You launched your Xiaohongshu ad campaign with confidence — the creative looked polished, the targeting parameters made sense, and the budget was in place. Then the dashboard loaded. Impressions crawled, clicks stalled, and ROAS told a story nobody wanted to read. If this sounds familiar, you are far from alone. At Hashmeta, our specialists have reviewed hundreds of underperforming campaigns across Xiaohongshu marketing accounts, and the same diagnostic gaps come up repeatedly: brands jump to optimisation tactics before they have correctly identified the root cause of the problem.

Xiaohongshu’s advertising ecosystem is layered. A campaign can fail at the impression stage, the click stage, the engagement stage, or the conversion stage — and each failure point has a distinct set of causes and remedies. Treating a targeting problem as a creative problem, or a bidding problem as a landing-page problem, wastes both budget and time. This article provides a structured, step-by-step troubleshooting decision tree that helps you isolate exactly where your campaign is breaking down and what to do about it — before you spend another cent on guesswork.

Why Diagnosis Before Optimization Matters

The instinct when a campaign underperforms is to act fast — swap the creative, raise the bid, broaden the audience. But Xiaohongshu’s ad platform punishes reactive, poorly-reasoned changes. Poorly managed ads do not simply waste budget in isolation; as Hashmeta’s performance team has observed, they can train the platform algorithm to associate your account with low-quality signals, compounding future performance problems. The platform rewards accounts that demonstrate consistent relevance and user value, which means every unjustified change you make during the learning phase carries a cost beyond the immediate spend.

Systematic diagnosis protects your account health. Before touching a single campaign setting, you need to understand which metric is failing, by how much, and relative to what benchmark. Average CTRs on Xiaohongshu typically range from 0.8% to 2.5%, with well-optimised campaigns exceeding 3%, while conversion rates for e-commerce objectives generally fall between 1% and 8% depending on the industry and offer. Knowing where your numbers sit within these ranges tells you which layer of the funnel deserves your attention — and which layers are actually performing fine.

How to Use This Decision Tree

Think of this framework as a clinical triage process. Each node represents a specific performance question. You answer it using your Aurora ad platform data, then follow the branch that matches your situation. Work sequentially from Node 1 through to Node 5 — do not skip ahead, because a problem at an earlier node often masks what looks like a problem at a later one. For example, a catastrophically low conversion rate can look like a creative failure when the real issue is that the ad is reaching completely the wrong audience in the first place.

Pull your campaign data for a minimum of seven days (ideally fourteen or more) before running through this tree. Shorter windows introduce too much statistical noise, especially for campaigns with modest budgets. With sufficient data in hand, open your Xiaohongshu advertising dashboard and work through each diagnostic node below.

Node 1: Are Your Ads Receiving Impressions?

Symptom: Your ad has been live for several days but impressions are negligible or zero despite having budget available.

Low or no impressions almost always point to a delivery problem rather than a creative problem — your ad is not even entering the auction competitively enough to be shown. The most common causes fall into three categories: bid, budget structure, or approval status.

Work through the following checks in order:

1. Check approval status first. Xiaohongshu’s review process is strict, and approximately 40% of first-time submissions face rejection or require revisions. Confirm your ad has passed review in the platform dashboard before assuming a delivery issue.
2. Evaluate your bid against the platform’s suggested range. The Aurora platform provides a bid suggestion range during campaign setup that reflects current auction conditions for your selected targeting parameters. If your bid sits significantly below the lower end of the suggested range, you are consistently losing auctions before your quality score even factors in. Raise your bid to at least the midpoint of the suggested range and monitor impression volume over the next 48 hours.
3. Check for audience over-narrowing. Hyper-specific targeting combinations — for example, a narrow age range layered with specific interests, device types, and geographic restrictions — can reduce your eligible audience to a size too small to generate meaningful delivery. Temporarily broaden one targeting dimension at a time to see which constraint is the binding one.
4. Review your daily budget cap. A daily budget that is too low relative to your target CPM means the campaign exhausts spend early in the day and goes dark during peak usage hours. Xiaohongshu usage peaks during commute times, lunch breaks, and evenings — if your budget runs out before these windows, you are structurally invisible to your best prospects.

If impressions recover after addressing one of these factors, pause here and allow the campaign to accumulate data before moving to Node 2. If impressions remain critically low despite correcting all of the above, escalate to a platform compliance review — your account or ad creative may have a flagging issue that requires specialist attention.

Node 2: Are Users Clicking? (CTR Diagnosis)

Symptom: Impressions are healthy, but your click-through rate is well below the 0.8–2.5% platform average — users are seeing your ad and scrolling past it.

A low CTR on Xiaohongshu is almost always a signal of poor creative-audience fit. Users are being served the ad, but the content is not resonating strongly enough to interrupt their scroll. This can stem from the creative itself, from a mismatch between the creative and the audience it is reaching, or from a disconnect between the ad’s visual language and Xiaohongshu’s native aesthetic norms.

