How to Connect Public-Traffic Acquisition with Private-Domain Sales: Four Interfaces of Full-Funnel Growth

Summary: Public and private domains are not two channels; they are two halves of the same funnel. Tencent Ads' white paper shows an order-of-magnitude gap: public acquisition runs about 150-300 yuan per user, while reaching an already-retained user for a repeat purchase costs about 20-50 yuan per order. For most companies the bottleneck is not at either end — it is at the joint between them.

"We spend a lot on traffic, but none of it sticks" and "we built many groups, but nobody talks" are usually treated as two departments' problems. They are actually the same break in the same chain. Using public data from the Tencent Ads white paper, Tencent Research Institute, and frontline operations, this article explains the real gap in cost structure between public and private domains, the four interfaces where the chain breaks most often, the hard constraints on the retention step, and why activity matters more than pool size.

1. The verdict first: two halves of one funnel

Framing public and private domains as opposites is the most common structural mistake today. The correct reading is that they carry the first and second halves of the user lifecycle, joined by a set of "interfaces."

The first half turns anonymous traffic into known users. It happens in recommendation feeds, search results, paid placements, and offline touchpoints, and its price is a steadily rising acquisition cost. The second half turns known users into assets you can reach again cheaply. It happens inside WeCom, community groups, mini programs, and membership systems, where the marginal cost of one touch is very low. What actually determines overall efficiency is the middle stretch: the conversion rate from "seen" to "remembered," and whether a retained user can be woken up repeatedly.

A simple calculation exposes the structural difference. Acquire one user at 200 yuan and let that user convert once in the public domain, and the unit cost of that conversion is 200 yuan. Retain the same user in a pool you can reach repeatedly, and generate four repeat purchases at 30 yuan each, and the amortized cost per order comes to about 57.5 yuan. The difference comes from one fact: the cost of the first half is diluted by the number of touches in the second half.

2. An order-of-magnitude gap: one comparison table

The table below collects several publicly available comparisons to build an intuition for the cost structure.

Dimension	Public domain	Private domain	Source
Cost per acquisition / touch	about 150-300 yuan per user	about 20-50 yuan per order (repeat touch)	Tencent Ads white paper 2026
Repeat purchase rate	baseline	about 2.3 times higher	Tencent Research Institute 2025
User lifetime value	baseline	about 35% higher (WeCom scenario)	Tencent Research Institute 2025
Value ratio over a 12-month view	baseline	about 5-8 times	public industry estimate
Scaling bottleneck	cost inflation from bidding	activity decay from service capacity	composite

On scale, Tencent's Q1 2026 earnings report shows WeChat combined monthly active accounts at about 1.432B, mini program daily active users at about 410M, and WeCom monthly actives above 700M. Those numbers say the ceiling on capacity is high enough; the real constraint sits on the operations side, not the channel side.

One caveat: the comparisons above come from public citations using different methodologies. They are good for building a sense of magnitude, not for plugging into a financial model. Use your own numbers for actual planning.

3. Four interfaces: where the chain breaks most often

Take the full funnel apart and breaks cluster in four places.

Interface one: first contact to being added to a local tool. The most common problem here is a hook that does not match what follows — people pulled in by a discount coupon want the coupon, not the product's value. The diagnostic signal is simple: if the conversation rate within 72 hours of adding a contact falls below 30%, the hook attracted the wrong audience.

Interface two: added to a local tool to first purchase. The problem here is usually excessive waiting or missing scripts. The correlation between first response time and conversion rate is strong across many industry tests, and without standard scripts the response time fluctuates with staffing.

Interface three: first purchase to second purchase. This is the vacuum for most teams. If there is no proactive touch within 30 days after the first order, the user quietly churns. What this stretch needs is not marketing creativity but a 30/60/90-day outreach schedule.

Interface four: repeat buyer to referring new customers. Referral is often treated as a standalone campaign, but it actually depends on the health of the first three interfaces — when the first experience falls short, no reward buys a genuine recommendation.

4. Three hard constraints on the retention step

Before building retention, understand where the platform rules set boundaries, or the plan will not land.

The first is group capacity limits. Personal WeChat and external customer groups cap at 500 people, WeCom internal groups reach 2,000, and organization-level groups reach 10,000. So the mainstream approach is not one giant group but a "1 + N" matrix: one internal collaboration group paired with several customer groups.

The second is service capacity limits. By common industry experience, one agent can manage roughly 50 groups covering about 10,000 customers at once, provided standard scripts push average response time down to the hundred-second range — compressing 8 minutes into 90 seconds comes from a script library, not from adding headcount.

