# Laptop SGN — E-commerce Analysis Report

**Client domain:** Tech e-commerce (Vietnam) · **Report date:** 15 November 2025
**Source:** Google Analytics 4 export · **Period:** 8 Aug – 7 Sep (30 days, back-to-school demand peak)
**Volume:** ~100,000 rows across session, user, item and transaction level
**Stakeholders:** Marketing, E-commerce / Merchandising, Growth

> Markdown transcription of the 41-slide analysis deck.
> Original deck: <https://tuanphanportfolio.my.canva.site/laptop-sgn-digital-marketing-analysis>

---

## 1. Dataset overview

Five working tables were assembled from the GA4 export.

| Table | Grain | What it carries |
|---|---|---|
| `traffic_sessions` | Session | Channel, session quality, revisit, bounce, session duration, new vs returning |
| `user_profile_summary` | User | Age, gender, new vs returning, signup, long-term behaviour |
| `page_and_funnel_events` | Event | Page views, funnel steps, purchase journey, session signals |
| `ecommerce_items_transactions` | Transaction / item | Revenue, transactions, products sold, AOV, product and category |
| `analyst_derived_segments` | Session / user | Pageview, duration, session-count, recency buckets and co-browsing flags |

> The transactions table was derived from GA4 source data rather than downloaded from GA4 as-is.

![Session table schema](../../assets/images/marketing/laptop-sgn/laptop-sgn-02.png)
![User profile schema](../../assets/images/marketing/laptop-sgn/laptop-sgn-03.png)
![Funnel event schema](../../assets/images/marketing/laptop-sgn/laptop-sgn-04.png)
![Transactions schema](../../assets/images/marketing/laptop-sgn/laptop-sgn-05.png)
![Derived segments schema](../../assets/images/marketing/laptop-sgn/laptop-sgn-06.png)

---

## 2. Analytics framework

**Governing question: how can we grow e-commerce revenue?**

| Pillar | Questions answered |
|---|---|
| **User acquisition** | Which channels bring the most traffic and users? Which channels drive repeat visits? |
| **User segmentation** | Who are our core user groups? Which segments create the most value? Which behaviours signal intent? Which product interests link to value? |
| **Funnel performance** | Where are the biggest drop-off points? How do users browse across categories before purchase? |
| **User retention** | How much more valuable are returning users? How does value change with frequency and recency? When should we re-engage? How should users be prioritised for CRM? |

---

## 3. Executive summary

**Growth is happening, but its quality is uneven.** Acquisition is stable overall, with mobile driving most traffic growth while desktop is under pressure from weaker SEO. The business is still bringing in users, but the traffic mix is shifting and becoming less balanced.

**The most valuable users are not the largest group.** The audience is young, mobile-led and slightly male-skewed, with the biggest volume from 18–24. The strongest commercial value comes from 25–34 — especially users who browse on mobile but convert on desktop.

**Desktop plays a disproportionate role in monetisation.** Mobile drives reach and discovery; desktop generates higher revenue per session. The business currently runs on a mobile-for-exploration, desktop-for-conversion pattern.

**The biggest revenue loss happens before checkout.** Conversion is improving overall, but leakage sits in the mid-funnel — Product Page to Cart, and Cart to Checkout. Checkout itself performs well.

**Retention is the strongest driver of long-term value.** Returning users generate much higher revenue and conversion than new users, and value rises sharply with repeat visits, recent return behaviour and repeat purchases.

> The challenge is not only to acquire more users, but to acquire the **right** users, convert them more efficiently, and retain them faster.

---

## 4. User acquisition

### Overall

| Metric | Value | Change |
|---|---|---|
| Sessions | 53K | +0.7% |
| Users | 24K | +0.6% |
| Pageviews | 268K | −0.8% |
| Bounce rate | 30.12% | +5.3% |
| % new sessions | 48.59% | −4.6% |
| % returning sessions | 51.41% | −2.8% |

Acquisition is stable but not strong in engagement depth: traffic and users grew, but lower pageviews suggest post-visit engagement needs work.

![Sessions trend](../../assets/images/marketing/laptop-sgn/laptop-sgn-07.png)
![Users trend](../../assets/images/marketing/laptop-sgn/laptop-sgn-08.png)

### Desktop by channel

Desktop growth is stagnating because **Google Organic — the largest traffic driver at 44% — declined 15% year on year.** Other channels grew, but not enough to offset the SEO drop. Traffic quality improved overall as bounce rate declined across most channels.

![Desktop channel breakdown](../../assets/images/marketing/laptop-sgn/laptop-sgn-09.png)

### Mobile by channel

Mobile growth is driven by strong SEO performance (**+30% YoY**), supported by significant expansion in paid and referral. Traffic volume is rising quickly, but the share of new users is declining across most channels — growth is increasingly driven by returning users. Improving bounce rates suggest better traffic quality.