Diagnose CTR problems using the following branches:

1. Audit your cover image and headline. Xiaohongshu is fundamentally a visual platform where aesthetic appeal significantly impacts performance. Images should be bright, clean, and professionally styled while maintaining an authentic, user-generated feel. Overly polished corporate photography consistently underperforms compared to lifestyle imagery that feels personally created. If your cover image looks like a product catalogue rather than a user’s personal post, your CTR will reflect it.
2. Check for creative fatigue. Creative fatigue occurs faster on Xiaohongshu than on platforms with larger user bases. If the same creative has been running for more than 7–14 days to the same audience segment, frequency may have reached the point where users have already dismissed your ad once and are now conditioned to ignore it. Introduce new creative variations and monitor whether CTR recovers.
3. Assess audience-message alignment. Even visually strong creative will underperform if it reaches the wrong audience. Review your targeting parameters and ask honestly: does this creative speak directly to the interests, values, and lifestyle of the specific audience segment I have defined? An influencer marketing lens is useful here — the most effective Xiaohongshu ads read like a trusted recommendation from someone the audience already follows, not a generic brand announcement.
4. Test UGC-style formats. Boosting creator-led or UGC-style content instead of running only brand-produced ads can meaningfully reduce CPC and improve engagement, because Xiaohongshu’s algorithm prioritises authentic, peer-style content. If your CTR is persistently low despite creative refreshes, consider whether your ad format itself is the barrier — Spark Ads that amplify existing organic notes, for instance, receive significantly higher engagement than standard ad creatives because users interact with them as native content.

Node 3: Are Users Engaging? (Engagement Rate Diagnosis)

Symptom: CTR looks acceptable, but likes, comments, saves, and shares are low relative to the volume of users who have viewed or clicked your content.

Engagement on Xiaohongshu carries a significance beyond vanity metrics. High engagement signals to the platform’s algorithm that your content is valuable, which can trigger broader organic distribution and improve your quality score over time. Low engagement — even with decent CTR — suppresses this virtuous cycle and leaves your campaign dependent entirely on paid delivery.

The save rate deserves particular attention here. Saving a post to a collection is a distinctly Xiaohongshu behaviour that signals strong purchase intent and deep content resonance. A high save rate with low comments suggests the audience finds the content genuinely useful but does not feel prompted to interact publicly. A low save rate alongside low comments is a clearer signal that the content is failing to create real value for the reader.

1. Evaluate content depth and utility. Xiaohongshu users expect authentic, valuable content rather than interruptive advertising. If your ad delivers a hard sell rather than genuinely useful information, lifestyle inspiration, or entertainment, users will click away without engaging. Restructure your content around a clear value proposition for the reader — a tutorial, a comparison, a lifestyle scenario — before making the brand or product the secondary focal point.
2. Check cultural and linguistic localisation. Engagement drops sharply when Chinese consumers detect content that has been translated rather than genuinely conceived for the platform. Terms suggesting superiority, unverified claims, or language that carries different implications in the Chinese cultural context will not just suppress engagement — they can trigger negative comments that further damage your quality score. Work with native Chinese speakers and consult a content marketing specialist to ensure your messaging is culturally resonant, not just linguistically accurate.
3. Review your CTA mechanics. Xiaohongshu users often need a specific prompt to comment or save. Adding a question to your caption, inviting users to share their experience, or structuring content as a listicle that users naturally want to save for reference can each lift engagement rates without changing the underlying creative.

Node 4: Are Clicks Converting? (Conversion Diagnosis)

Symptom: Impressions, CTR, and engagement all look reasonable, but the conversion rate is low — users are clicking your ad and not completing the desired action.

This is one of the most frustrating failure modes because the top-of-funnel metrics look encouraging. The problem, however, has almost certainly shifted from the ad itself to what happens after the click. Low conversion rates with healthy upstream metrics point to landing page issues or a mismatch between the ad’s messaging and the offer it leads to.

1. Audit the post-click experience for message continuity. If your ad promises a specific product, price, or outcome, the destination page must immediately deliver on that promise. Any friction — a generic homepage, a slow-loading page, a page in the wrong language, or a price that differs from what the ad implied — will cause users to abandon. Low conversion rates point to landing page issues or misalignment between ad messaging and offer more often than they point to a creative failure.
2. Evaluate page load speed and mobile optimisation. Xiaohongshu is a mobile-first platform. A landing page that takes more than two to three seconds to load on a mobile connection will haemorrhage conversions regardless of how compelling the ad was. Use Xiaohongshu’s native in-app landing pages or a mobile-optimised destination built specifically for this traffic. An ecommerce web design specialist can identify and resolve technical barriers that silently suppress conversion rates.
3. Check whether your audience is at the right funnel stage. Xiaohongshu functions as a discovery and consideration platform at its core. Users often visit the platform to research products, not to make immediate purchase decisions. If you are using a CPA conversion objective with a cold audience that has never encountered your brand before, your conversion rate will naturally be low — not because the ads are failing, but because you are asking for a commitment the user is not yet ready to make. Layer in retargeting campaigns for users who have previously engaged with your notes or visited your profile before applying hard conversion pressure.
4. Verify your tracking implementation. Before concluding that conversions are low, confirm your tracking is working correctly. Incorrect tracking implementation, broken links, or unapproved third-party technologies are a common source of underreported conversions. Use Xiaohongshu’s native tracking tools and validate them through the platform’s testing tools before drawing conclusions from conversion data. Considering using AI marketing tools to assist with attribution modelling across the full customer journey.

