It was not until my fourth complete low season creating smoking cessation and nasal inhaler content that I truly broke a bad habit: whenever I saw public domain readership cut in half, I would rush to post daily, chase hot topics, and increase ad spend. The result was always the same: more competition, higher costs, lower conversions.

From July to August 2023, the average daily reads of my "Smoking & Oral Health / Respiratory / Withdrawal Reactions" series dropped from about 12,000 in peak season to around 3,500; daily follower additions from external links dropped from 40 to 12. The mistake I made that year was typical: I wrote 18 new articles in two weeks, 11 of which had fewer than 2,000 reads and nearly zero conversion. On the night of August 12 that year, I opened my backend and sorted by "cumulative reads over past 12 months x follower addition rate x whether product links can be attached," and I discovered: what truly continues to generate orders in the low season is not new hot topics, but old articles that already proved their conversion but are gathering dust in the archive, plus inactive existing users in private domain, plus combo packages more stable than single products.

My clear view: In the low season for tobacco-related content, profit does not depend on "waiting for traffic to warm up," but on "repricing existing assets." Public domain has strong seasonality for customer acquisition, while private domain and combo sales turn the low season into a window for system repair, repeat purchases, and raising average order value. The three paths below are all practices I executed week by week and adjusted — not a conceptual list.

1. First Recognize: Where Exactly Is the Low Season "Low"

Using the content types I track as examples (long educational articles, short video narration, community Q&A):

What is low is unfamiliar traffic and impulsive clicks, not the smoking cessation demand itself. People who want to quit do not stop struggling just because it is July or August. What changes: platform recommendations cool down, competing content increases, users become less willing to click.

There are three most loss-making strategies in the low season, all of which I have tried:

1. Fighting declining reads with more new articles — output doubles, marginal conversion approaches zero.
2. Cutting prices in public domain — the price anchor is broken, making it harder to raise in peak season.
3. Sending only ads in private domain without updates — group message open rate drops from 40% to below 15% in a week.

The correct priority should be reversed: Public domain does refresh and precision targeting (or even less spending); private domain does activation and accompaniment; the conversion layer does combos rather than single SKU hard push.

---

2. Old Article Refresh: Turning Inventory Content Into a "Second Spring Production Line"

2.1 First Build a "Refreshable Asset Table," Not by Gut Feeling

In August 2023, I screened 27 out of about 140 historical articles into the refresh pool, with only four criteria (at least three must be met to enter):

• Cumulative reads >= 20,000, or once entered recommendation in a single day
• Clear action at the end or in comments (follow, request checklist, ask about product)
• Core thesis still valid within 12 months (physiological mechanisms, withdrawal timeline, care steps prioritized)
• Title or cover looks "dated" under current platform style (font, tone, information structure outdated)

Evergreen content (how-to guides, timelines, checklists, mechanism explanations) is naturally suitable for reuse. The industry has always emphasized that evergreen depends on regular updates rather than writing once and leaving it. My experience: in the low season, refreshing only 2-3 articles per week is more cost-effective than starting 5 new topics.

2.2 Refresh Is Not Changing the Title, It Is Changing the "Decision Path"

I decompose an old article into five layers, at least three must be changed:

Layer	What to Change	Example
Hook	Title, first 80 chars, cover info points	"When does bad breath disappear after quitting" -> "My breath changes recorded at Day 7/14/30 and how I handled them"
Evidence	Add 1 timestamped case, 1 number comparison	Specify: March 2024, 86 people checked in the community, Day 14 most frequent feedback was...
Structure	Add subheadings, step numbering, comparison table	Convert prose into "Phenomenon -> Cause -> 3 Steps You Can Do Today"
Action	Single CTA at the end	Keep only one: get the "21-Day Oral/Respiratory Withdrawal Comparison Table"
Distribution	Slice into short videos / community newsletter	Long article split into 3 45-second voice clips + 1 community poll

In January 2024, I refreshed "Oral Changes Timeline After Quitting Smoking" (original draft November 2022). Specific actions:

• 2 hours to rewrite the opening, delete 400 words of emotional lead-in, replace with a table "Days 1-3 / Day 7 / Day 30"
• Add 3 anonymous feedback entries from 2023 community (keep only days and symptoms, no medical promises)
• Change end-of-article CTA from "Learn about products" to "Reply with keyword to get comparison table"
• Published to public domain at 20:30 that day, sent the table image to 3 old groups at 9:00 next day

Result: about 9,100 reads in 72 hours (original peak was 16,000, but 9k in low season exceeded expectations); 67 follower additions; 19 of them bought the Starter Combo within 7 days, with average order value about 38% higher than single consumables. The same week, a hot topic commentary I newly wrote had 2,400 reads, 4 follower additions, 0 conversion.

Personal assessment: the ROI of writing new articles in the low season often cannot beat "refreshing one old article with proven conversion." New articles solve topic coverage; refresh solves cash flow.