![Mobile channel breakdown](../../assets/images/marketing/laptop-sgn/laptop-sgn-10.png)

### Which channels drive repeat visits?

Among paid channels, **Google Paid Search contributes the most to future organic revisits**, and the effect strengthens among users with more than 14 sessions — it brings in more qualified users who return organically later. Desktop users show a stronger organic return pattern than mobile. **Google Display shows the weakest contribution.**

![Returning visits via organic by channel](../../assets/images/marketing/laptop-sgn/laptop-sgn-11.png)
![Returning visits by session count](../../assets/images/marketing/laptop-sgn/laptop-sgn-12.png)

### Acquisition recommendations

- **SEO** — investigate the desktop decline by identifying affected pages, keywords and ranking losses; prioritise recovery on the highest-impact landing pages.
- **Performance marketing** — keep investing in Google Paid Search; strengthen mobile new-user prospecting so growth is not over-dependent on returning users; assess Google Display critically and reallocate if it keeps underperforming.
- **Analytics / BI** — review the decline in mobile new-user share; add value-based channel KPIs (revisit rate, bounce rate, new-user share) rather than sessions alone.
- **UX / CRO** — improve the mobile landing experience for first-time users.
- **Content** — tailor messaging by channel role: acquisition-led for mobile prospecting, recovery-focused for desktop SEO pages, high-intent for Paid Search.

---

## 5. User segmentation

### Who are our users?

- **Device:** mobile-led — 57.8% of sessions.
- **Gender:** relatively balanced, male slightly ahead at 52.3% vs 47.7%.
- **Age:** heavily concentrated in 18–24 (51.5%) and 25–34 (39.0%). Users 35+ are a small share.

![Sessions by gender](../../assets/images/marketing/laptop-sgn/laptop-sgn-13.png)
![Sessions by device](../../assets/images/marketing/laptop-sgn/laptop-sgn-14.png)
![Sessions by age group](../../assets/images/marketing/laptop-sgn/laptop-sgn-15.png)

Device preference varies by age: 18–24 skew more to desktop, while 25–34 are more mobile-oriented (using mobile 20% more often than desktop). 32% of mobile users are female — 5% higher than desktop.

![Device and gender](../../assets/images/marketing/laptop-sgn/laptop-sgn-16.png)
![Device and age group](../../assets/images/marketing/laptop-sgn/laptop-sgn-17.png)

### Product interest by gender

Brand preference differs clearly by gender:

- **Male** — over-index on **Samsung (40.7%)** and **Xiaomi (26.8%)**: performance, specs, broad product coverage.
- **Female** — stronger preference for **Apple (30.4%)** and **OPPO (19.8%)**: design, camera experience, lifestyle appeal.
- Samsung is the broadest brand across both genders but skews male; Realme is small and balanced.

![Item sales share by brand and gender](../../assets/images/marketing/laptop-sgn/laptop-sgn-18.png)

### Which segments create the most value?

**Desktop users generate higher revenue per session than mobile across all age groups.** The highest-value segment on both male and female sides is **25–34 on desktop**. This pattern is much weaker on mobile, where revenue per session is similar across ages.

> Revenue per session = AOV × CR = (Revenue / Transactions) × (Transactions / Sessions). Currency: VND.

![Male revenue per session by age and device](../../assets/images/marketing/laptop-sgn/laptop-sgn-19.png)
![Female revenue per session by age and device](../../assets/images/marketing/laptop-sgn/laptop-sgn-20.png)

### Which behaviours signal purchase intent?

- **Pageviews** — sessions with 1–2 pageviews dominate volume, but sessions with **3+ pageviews (~25% of total)** are more engaged and convert better. Conversion rises as pageviews increase.
- **Session duration** — short sessions dominate, but sessions of **8 seconds or more (~25% of total)** convert materially better.

![Pageviews vs conversion rate](../../assets/images/marketing/laptop-sgn/laptop-sgn-21.png)
![Share of sessions by pageview](../../assets/images/marketing/laptop-sgn/laptop-sgn-22.png)
![Session duration vs conversion rate](../../assets/images/marketing/laptop-sgn/laptop-sgn-23.png)
![Share of sessions by duration](../../assets/images/marketing/laptop-sgn/laptop-sgn-24.png)

### Segmentation recommendations

- **Merchandising** — tailor homepage visibility, collections and promotions for 18–24 (reach), 25–34 (value), mobile explorers, desktop converters, and the Samsung/Xiaomi vs Apple/OPPO brand groups.
- **Content** — performance and specs messaging for Samsung–Xiaomi audiences; design, camera and lifestyle for Apple–OPPO; premium and upgrade framing for 25–34.
- **Performance marketing** — prioritise 25–34 in acquisition and remarketing; route each segment to the right landing page by age, gender interest and device.
- **CRM** — retarget high-intent users (3+ pageviews, longer sessions) with follow-up based on profile, product interest and browsing depth.
- **UX / CRO** — optimise mobile for discovery and desktop for conversion.
- **Analytics** — track segment value using revenue per session, conversion rate, page depth and session duration.