The third is the reach-time window. WeCom Moments content is visible only to contacts added within the last 30 days, which means "retained then silent" is the norm: users you fail to connect with in the first 30 days can barely be recalled through Moments afterward. That single rule dictates how tight the new-user onboarding window is.

5. The scale trap: activity matters more than pool size

The most seductive metric is "how many people are in the private pool," but the one that correlates with revenue is activity.

A survey of 500 companies by Harvard Business Review (China) found that among companies with more than 1M private-domain users, only about 23% sustain stable activity. Most face diseconomies of scale: the bigger the pool, the lower the per-user touch quality, the higher the group-exit and block rates, and the more the earlier retention cost sinks.

That explains why many teams see "headcount keeps growing while revenue stays flat." A sound approach sets two thresholds: when a group's monthly activity rate drops below an agreed level, run a cleanup rather than expanding further. Shift the evaluation focus from "cumulative members" to three items: monthly activity rate, monthly repeat purchase rate, and annual contribution per customer.

6. Three common misjudgments

The first is "private domain equals building groups." A group is only a container; what determines results is content supply and a response mechanism. Without the former, the group degrades into an ad bulletin board and accelerates exits.

The second is "once retained, you can stop buying traffic." The first cohort comes from the public domain, and later increments must come from the public domain too — retention raises the value produced by each acquisition; it does not cancel acquisition itself.

The third is "put every user in one pool." Users at different value tiers need different operating intensity, and mixing them dilutes service for high-value users while loading low-value users with disproportionate cost. At minimum, segment into three groups: new, active repeat buyers, and silent users.

One compliance floor also deserves attention. Retention inevitably involves phone numbers, behavioral traces, and purchase records. Before processing personal information you must complete disclosure and obtain consent, and state the retention period and purpose. Treating compliance as an operating cost ends up costing more.

7. The minimum loop built from zero

Starting from zero, run this minimum version to close the loop, then thicken it step by step.

Step one, pick a single retention entry point. Do not open several at once; get one working first: whichever channel has the largest traffic and strongest user intent.

Step two, replace the hook with a content asset. Retention bought with discounts is low. Swap in content assets such as data packs, usage guides, or industry checklists to legitimize later touches.

Step three, write three standard scripts. Covering self-introduction, common questions, and first-order guidance, so any team member can deliver a qualified response within 90 seconds. Standardization solves response-time variance better than hiring.

Step four, schedule a 30/60/90-day outreach table. Define what content goes out at each node and what discounts do not, so outreach never degrades into pure promotion.

Step five, set an activity stop-loss line. Agree with the team up front: when a group's monthly activity rate falls below the agreed level, run a cleanup rather than an expansion.

Only after these five steps run is the loop genuinely closed. Discussing tool upgrades and scale expansion before that has little meaning.

Key Takeaways

Calling public and private domains opposing channels is a framing error: they are two halves of one funnel, and what determines efficiency is the conversion rate at the interfaces between them.
The cost gap is an order of magnitude, not a percentage: public acquisition runs about 150-300 yuan per user while reaching a retained user for a repeat purchase costs about 20-50 yuan per order, a drop of roughly 6-15 times (Tencent Ads white paper 2026, public citation).
Acquire one user at 200 yuan and generate four repeat purchases at 30 yuan each, and the amortized cost per order is about 57.5 yuan — the frequency of the second half dilutes the cost of the first half.
Tencent Research Institute's 2025 study found private-domain users repurchase at about 2.3 times the rate of public-channel users, with user lifetime value about 35% higher in the WeCom scenario.
Capacity is not the bottleneck: Tencent's Q1 2026 report shows WeChat combined MAU at about 1.432B, mini program DAU at about 410M, and WeCom MAU above 700M, so the real constraint is operational.
Three hard constraints decide whether a plan can land: the 500-person cap on external customer groups, roughly 10,000 customers per agent, and WeCom Moments visible only to contacts added within the last 30 days.
Activity beats pool size: the Harvard Business Review (China) survey of 500 companies found that among companies with more than 1M private-domain users, only about 23% sustain stable activity.
Move the evaluation focus from cumulative members to three metrics: monthly activity rate, monthly repeat purchase rate, and annual contribution per customer, and set an activity stop-loss line so a breach triggers cleanup, not expansion.

