Why Doing Too Many Tobacco Topics at the Same Time Significantly Reduces Overall Profits

I've seen a very typical content operation state: a team starts with only two or three people, with very focused topic selection, everyone studies the same type of user problems every day, and articles, short videos, and data reviews all revolve around a narrow direction. A few months later, the team expands to five or six people, and the person in charge starts thinking "we can't just do one direction," so they add tobacco history, industry observation, consumer culture, health knowledge, smoking cessation services, cigar culture, oral health, industry policy, equipment reviews, and a whole bunch of other columns.

On the surface, this is business expansion.

In reality, many teams begin their profit decline phase from here.

The problem isn't that there's more content, but rather that content starts competing for the same limited production resources without forming strong enough revenue synergy.

My judgment has always been quite direct:

The most dangerous state for a content team is not having no direction, but having too many directions, each of which seems reasonable.

Because "seems reasonable" and "makes money" are two different things.

Especially for tobacco-related content, internet advertising itself has clear regulatory boundaries. The "Internet Advertising Management Measures" clearly stipulate that it is prohibited to use the internet to publish tobacco (including e-cigarette) advertisements.

Therefore, when creating tobacco-themed content, you cannot simply equate "increasing the number of topics" with "increasing business opportunities."

What should really be managed is legal and compliant information content capabilities, user relationships, data assets, and adjacent service opportunities.

I. The Most Misleading Metric: Increased Content Volume

Assume a 5-person content team initially only works on three topics.

Monthly production:

• 30 in-depth articles;
• 90 short-form content pieces;
• 300 user question compilations;
• 500 search keywords;
• Approximately 800 user feedback items.

Because the topics are relatively focused, the team gradually develops a very familiar working method.

Editors know what users care about most.

Writers know what angles generate reads.

Data analysts know which keywords are worth continuing to pursue.

The person in charge can even tell from the comments section what should be done next week.

Six months later, what this team truly owns is no longer "120 articles," but a set of content assets.

Then the person in charge makes a decision:

Add 7 more topics.

The original 3 topics become 10.

Looks very reasonable.

But the 5 people's time hasn't increased.

So where each topic could previously invest 40 units of production resources per month, now each can only average 12-15 units.

The first problem to appear isn't declining traffic.

It's declining familiarity.

Writers start frequently looking up information.

Editors need to re-understand users.

Data analysts re-establish keyword classifications.

Designers redo visual templates.

The person in charge starts attending more topic selection meetings.

A piece of content that originally took 4 hours to complete might now take 6 hours.

A topic that could be judged in 20 minutes might now take 1 hour.

This is a type of cost that is very easily missed by financial statements in content operations:

Complexity cost.

It doesn't appear separately on payroll, but it constantly erodes profits.

II. After Increasing Topics, What Really Increases is "Switching Cost"

I quite dislike a certain content team management approach:

"This direction has opportunities recently, let's do it too."

Then a quarter later:

"That direction is also trending, let's arrange another column."

In the end, the whole team has over a dozen columns, but none is truly done in depth.

The problem is that people are not servers.

Studying tobacco industry policy today, researching smoking cessation services in the afternoon, studying cigar culture in the evening, and doing oral health the next day.

They all look "tobacco-related."

But the user problems behind them are not exactly the same.

Why do users look at industry policy?

Perhaps to judge industry trends.

Why look at smoking cessation content?

Perhaps to reduce dependence.

Why look at cigar culture?

Perhaps out of interest and cultural consumption.

Why look at oral health?

Perhaps due to health anxiety.

Although these users may all be interested in the broad term "tobacco," their motivations and needs are completely different.

If an account jumps back and forth between these needs every day, it becomes difficult for platform algorithms, search systems, and even users themselves to form a stable perception.

The most common situation at this point is:

Traffic doesn't significantly decline.

But user quality starts declining.

More troublesome is that after commercialization efficiency declines, the team easily misjudges the cause.

The person in charge will say:

"Traffic hasn't been good lately."