2.3 Refresh Production Line (My Fixed 90-Minute Template)

1. 0-15 min: Open old article backend, note down historical highlight paragraphs, negative comments, sections with high bounce rate
2. 15-45 min: Rewrite hook + add one table + delete outdated data
3. 45-60 min: Keep only one CTA, check for exaggerated efficacy claims

4. 60-75 min: Create 3 short video script outlines (each with one number or one step)

5. 75-90 min: Schedule: public domain publish date / private domain secondary touch date / combo pack hook sentence

Pitfall: Publishing 3 refreshed articles on the same day — the platform treated it as a "copycat account," recommendations went silent. Later changed to "1 article every other day + only 1 short video on that day," much more stable.

---

3. Private Domain Activation: When Public Domain Cools, the Profit Stage Is in WeChat

Private domain is not just "having a group." The common industry framework is: acquisition -> retention -> activation -> conversion. In the low season I only focus on three things: recall, give tasks, give combo reasons.

3.1 Segment, Not Send the Same Ad to Everyone

I roughly divide the private domain into four layers (numbers approximate from a June 2024 export):

• A Purchased Users ~ 420 people: bought tools/kits
• B Strong Intent ~ 380 people: received materials, asked about price but did not buy
• C Silent Followers ~ 1,100 people: added WeChat over 30 days, no interaction
• D Group Only, No Private Chat ~ 600 people

The low season mistake is blasting "limited-time discount" to all ABCD. In August 2023 I tried 10% off for everyone. Result: Layer A felt offended ("early buyers lost out"), Layer B kept waiting, Layer C directly complained about spam. That week: 23 complaints/friend deletions, only 6 more conversions, net effect negative.

Later changed to:

Layer	Contact Method	Content	Frequency (Low Season)
A	1-on-1 or small group	Usage review, next stage combo suggestion	Every 10-14 days
B	Tagged mass send	Send a "Decision Comparison Table" + limited Q&A slots	Once every 7 days
C	Value recall	3 consecutive value messages before mentioning products	14-day wake-up sequence
D	In-group tasks	Check-in, Q&A relay, do not push conversion first	2 value msgs/week, 1 light conversion

3.2 A Tested Effective "7-Day Silent User Wake-Up" Sequence

Target: Layer C, added WeChat >= 45 days, never replied.

July 8-14, 2024, I did a small sample of 300 people (not full database blast):

• D1: Send only one sentence: "If you're still here, reply 1, I'll send you the 'First Week Withdrawal Body Signal Comparison'." — Reply rate about 18%
• D2 (only repliers): Send PDF + ask "What day/symptom are you stuck on?"
• D3-D4: Give 1 actionable action based on reply (hydration, oral care, substitution behavior), do not mention price
• D5: Send a link to a refreshed article + "Table 3 in the article might help you"
• D6-D7: For those still interacting, provide Starter Combo explanation + 2 real usage boundaries (who it is not suitable for, cannot replace medical treatment)

Out of 300 people, about 54 had effective conversations, 11 conversions, profit was higher than my same-week public domain new customers because ad spend was nearly zero and customer service time was concentrated on people who actually replied.

Key discipline: No hard ads for the first 4 days. I tried inserting a link on D1, reply rate dropped below 6% and the whole sequence was ruined.

3.3 Community Activation: Replace "Notifications" with "Tasks"

The low season group's biggest fear is becoming an ad board. My current weekly rhythm (example: about 180-person withdrawal support group):

• Monday: One quantifiable small goal for the week (e.g., "record craving frequency")
• Wednesday: Anonymous Q&A 30 minutes (I set a timer, stop when time is up)
• Friday: Old article slice + one table
• Sunday: Sync "Combo Pack This Week Q&A Slots" only to A/B tags

In May 2024, after changing from "posting products every day" to the above rhythm, group message opens (approximate reads via WeCom stats) went from below 20% to about 35%. In that month's private domain conversions, 41% of orders came from people who first participated in Wednesday Q&A. Accompaniment creates interpretation rights, not purchase pressure.

---

4. Product Combo Sales: Raise Order Value Through Structure in Low Season, Don't Fixate on Discounts

Single-product thinking can still rely on volume when traffic is high. When volume is insufficient in low season, you must make each order more worth closing and more worth serving.

4.1 How I Changed from "Selling One Device" to "Selling a Stage Plan"

Early on I only pushed the nasal inhaler as a single product. In low season the script easily became a price negotiation. Users asked: "Can you make it cheaper?" You had no card to play except lowering the price.

From November 2023, I fixed a three-layer combo (names can be adjusted for compliance, logic is universal):

1. Awareness Pack (Low price/Free)
2. 21-day comparison table, withdrawal reaction handbook, oral/respiratory care checklist
3. Purpose: complete screening, no profit or thin margin
1. Starter Tool Kit (Main push)
2. Main tool + 2-week consumable rhythm explanation + 1 15-minute usage review (voice/text)
3. Price anchor: more convenient than "single tool + after-sales private chat troubleshooting"
1. Consolidation Pack (Advanced)
2. Starter pack content + 4-week check-in slots + relapse scenario plans (social events, overtime, emotional fluctuations)
3. Sold only to those who completed the starter or have clear high-intensity needs

The low season main push is always 2, not 3. I tried pushing the high-unit-price consolidation pack in low season, hesitation lengthened, conversion cycle dragged from 3 days to 2 weeks, cash flow became more difficult.