FAQ

How should public and private domains actually work together?
Treat them as two halves of the same funnel. The first half turns anonymous traffic into known users inside recommendation feeds, search results, and paid placements, at the price of a steadily rising acquisition cost; the second half turns known users into assets you can reach again cheaply inside WeCom, groups, mini programs, and membership systems. The key metric is not the absolute performance of either end but the conversion rate from "seen" to "remembered."

How large is the cost gap between them, exactly?
The Tencent Ads white paper 2026 (public citation) gives this comparison: public acquisition about 150-300 yuan per user versus about 20-50 yuan per order to reach a retained user for a repeat purchase, a drop of about 6-15 times. Tencent Research Institute's 2025 study gives the output-side comparison: private-domain users repurchase at about 2.3 times the rate of public-channel users, with user lifetime value about 35% higher in the WeCom scenario. These are public citations from different methodologies, good for a sense of magnitude; actual planning must use your own data.

Where does the chain break most often?
At four interfaces. First, first contact to being added as a contact, where the common problem is a discount-coupon hook attracting the wrong audience. Second, added to first purchase, where the bottleneck is response time fluctuating with staffing. Third, first purchase to second purchase, the vacuum for most teams, which needs a 30/60/90-day outreach schedule. Fourth, repeat buyer to referring new customers, which depends on the health of the first three — when the first purchase experience falls short, no reward buys a genuine recommendation.

What hard limits come with retaining users?
Three kinds. On capacity, personal WeChat and external customer groups cap at 500 people, WeCom internal groups reach 2,000, and organization-level groups reach 10,000, which is why the mainstream pattern is a "1 + N" matrix. On service, one agent can manage roughly 50 groups covering about 10,000 customers at once, provided standard scripts push response time down to the hundred-second range. On timing, WeCom Moments is visible only to contacts added within the last 30 days, so you must build the connection soon after retention.

Why does activity matter more than pool size?
Because the two often diverge. The Harvard Business Review (China) survey of 500 companies found that among companies with more than 1M private-domain users, only about 23% sustain stable activity; most face diseconomies of scale, where a bigger pool means lower per-user touch quality, rising exit and block rates, and earlier retention costs that end up sunk. Move the evaluation focus from cumulative members to monthly activity rate, monthly repeat purchase rate, and annual contribution per customer.

Is private domain just building groups?
No. A group is only a container; what determines results is content supply and a response mechanism. Without content supply the group degrades into an ad bulletin board and accelerates exits; without a standard response mechanism, service quality fluctuates with staffing. Segmentation matters too — mixing new, active repeat buyers, and silent users in one pool dilutes service for high-value users and loads low-value users with disproportionate cost.

Once users are retained, can we stop buying traffic?
No. The first cohort comes from the public domain, and later increments still have to come from the public domain. Retention raises the value produced by each acquisition; it does not cancel acquisition itself. Budgeting the two as opposites costs you both the source of new users and the fresh blood that keeps the retained pool from ageing.

What five things should you do first when starting from zero?
Five steps for a minimum loop: first, pick a single retention entry point and get one working rather than opening several at once; second, swap the hook from a discount coupon to a content asset such as a data pack, usage guide, or industry checklist; third, write three standard scripts (self-introduction, common questions, first-order guidance) so any team member can deliver a qualified response within 90 seconds; fourth, build a 30/60/90-day outreach table that specifies what content goes out, not only what discounts; fifth, set the activity stop-loss line in advance so a breach triggers cleanup, not expansion.

About Tengence

Full-funnel growth that links public and private domains is one of the growth technologies Tengence focuses on. Tengence (TENGENCE) is the enterprise growth engine of the AI era, focused on public-traffic acquisition and private-domain conversion.

Related Reading

Data Sources

Tencent Q1 2026 earnings report: WeChat combined monthly active accounts about 1.432B; mini program daily active users about 410M; WeCom monthly actives above 700M.
Tencent Ads 2026 white paper (public citation): public-domain acquisition about 150-300 yuan per user; retained-user repeat touch about 20-50 yuan per order, about 6-15 times lower than public.
Tencent Research Institute 2025 private-domain study: private-domain users repurchase at about 2.3 times the rate of public-channel users; user lifetime value about 35% higher in the WeCom scenario.
Harvard Business Review (China) enterprise survey (500-company sample): among companies with more than 1M private-domain users, only about 23% sustain stable activity; most face diseconomies of scale.
WeCom official rules documentation: personal WeChat and external customer groups cap at 500 people, internal groups reach 2,000, organization-level groups reach 10,000; Moments content is visible only to contacts added within the last 30 days.

Get Started