35% | Whether welcome message looks like a bulk ad |
| Valid conversation to payment | When product and price match, 8%–15% has appeared | Whether wrong unit price is sold, whether screening is insufficient |
| Existing customer referral share | Target >15% | Whether delivery and camp closing ceremony exist |
The meaning of private domain is not "harass them once more," but spreading the one-time content cost over a higher conversion rate. The same SEO article, with conversion from 1% to 2%, halves the CAC — often more cost-effective than writing ten more homogeneous articles.
5. My Current Resource Allocation (Clear Positioning)
If I can only allocate 100% energy/budget, my default configuration at unit price 300–1500 and team <=3 people is:
| Area | Share | Reason |
|---|---|---|
| Search-oriented long articles + old article updates | 45% | Structurally reduce CAC, can accumulate |
| Private domain sequences, tags, small camp delivery | 25% | Key lever to turn traffic into profit |
| Deep distribution on 1–2 platforms (Zhihu or short video, choose strength) | 15% | Supplement new customers, do not covet many platforms |
| Small paid tests (search terms or news feed content) | 10% | Only to learn intent and copy, not for volume |
| Referral and camp closing operations | 5% | Lowest CAC source, needs mechanism not luck |
Things I Clearly Oppose:
- Talking about "scaling up advertising" without calculating per-customer gross profit
- Using scare tactics to stack plays, then pretending to be professional in private domain — trust debt will come back with interest
- Running five platform matrices simultaneously without a unified post-WeChat path
- Making "harm reduction product recommendation" the only conversion point without compliance and evidence boundaries (both crosses the line and turns the account into a short-term harvester)
6. Four Working Principles to Squeeze CAC Into the Profit Line
Principle 1: Set affordable CAC first, then choose channels.
Derive from gross profit, not forward from platform average prices. A 399-yuan product should not fantasize about news feed breaking through head-on.
Principle 2: Intent > Coverage.
Do less "all smokers" and more "smokers preparing to take action" and their family members. Family keywords (secondhand smoke, persuasion scripts) sometimes convert better than smoker keywords.
Principle 3: Content assetization, advertising experimentation.
80% effort goes into accumulable articles and script libraries; advertising budget is tuition — only counts as well spent if you have learning notes.
Principle 4: Conversion rate is a cheaper cost-reduction tool than traffic.
A 20% improvement at each funnel level often improves CAC more than doubling traffic. This is especially true in tobacco science — users are heavy deciders, trust chains are long; whoever shortens the chain has lower costs.
7-Day Action Checklist — Executable Immediately
Day 1
List all your monetizable products: unit price, variable cost, per-customer gross profit, affordable CAC (suggest <= gross profit / 1.5). If you have no product, first set a low-fulfillment-cost "assessment + resource pack" or small camp, otherwise all subsequent numbers will spin in the air.
Day 2
Export the last 30 days' friend sources and conversion sources, calculate the real paying CAC. If you cannot calculate it clearly, from today add remark tags to each WeChat addition channel.
Day 3
Cut or pause advertising plans whose paying CAC is more than 1.5 times the affordable line; keep one small-budget group only for testing copy.
Day 4
Build a keyword list: 10 problem keywords, 5 solution keywords, 3 transaction keywords. Plan 1 pillar article outline for each solution keyword.
Day 5
Unify all content in advertising or promotion to a single CTA, and rewrite the WeChat add welcome message (commit in writing what deliverable will be provided within 48 hours).
Day 6
Tag WeCom friends with 4 basic tags (smoking history, daily count, failure count, trigger scenarios), maintain only one active small group within 30 people or a waiting list for the next camp opening.
Day 7
Review: which URL or answer contributed the most to WeChat additions and conversions? Adjust the next two weeks' writing schedule to "doubling down on high-conversion topics + updating old pillar articles," rather than chasing another hot topic.
Tobacco science does not lack "correct words"; what it lacks is turning correct words into a cost-controllable acquisition system. Once you accept that "readership is not revenue," channel selection suddenly becomes much simpler: keep what accumulates intent traffic, keep what increases private domain conversion — the rest, no matter how popular — is just burning money and enthusiasm for the platform's algorithm.
Squeezing CAC down to where profit can cover it is not about being stingy. It is about respecting every person who takes the time to read your article seriously: do not feed their attention to inefficient channels, and do not gamble your cash flow on clicks without compounding.