Node 5: Are Your Costs Spiralling? (Efficiency Diagnosis)

Symptom: The campaign is generating impressions, clicks, and some conversions, but CPC, CPM, or CPA are climbing faster than your results justify — the campaign is becoming progressively less efficient over time.

Escalating costs with stable or declining performance is a sign that your quality score is eroding, competition in your auction is intensifying, or your campaign structure has a fundamental inefficiency that compounds over time. Brands running broad, untargeted campaigns waste 40–50% of their budget on irrelevant impressions, and that waste actively signals poor quality to the platform algorithm.

1. Segment your audience and pause underperformers. Analyse performance across demographic segments, interest categories, and placements. High-performing segments being averaged with low-performing ones distort your efficiency metrics and your quality signals simultaneously. Create separate ad groups for your best-performing audience segments, pause the weakest segments, and reallocate budget to proven combinations.
2. Implement dayparting. Concentrating spend during high-engagement windows — evenings and weekends when Xiaohongshu usage peaks — improves the ratio of engaged impressions to total impressions, which supports quality score over time. Spreading budget evenly across low-traffic hours dilutes this signal and inflates effective CPM.
3. Reassess your bidding strategy against your campaign objective. A CPM bid strategy for a conversion-focused campaign introduces structural inefficiency from the outset. Ensure your bidding model aligns with your objective: CPM for awareness, CPC for traffic-focused campaigns, and CPA for advertisers with sufficient performance history to support algorithmic optimisation. An AI marketing agency approach to bid management — using data signals to continuously refine bid levels rather than setting and forgetting — consistently delivers lower cost-per-acquisition outcomes.
4. Review for quality score decline indicators. While Xiaohongshu does not display quality scores directly, declining CTR, falling engagement rates, and rising CPMs over time are reliable proxy signals that your quality score is eroding. Improving quality score by even one point can reduce costs by 5–15%, which means proactive quality maintenance is one of the highest-leverage activities in ongoing campaign management.

When to Escalate: Signs You Need a Specialist

This decision tree covers the majority of underperformance scenarios that brand teams can resolve independently. However, some situations warrant escalation to an experienced Xiaohongshu marketing specialist. If you have worked through all five nodes, implemented the recommended fixes, and your campaign metrics have not improved meaningfully after 14 days, one of the following deeper issues is likely at play.

• Account-level quality penalties — repeated ad rejections, policy violations, or a history of low-quality campaigns can create account-level signals that suppress performance across all new campaigns, regardless of how strong the individual creative is.
• Category or product restrictions — some product categories face structural advertising limitations on Xiaohongshu that require a fundamentally different strategy, such as focusing on adjacent compliant content rather than direct product promotion.
• Attribution complexity — if your conversion journey spans multiple touchpoints across Xiaohongshu, WeChat, and external e-commerce destinations, standard platform attribution will undercount performance and lead to incorrect optimisation decisions. Sophisticated multi-touch attribution requires specialist tooling and methodology.
• Scaling ceilings — campaigns that perform well at low spend but collapse when budgets are increased typically have an audience size or creative pool problem that requires a structured scaling strategy rather than a simple budget increase.

Hashmeta’s team of Xiaohongshu specialists uses proprietary data frameworks and platform expertise built across more than 1,000 brand engagements to diagnose and resolve these deeper structural issues. Whether you need an independent audit or full campaign management, our AI-powered marketing methodology means recommendations are driven by data, not guesswork. You can also leverage tools like AI Influencer Discovery to identify creator partnerships that strengthen both organic and paid performance simultaneously.

Conclusion

An underperforming Xiaohongshu ad is rarely a single problem — it is a layered set of questions that need to be asked in the right sequence. Working from impression delivery through to cost efficiency, this decision tree gives you a logical path from symptom to diagnosis to action, rather than a list of generic optimisation tips that may or may not apply to your specific situation.

The discipline of diagnosing before optimising is what separates brands that consistently improve their Xiaohongshu results from those that cycle through creative changes and bid adjustments without ever understanding why performance fluctuates. Use this framework as a standing audit protocol — not just a one-time rescue exercise — and your campaigns will compound in quality and efficiency over time.

For brands that want expert-led diagnosis and hands-on campaign management, Hashmeta’s Xiaohongshu marketing team is ready to help. Our data-driven approach, combined with deep platform expertise across Singapore, Malaysia, Indonesia, and China, means we identify root causes quickly and implement fixes with precision. Reach out today to get a customised audit of your current campaigns.