Private Traffic Operations in Practice: Building a Repeat Purchase Engine in 3 Months, with the SOP
Summary: Acquisition cost in public channels keeps rising, so retaining the users you already paid for is the cheapest growth left. This playbook gives you the cost data, three myths to drop, a five-step private-domain SOP and one anonymized case that lifted repeat purchase by double digits in 90 days.
Introduction
Acquisition cost in public channels keeps climbing, and with it the long-term value of the users you already own. According to iResearch's white paper on cross-border ecommerce private domain operations in China for 2024, compiled by 10100.com, leading brands see repeat purchase rates among private-domain users 2.3 times higher than in public channels. CIC data puts China's private-domain operations solutions market at RMB 11,900,000,000 in 2024, with RMB 25,800,000,000 projected for 2029 (Tencent News).
This article is written for ecommerce, beauty, mother-and-baby and B2B categories where repurchase happens often. It delivers a playbook you can run: first clear away three common myths, then use a five-step SOP to take a private-domain pool from 0 to 1, and finally check it against an anonymized case where the repeat purchase rate rose 200% in 90 days. Read it and act on it the same day.
1. Why Private Domain Is Not Optional: Hard Numbers on Cost and Repurchase
Public traffic belongs to platforms that control distribution and pricing, so reaching users means paying for access again and again. Private domain is a user base the company controls, reachable repeatedly at very low cost.
| Dimension | Public traffic | Private traffic | Difference |
|---|---|---|---|
| Acquisition cost | 150–200 yuan per user, higher in some categories | 60–80 yuan per user | About 60% lower |
| Repeat purchase rate | 8–12% | 20–35% | 2 to 3 times higher |
| Average order value | Baseline | +30–50% | Significant lift |
| User lifetime value | Baseline | +150% | More than doubled |
| Reach cost | Pay every time | Almost free | Fundamental difference |
| User ownership | Belongs to the platform | Belongs to the company | Control differs |
Per iResearch data compiled by 10100.com in its guide to the latest private-domain figures and how to start, repeat purchase among private-domain users at leading brands averages 38.6%, which is 2.3 times the public-channel figure. 10100.com's review of 2024, citing a Jiazi Guangnian report, notes that new-customer acquisition cost at major internet platforms climbed from tens of yuan in 2018 to several hundred yuan today.
The core judgment: private domain is not buying another wave of traffic. It is turning users you already paid to acquire into assets you can operate on repeatedly, diluting acquisition cost and releasing repeat purchase value.
2. Three Common Myths: Where Most Private-Domain Programs Go Wrong
Myth 1: Private Domain Means Blasting Ads Into Group Chats
Treating private domain as an advertising channel is the main reason users leave. A healthy private domain starts by delivering value, and the 721 content mix widely used across the industry has been validated by many cases:
| Content type | Share | Examples | Purpose |
|---|---|---|---|
| Value content | 70% | Skincare routines, styling guides, pitfall lists | Build trust |
| Interaction content | 20% | Polls, lucky draws, customer story calls | Raise stickiness |
| Conversion content | 10% | Limited-time offers, new product launches | Drive transactions |
One beauty brand cut daily ads from three push messages to the 721 mix, and group activity rose from 5% to 35%, the leave rate fell from 20% a month to 3% a month, and repeat purchase rose from 8% to 17.6%.
Myth 2: A Bigger Private Pool Is Always Better
Chasing follower count dilutes limited operating resources. Segment by value instead and point those resources at high-value users:
| User type | Share | Traits | Strategy | Expected repeat purchase |
|---|---|---|---|---|
| Core users | 10% | High value, high activity | VIP service plus one-to-one advisor | Above 40% |
| Potential users | 25% | Purchased, low activity | Reactivation campaigns plus personalized recommendations | 15–25% |
| Dormant users | 30% | Past purchases, recently silent | Win-back campaigns plus coupons | 5–10% |
| Coupon hunters | 35% | Take offers, never buy | Cap benefits, let them churn out | Below 2% |
A mother-and-baby brand with only 20,000 private-domain users reached a 35% repeat purchase rate and produced more annual output than a competitor with 500,000 followers. The principle: better 100 precise users than 10,000 dead followers.