So they continue adding topics.

But the real problem might be:

Users fundamentally don't know what problem you're really good at solving.

III. An Anonymous Case: From 10 Topics Cut to 4 Topics

The following case is a case review based on common content team operating structures and does not correspond to any specific company.

In spring 2025, a 5-person content team based in Hangzhou decided to expand their tobacco-related content business.

The team originally mainly worked on 3 directions:

1. Industry information;
2. Tobacco culture;
3. Health-related knowledge.

The person in charge believed there were too many market opportunities at the time, so they added 7 new directions.

Three months later, the team increased from about 120 pieces of content per month to 210.

The surface data looked very impressive:

Monthly total reading volume increased by approximately 31%.

The person in charge was very excited.

But the finance person discovered something uncomfortable:

Monthly revenue only increased by approximately 8%.

At the same time, content staff overtime increased significantly.

The original monthly content production cost was about 68,000 yuan; three months later it approached 94,000 yuan.

Adding data tools, outsourced design, review, and management costs, the overall operating profit rate dropped from approximately 27% to around 16%.

At this point, the person in charge's initial judgment was:

"Although traffic growth isn't fast enough, the direction is right. Let's persist for another three months."

I believe this is the most dangerous step for many teams.

Because persisting at this point means continuing to invest resources in a model that has already proven to have declining efficiency.

The team then recalculated the 10 topics based on four indicators:

Effective user count, per-article production cost, user return rate, and subsequent service relevance.

The results were very clear.

Among them, 3 topics had high traffic but low user return rates, and content production costs were significantly above average.

Another 2 topics had many comments but hardly formed any subsequent user relationships.

In the end, only 4 topics were kept:

• One carries core professional recognition;
• One carries stable search traffic;
• One carries health education and adjacent needs;
• One carries small-scale experimentation.

The other 6 topics were all suspended.

Not deleted.

But suspended.

Because business decisions fear two extremes the most:

One is "do everything."

The other is "cut everything."

The truly effective method is to layer resources.

IV. The Most Important Thing I Believe: Traffic Cannot Directly Equal Profit

Many content teams have a very simple financial model:

Reading volume × commercial value = revenue.

Reality is far from this simple.

It should at least be broken down into:

Effective user count × user value × sustainable operation period − content costs − management costs − risk costs

When calculated this way, many so-called "high-traffic topics" immediately reveal problems.

For example:

Topic A brings 1 million reads in a month.

Topic B only brings 400,000 reads.

Topic A requires 30 person-days.

Topic B only requires 10 person-days.

Topic A users on average read only once.

Topic B users read 4-5 pieces of content consecutively.

If you only look at reading volume, A obviously looks better.

If you look at profit, B might be the topic truly worth investing in.

This is also why I don't recommend content managers use "number of viral hits" as a core business indicator.

A viral hit only proves that one particular content piece succeeded.

It cannot prove that this topic is worth long-term operation.

V. The More Topics, The Harder It Is for Content Assets to Form Compound Interest

A truly valuable content team should not start from zero every month.

It should get faster and faster.

In the first year, writing one piece of content takes 6 hours.

In the second year, it might only take 4 hours.

In the third year, even 2-3 hours to complete.

Why?

Because past work continuously accumulates:

• Keyword library;
• User question library;
• Case library;
• Title library;
• Data library;
• Expert resource library;
• Content templates;
• Comment feedback;
• Search demand;
• User profiles.

These are the real assets of content operations.

But if the team keeps switching topics, this compound interest gets interrupted.

For example, a writer has already written 100 industry observation articles in a row.

They know:

What problems users truly care about;

What data needs verification;

What expressions easily cause misunderstanding;

Which materials are worth long-term tracking.

Suddenly asking them to do a completely different health service feature means temporarily freezing part of the experience they've accumulated.

So I've always believed:

The biggest waste for a content team is not producing one less article, but frequently changing tracks for someone who has already developed professional competence.

VI. Focusing Doesn't Mean "Only Doing One Topic"

A misconception needs to be corrected here.