4.2 A Real Comparison (Same Customer Service, Same Week Traffic Structure)

Third week of August 2024, public domain added about 90 followers (quite light):

Strategy	Script Focus	Conversions	Avg Order Value	Returns/Disputes
Only push single product	Features and price	7	Baseline 1.0	2 "don't know how to use" complaints
Push starter combo	14-day rhythm + review	9	~ 1.35	0 serious disputes
Force consolidation pack	Long-term accompaniment	3	~ 1.9	1 "too expensive, didn't finish"

Solid conclusion: the optimal solution in low season is often "medium order value + high sense of completion," not the highest order value. The value of a combo is not filling the box, but reducing usage failure rate, thereby reducing after-sales friction and increasing referrals.

4.3 Combo Scripts Only Bind to "Stage," Not to Fear

Phrases I prohibit my team from using: "If you don't quit now..." "Must buy the full set to be effective."

Allowed phrases:

• "What week are you on? Which line on the comparison table bothers you the most?"
• "The single tool solves 'hand/nasal inhalation action replacement,' the comparison table solves 'do you know whether today is normal or not.'"
• "The starter pack is for people still struggling; the consolidation pack is for people who have stabilized but fear relapse at social events."

In February 2024 there was a negative incident: an assistant, to boost performance, blasted "last 6 hours" to 50 Layer B people. 5 orders came that day, but 3 were returned within a week, and two old customers questioned credibility in the group. In low season, credibility costs more than volume, this lesson only needs to be learned once.

4.4 How Combo, Old Articles, and Private Domain Fit Together

A complete chain looks like this:

1. Public domain refreshed article -> only direct to "comparison table"
2. Private domain D1-D4 only solve one specific discomfort
3. D5 send the table from the refreshed article, establish professional credibility

4. For those still interacting, recommend starter combo by stage

5. Proactive review nodes at Day+3, Day+7, Day+14 after conversion (reduce returns, sow consolidation pack seeds)

Low season profit is often hidden in Day 14 of Step 5, not in the read count of Step 1.

---

5. Low Season Weekly Battle Table (Ready to Copy)

Taking a solo or ultra-small team as example, I planned about 20 hours per week for content + private domain:

Time Slot	Action	Output Goal
Monday morning	Select 2 old articles into refresh pool, write asset table	Topics locked
Monday afternoon	Complete 1 refresh + CTA	1 long article
Tuesday	Slice into 3 short videos / graphic clips	Distribution materials
Wednesday	Community Q&A 30 min + Layer B comparison table	Activation
Thursday	Layer C wake-up sequence progress (only handle repliers)	Conversations
Friday	2nd refresh or deep review notes	Content
Saturday	Layer A usage follow-up 10-15 people	Repurchase / referrals
Sunday	Check numbers: reads, follows, conversations, combo ratio	Parameter adjustment

Resource allocation I recommend: Content 40% (nearly all refresh and slicing) / Private domain 45% / Product and script refinement 15%.

The least valuable time spent in low season is "new hot topics just for posting's sake."

---

6. Four Principles I Repeatedly Remind Myself

1. Low season is the season to repair the system, not to prove you can still post daily.
2. Old article refresh is about "redoing the decision path," not changing to a shocking title.
3. Private domain activation is about segmentation and restraint; those who do not mention money in the first half deserve to talk about combos in the second half.

4. Combo selling is about stage matching and sense of completion; unit price increase is a result, not a script goal.

If you are now in a month where reads are cut in half and follower additions are slowing, do not first ask "what viral piece to write next." First open your backend, sort your old articles by conversion rate and take the top 20. Export your private domain, mark those who bought or asked about price within 90 days. Organize your offerings into three layers: Awareness Pack / Starter Kit / Consolidation Pack. Once these three steps are done, your low season profit source will change from "recommendation traffic left to luck" to "operational actions you can review weekly."

In my next low season, I will still write my KPIs as three numbers: refreshed article conversion share, private domain wake-up 7-day conversion, starter combo share of total orders. Public domain popularity is only a reference, no longer an emotional switch. People who actively create profit are usually not the ones with the highest traffic, but those who best deploy their existing assets when traffic declines.

8k–15k → 2k–5k
Avg public domain reads per article (peak → low)
2.5%–4% → 1%–1.8%
End-of-article follower add rate change
8%–12% → 3%–6%
First-time buyer conversion rate change
35% → 55%–70%
Repeat customer contribution attainable range

Wrong Low-Season Practices

Fighting read decline with more new articles, slashing prices in public domain, only ads in private domain

Correct Practices

Public domain refresh & precision, private domain activation & tasks, combo layer conversion

Note: “Low season” here refers to the periodic decline of public domain traffic in the content industry; division may vary by platform and topic.