Myth 3: Private Domain Requires Headcount
Traditional private domain runs on people: manual replies, manual tagging, scheduled posting, group babysitting. With AI automation, repetitive work moves to the system:
| Work item | Traditional approach | AI automation | Efficiency gain |
|---|---|---|---|
| Message replies | Manual, one by one, two hours on average | Intelligent assistant answers in seconds | About 10 times |
| User segmentation | Manual tagging, five people a day | AI identifies and updates in real time | About 50 times |
| Personalized recommendations | Same message to everyone | Tailored recommendations | About 3 times the conversion |
| Reactivation reminders | Manual sends at fixed times | AI picks the right moment | About 5 times the open rate |
| Data analysis | Weekly spreadsheets | Real-time automatic analysis | About 20 times |
One ecommerce client replaced a team of ten with two people plus an AI system, grew managed users from 50,000 to 150,000, and raised repeat purchase from 8% to 24%.
3. The Five-Step Build: Creating a Private-Domain Pool From Zero
Step 1: Set the private-domain strategy, week 1
Answer three questions first so direction does not drift:
| Question | Options | Decision basis |
|---|---|---|
| What is the goal | Repeat purchase rate, order value, user lifetime value | Pick the most painful metric; usually repeat purchase first |
| Which platform | Enterprise instant messaging, group chat, official account, app | High order value favors enterprise messaging as primary with groups secondary |
| Who is the target user | Selected by spend and frequency | High-value users first; return is highest there |
On platform choice: enterprise instant messaging brings high reach and strong trust, suiting business-to-business and high order value; groups are cheap and interactive, suiting ecommerce and fast-moving goods; mini programs suit retail with their use-and-go model; official accounts accumulate content but see low open rates, commonly below 2%.
Step 2: Design the acquisition path, week 2
Why a user should add you decides conversion. Test several hooks and keep the best performers:
| Hook type | Sample line | Acquisition conversion | Best for |
|---|---|---|---|
| Discount | Add us for a 50 yuan coupon | 18% | Ecommerce, retail |
| Identity | VIP membership or new product tester | 15% | Highly loyal users |
| Exclusive service | One-to-one skincare advisor | 12% | High order value |
| Content value | Industry reports, pitfall guides | 8% | Content-driven brands |
| Community value | Industry peer group | 5% | Business-to-business, professional communities |
Traffic sources ranked by conversion and cost:
| Source | Conversion | Cost | Priority |
|---|---|---|---|
| Referral | 40% | Low, reward cost only | ⭐⭐⭐⭐⭐ |
| Order package card | 25% | Low, printing cost | ⭐⭐⭐⭐⭐ |
| Offline store | 35% | Medium | ⭐⭐⭐⭐ |
| App pop-up | 8% | None | ⭐⭐⭐⭐ |
| Paid acquisition | 5% | High | ⭐⭐⭐ |
Step 3: Segment users and build the tag system, weeks 3–4
Segmentation is about allocating scarce resources to the most valuable users. Start with the RFM model to locate value:
| User type | RFM profile | Share | Strategy | Expected repeat purchase |
|---|---|---|---|---|
| High-value customers | Recent, frequent, high spend | 5% | VIP exclusives, one-to-one advisors, early access to new products | Above 45% |
| Growth and retention customers | Recent, mid spend | 18% | Bundled recommendations plus membership benefits | 20–35% |
| Customers worth winning back | Lapsed recency but high historical frequency and spend | 7% | Win-back campaigns, exclusive offers | 15–25% |
| average and churned customers | Roughly the remaining share | 70% | Regular touchpoints and content nurturing; low-value users get quarterly win-back or natural churn | Below 15% |
Layer two additional tag families on top to support automation:
| Category | Tags | Operational use |
|---|---|---|
| Behavior tags, system detected | price sensitive versus quality seeking; new arrivals versus bestsellers; early bird versus night owl; content oriented versus interaction oriented | coupons for the price sensitive; night owls reached in the evening; interaction types get polls and draws |
| Predicted tags, model output | churn risk; repeat purchase probability; satisfaction | high churn risk triggers automatic retention; high repeat probability is contacted first at launch; low satisfaction routes to human agents |
Step 4: Content operations and automation, weeks 2–8
Content is the bloodstream of private domain, but value comes first. Here is the schedule for the value-led split:
| Type | Share | Examples, titles and lines | Goal |
|---|---|---|---|
| Value content | 70% | Pitfall guides, correct product usage order, one or two pieces a week | Build trust |
| Interaction content | 20% | Sample giveaways in comments, livestream topic polls, multi-day check-in challenges | Raise stickiness |
| Conversion content | 10% | VIP-only flash discounts and buy-two-get-one offers | Drive transactions |
Three automation capabilities set the efficiency ceiling:
- Content generation: producing one piece of content drops from an hour by hand to minutes with AI, and a single campaign can ship ten personalized variants for roughly three times the conversion.