Focusing doesn't mean:

I only do one column.

This understanding is too mechanical.

True focusing should be:

Around the same core group of users, solving a set of related problems.

For example, if a team's core users are adult users interested in tobacco industry information, health impacts, and smoking cessation-related services, they can form:

Core topic

Industry and professional information.

Traffic topic

Stable search-type questions.

Relationship topic

Health knowledge and behavior change content that users care about long-term.

Experimental topic

Small-scale validation of adjacent service directions.

These four types of topics don't conflict.

The real question is:

Do they serve the same core user?

If users highly overlap, topics can form traffic transfer between each other.

If users are completely different, it creates "internal competition" in content operations.

VII. How I Would Judge Whether a Topic Should Stay

If I took over a content team that was already quite messy, I wouldn't immediately ask:

"Which topic has the highest traffic?"

I would first create a topic profit statement.

Each topic would be evaluated on at least 10 indicators:

Then score each item.

I especially value three things:

User overlap, production efficiency, long-term accumulation capability.

Because these three indicators determine whether a topic can generate compound interest.

If a topic has high traffic but requires re-research every time, re-finding materials, re-building the content system, and the users don't have much overlap with core users, I'd rather do less.

VIII. What Should Be Cut Most Isn't Necessarily the Worst Topic

This is a mistake operators are particularly prone to making.

Many people only look at one thing when cutting topics:

"This column has the lowest traffic."

Actually, that's not necessarily true.

I prefer to cut three types of topics.

Type One: Decent Traffic but Completely Unable to Retain Users

This type of content very easily creates illusions for the person in charge.

The data dashboard looks great.

But users come and leave.

Next month, they still need to buy traffic again and produce content again.

This kind of traffic hasn't formed assets.

Type Two: Content Cost Significantly Higher Than Other Topics

For example, producing the same 10 pieces of content:

Topic A requires 8 person-days.

Topic B only requires 3 person-days.

If the final effective user value generated by both is similar, then A is actually a profit black hole.

Type Three: No Synergy with Core Business

This is the type I'm most willing to cut.

A topic itself might not be bad.

It might even have traffic.

But if the users it attracts are completely different from your core users, it will continuously consume team resources.

Not every good opportunity is your opportunity.

This is especially important for small teams.

IX. Giving the Team a "Resource Ratio" Is More Effective Than Shouting "Focus"

If a team only has 5 people, I usually don't allow 10 topics to evenly distribute resources.

A more practical approach is:

60% of resources to core topics.

Including lead writers, main data analysis, and most in-depth content.

20% to profit-related adjacent topics.

The focus is not on directly promoting tobacco products, but on building long-term value around legal and compliant professional information, health education, service-oriented content, etc.

10% to traffic experiments.

Testing new search demands, user problems, and content formats.

10% reserved for truly new opportunities.

The significance of this ratio isn't the 60%, 20%, 10%, 10% themselves.

But rather:

No new topic can indefinitely occupy core resources.

A new direction wanting to upgrade from 10% to 30% must prove itself.

Otherwise, it can only ever be an experiment.

This shifts the team from "feeling this direction is good" to "data proves this direction is worth increasing resources."

X. I Recommend a 30-Day "Topic Subtraction"

If the team already has 10-15 topics, I wouldn't recommend an immediate large-scale restructuring.

First do a 30-day experiment.

Days 1-7: Statistics

Pull out all topics from the past 90 days.

Statistics:

• Content volume;
• Reading volume;
• Effective users;
• Return visits;
• Saves;
• Comments;
• Content production time;
• Outsourcing costs;
• Management time.

Don't rely on feelings.

Days 8-14: Layering

Divide topics into:

A: Core

Must continue investing.

B: Synergy

Can continue developing around the core topic.

C: Experiment

Low-resource validation.

D: Stop

Temporarily stop production.

Days 15-21: Resource Reallocation

Stop D category.

Reduce C category.

Put all the saved personnel time and budget into A category.