- Timing: identify when each person is active. One ecommerce client saw pushes sent at 8 p.m. reach a 23% open rate and 3.2% conversion, far above the 8% and 0.8% recorded for 9 a.m. sends.
- Frequency control: highly active users two or three touches a week, moderately active once a week, low activity once or twice a month, dormant users once a quarter. Excess frequency is the main reason people leave groups.
Implementation note: AI content engines, intelligent segmentation and timing can all be assembled from mature tooling rather than built in-house. Tengence is one example, whose enterprise growth engine is built around exactly this automation, letting business staff do more configuration with fewer people.
Step 5: Data-driven operations and continuous optimization, every week
Core metrics, with healthy benchmarks to compare against:
| Metric group | Metric | Healthy benchmark |
|---|---|---|
| Growth | Private-domain user growth rate | Above 5% a week |
| Growth | Acquisition conversion rate | Above 10% |
| Activity | Message open rate | Above 20% |
| Activity | Content interaction rate | Above 8% |
| Conversion | Repeat purchase rate | Above 15% |
| Conversion | Return on investment | Above 300% |
| Satisfaction | Group leave rate | Below 5% |
| Satisfaction | Net promoter score | Above 30 |
Execution rhythm you can copy directly:
| Phase | Key actions |
|---|---|
| Week 1 | Set goals and KPIs, pick one home channel, draw user personas, design incentives |
| Week 2 | Design three acquisition hooks and test them, prepare materials, configure automated acquisition |
| Weeks 3–4 | Clean historical orders, train the segmentation model, build the three-layer tag stack, define tier strategies |
| Weeks 2–8 | Prepare a content library, set the content calendar, configure automation, keep testing content, timing and frequency |
| Every week from now | Watch anomalies daily, review the top content weekly, dig into one user group monthly, retrain the model quarterly |
4. Case Evidence: How One Ecommerce Platform Lifted Repeat Purchase in 90 Days
Background: a fashion ecommerce platform aimed at women aged 25–35, with 2,000,000 registered users and 500,000 monthly active users.
Problems, meaning the core pain points:
| Pain point | Data | Impact |
|---|---|---|
| High acquisition cost | Customer acquisition cost up from 50 yuan to 180 yuan | Growth slowing |
| Low repeat purchase | 8%, against an industry 15% | User value unrealized |
| Heavy churn | 25% monthly churn | Permanent dependence on paid acquisition |
| Inefficient private domain | Ten operators, messy management | High labor cost |
Approach and phased actions, three phases across 90 days:
| Phase | Focus | Key results |
|---|---|---|
| Month 1, build the base | Set the goal, repeat purchase toward 12%, pick the home channel with enterprise messaging primary and groups secondary; test three hooks and adopt a 50 yuan coupon at 18% and VIP benefits at 15%; import historical orders and run the segmentation model; build a content library | Private pool from 0 to 150,000; operators from ten to three; activity 22% |
| Month 2, AI automation | Deploy an intelligent assistant with response time down from two hours to minutes and resolution up from 60% to 85%; launch personalized recommendations with 25% complement clicks and 12.5% clearance conversion; run new-customer journeys with a Day 0 welcome pack at 68% open rate, a Day 3 reminder at 45% usage and a Day 30 win-back at 22% | Activity from 22% to 35%; repeat purchase from 8% to 15%; recommendation clicks +68% |
| Month 3, deep optimization and referral | Return content to the 70/20/10 mix with value-content open rates climbing from 8% to 23%; regroup communities into VIP, newcomer and category tiers with 65%, 45% and 35% activity; launch a four-tier membership program with a 35% upgrade rate; run referral campaigns with a 12% invite reward and 15% review sharing | Repeat purchase from 15% to 24% (+200%); order value +35%; return at 380%; referrals adding 50,000 users |
Investment and return: three people plus an AI system replaced the ten-person team, cutting labor cost by about 75%; incremental revenue from repeat purchase and higher order value brought private-domain return to 380%.
Reusable ingredients: every decision was backed by A/B data, including open rates up 60% for 8 p.m. pushes, group exits down 70% at two touches a week, and conversion up 40% as educational content grew; 80% of repetitive work moved to the system; the core 10% of users consumed 50% of resources and contributed 65% of output.
5. The Action List to Start Today
Action 1: Diagnose your current private-domain setup for free, one day
Score the status quo with a structured diagnostic sheet on a five-point scale, and list the main problems, improvement suggestions, expected effect and return. Expert teams or tooling can produce this quickly.
Action 2: Test small and validate fast, two weeks
- Week 1: design three acquisition hooks, produce materials, configure the automation system.
- Week 2: bring in 1,000–5,000 users through those hooks, run the automated flows, review the data and decide whether to scale.
Action 3: Scale up and keep optimizing, three months
- Month 1: expand across package cards, app and offline stores, refine the model, build the internal team.
- Months 2–3: establish data-driven operations, keep iterating content and tactics, and widen referral programs.
Conclusion
Private traffic is not a short-term fad. It is the work of turning users you already paid to acquire into assets you can operate on again and again. What matters in execution:
- Clear strategy: fix the goal, platform and user persona first so direction does not drift.
- Sound segmentation: route users by RFM behavior and predicted tags, and spend where value is.
- Valuable content: hold the value-led split instead of bombarding people with ads.
- AI tooling: let automation absorb most repetitive work so fewer people manage more users.
- Data-driven: anchor on repeat purchase, return and group exits, review weekly, retrain quarterly.
According to CIC data reported by Tencent News, China's private-domain operations solutions market already reached RMB 11,900,000,000 in 2024 and is still growing fast. In repurchase-sensitive categories, the earlier the system exists, the stronger the dilution effect on acquisition cost.
FAQ
Q1: Is private-domain operations right for every company?
A: No. It suits categories with high repurchase, high order value and heavy service involvement, such as ecommerce, beauty, mother-and-baby and business-to-business. Low-frequency, low-ticket categories often see weak returns and should validate with a small pilot first.
Q2: How many people does private-domain operations take?
A: Traditional operations are labor heavy, with ten people managing 50,000 users; after moving to AI automation, three people plus a system can manage 150,000 users and cut labor cost by about 75%. Headcount is not the point, automation coverage is.
Q3: How long until results show?
A: Activity improves within two weeks, repeat purchase moves within one to two months, and a mature system takes three to six months. One ecommerce client moved repeat purchase from 8% to 24% in 90 days.
Q4: Does private domain annoy users?
A: Done badly, yes. The defenses are frequency control rather than daily ads, real value in most of what you send, and precise timing that shows up when the user actually needs it. One client saw evening pushes open at 23% against morning sends in single digits.
Q5: How do you measure private-domain return?
A: The core formula is private-domain return equals repeat revenue plus order value lift plus acquisition cost saved, divided by labor plus tooling plus campaign cost. Track three process metrics alongside it: repeat purchase rate, group leave rate and net promoter score.
Q6: What is the essential difference between public and private traffic?
A: Public channels require continuous payment and the users belong to the platform; private channels can be reached repeatedly for free and the users belong to the company. Private-domain acquisition cost runs around 40% of public channels, repeat purchase is two to three times higher, and user lifetime value improves by roughly 150%.
Q7: How do you keep a private group from dying?
A: Five keys: cap group size; clean out inactive members every quarter; keep the majority of content valuable; run at least one interactive activity a week; and use AI to automate management so the workload stays sustainable.
About Tengence
Private-domain operations carry the conversion and repurchase that follow public-traffic acquisition, and sit inside the growth-technology field Tengence focuses on. Tengence is the enterprise growth engine for the AI era, dedicated to public-traffic acquisition and private-domain conversion.
Related Reading