This step is critical.

If you just "cut content," the team won't feel profit improvement.

The released resources must be reconcentrated.

Days 22-30: Compare Results

Compare before and after adjustments:

• Per-article cost;
• Effective user cost;
• User return visits;
• Content production efficiency;
• Core topic growth;
• Team overtime hours;
• Manager meeting time.

Finally, look at one indicator:

Has the profit contribution per unit of resource improved?

If it has improved, focusing is effective.

If not, re-examine whether the core topic was chosen correctly.

XI. Another Often Overlooked Cost: The Boss's Own Time

Small teams especially easily overlook this item.

Assume the person in charge spends daily:

• 30 minutes reviewing topics;
• 30 minutes handling cross-team coordination;
• 1 hour looking at data;
• 30 minutes resolving conflicts between columns.

2.5 hours per day.

Over 22 working days a month, that's 55 hours.

If topics increase from 4 to 12, the person in charge might not simply increase work by 50%.

Because more coordination relationships arise between topics.

With 4 topics, the combination relationships are relatively limited.

With 12 topics, communication, scheduling, resource contention, data interpretation, and personnel arrangement significantly increase.

So the truly frightening thing about multi-topic operations is:

Complexity doesn't grow linearly.

This is also why a 10-person team isn't necessarily more profitable than a 5-person team.

After headcount increases, if the management system hasn't been upgraded synchronously, what increases might just be coordination costs.

XII. What's Truly Worth Pursuing Isn't "Topic Count," But "Synergy Between Topics"

I imagine a mature content team as a tree.

There's only one trunk.

There can be many branches.

The trunk is the core user and core capability.

The branches are the different content topics.

Without a trunk, a dozen topics are just a dozen small grass plants.

Each needs its own watering.

None goes deep enough.

If there's a clear trunk, adjacent topics can share:

• Users;
• Data;
• Writers;
• Search terms;
• Content materials;
• Professional knowledge;
• Distribution channels;
• User feedback.

At this point, adding one topic can truly increase profits.

Otherwise, what increases is often just workload.

XIII. The Very Simple Question I Would Finally Give Operators

If a new topic comes in, I wouldn't first ask:

"Does this direction have a market?"

Because most directions can find a market.

I would ask:

"Why must this topic be done by us?"

Then ask three more questions:

Is it our core user?

Can it reuse our existing capabilities?

The longer we do it, the more profitable it becomes?

If all three answers are "yes," it's worth continuing.

If only the first answer is "yes," it needs observation.

If the only reason is "the market is big," I usually won't invest too many resources.

Because a big market doesn't mean you can make money.

Conclusion: Doing Less Doesn't Equal Giving Up Opportunities

The biggest psychological barrier for content operators is the fear of missing opportunities.

Seeing a new direction, they think:

"If we don't do it now, what if others do it later?"

But after truly operating a content business, you discover another type of loss is more serious:

You try to grab everything, and in the end, no direction forms a true competitive advantage.

Focusing doesn't really solve the "what to do" problem; it solves:

With limited people, money, time, and attention, where should compound interest be formed.

Especially for tobacco-related internet content, commercialization itself is subject to clear advertising regulatory constraints, so the simple logic of "more topics = more monetization opportunities" cannot be adopted. Current rules clearly prohibit internet tobacco advertising, including e-cigarette advertising; local regulatory authorities also continuously emphasize the compliance boundaries of internet tobacco-related operations and advertising activities.

Therefore, I tend to understand tobacco-related content operations as a constrained content asset management problem.

The core isn't to spread topics wider and wider.

But to find a stable core user, and then around this user, organize professional content, search demand, health education, adjacent services, and experimental directions.

Truly profitable teams often don't know the most, but know what not to do.

When a 5-person team can very clearly state:

"Among these 10 directions, we only concentrate resources on 4, and temporarily don't do the other 6."

This isn't conservatism.

It shows they've started transforming from "content producers" into true operators.

📊 Key Data Overview

Margin vs After

Resource Allocation