---

## 6. Funnel performance

### Overall conversion

| Metric | Value | Change |
|---|---|---|
| Revenue | 3.26T VND | +7.3% |
| Transactions | 971 | +5.2% |
| Products sold | 1,446 | +4.5% |
| Conversion rate | 1.8% | +2.0% |
| Average order value | 3.4M VND | +2.0% |

### Where are the biggest drop-off points?

| Step | Users | Continue | Drop |
|---|---|---|---|
| Catalog page | 30,909 | 39.25% | 60.75% |
| Product page | 16,163 | **21.46%** | **78.54%** |
| Cart page | 3,665 | **27.47%** | **72.53%** |
| Checkout page | 1,020 | 63.46% | 36.54% |
| Purchase | 711 | — | — |

- **Top drop-off:** Product Page — only 21.46% of visitors move to cart. This is the main bottleneck.
- **Secondary:** Cart Page — only 27.47% continue to checkout.
- **Strongest step:** Checkout converts to purchase at 63.46%, so checkout is not the problem.

> The funnel problem is **mid-funnel friction, not end-funnel friction**. Users get lost before checkout — deciding whether to add to cart, and continuing from cart to checkout.

### Category exploration before purchase

Desktop shows higher co-browsing than mobile across all categories (**desktop average 32% vs mobile 18%**) — mobile behaviour is more targeted, desktop supports broader discovery.

- **Highest co-browsing:** Xả Hàng, Khuyến Mãi, Trade-in Phone, Apple — users explore across categories in these sessions.
- **Lowest co-browsing:** Phone and Laptop — more focused, goal-oriented browsing.
- **Middle:** Tablet and Accessories.

![Co-browsing sessions per product category](../../assets/images/marketing/laptop-sgn/laptop-sgn-26.png)

### Funnel recommendations

- **UX / CRO** — optimise Product Page and Cart Page: clearer CTA placement, checkout path, promo visibility, mobile browsing flow.
- **Content** — strengthen product decision support with clearer pricing, trust signals, and shipping/payment information to raise add-to-cart intent.
- **Merchandising** — run continuous A/B testing on product and cart pages; improve exploration in low co-browsing categories (Phone, Laptop) via related products and bundles; use high-browsing categories as discovery hubs.
- **Analytics** — track product-page depth, add-to-cart rate, cart-to-checkout rate and cross-category browsing rate as shared metrics.

---

## 7. User retention

### Returning vs new users

Returning users drive substantially higher value: **VND 2.3T revenue vs VND 0.9T** from new users, with a much stronger conversion rate (**2.50% vs 1.07%**).

![New vs returning user value](../../assets/images/marketing/laptop-sgn/laptop-sgn-27.png)

### Value by frequency and recency

User value increases with frequency: although 1-session users contribute the largest session volume, revenue per user rises steadily with session count, peaking above 5 sessions. The same pattern appears in recency — users returning after more prior sessions or on shorter revisit cycles generate much higher value per user.

![Value by session count](../../assets/images/marketing/laptop-sgn/laptop-sgn-28.png)
![Value by days since last session](../../assets/images/marketing/laptop-sgn/laptop-sgn-29.png)

### When should we re-engage?

- New users take about **14 days on average** to return after their first session.
- The biggest retention drop happens **between sessions 2 and 6**.
- Once users reach around **6 sessions**, they return roughly within a week — a strong signal of long-term retention.
- Revisit time decreases steadily as sessions accumulate, showing habit formation.

> Re-engagement should focus on accelerating repeat visits during the first few sessions.

![Average days since last session](../../assets/images/marketing/laptop-sgn/laptop-sgn-30.png)

### Long-term behaviours that matter

- **Frequency** — users with 6+ sessions are ~25% of total sessions and convert more strongly.
- **Recency** — around **80% of returning sessions happen within 2 days** of the previous visit, and these convert better than less recent visits.
- **Repeat purchase** — as users accumulate purchases, time to next purchase shortens.

![Frequency vs conversion rate](../../assets/images/marketing/laptop-sgn/laptop-sgn-31.png)
![Recency vs conversion rate](../../assets/images/marketing/laptop-sgn/laptop-sgn-33.png)
![Time to next purchase](../../assets/images/marketing/laptop-sgn/laptop-sgn-35.png)
![Repeat purchase rate](../../assets/images/marketing/laptop-sgn/laptop-sgn-36.png)

### RFM-based segmentation

![RFM segmentation matrix](../../assets/images/marketing/laptop-sgn/laptop-sgn-37.png)

| Segment | Definition | Action |
|---|---|---|
| **High value active** | 0–2 day recency, 2–5 or 6+ orders | Prioritise for revenue expansion: new arrivals, cross-sell, upsell, premium content |
| **Welcome** | Recent, 1 order | Onboarding, trust-building, best sellers, first-repeat incentive — avoid heavy discounting |
| **To win back** | 3–6 day recency, 2+ orders | Timely nudges: reminders, recently viewed, relevant category content |
| **Lapsing** | 7–14 day recency, 2+ orders | Act before reactivation probability declines |
| **Lapsed / churn risk** | >14 day recency, 2+ orders | Targeted win-back: tailored discounts, urgency, value offers |

The customer base is still concentrated in low-frequency, low-order-value segments — especially 1-order users — so the biggest growth opportunity is moving first-time buyers into repeat purchase behaviour.

### Retention recommendations

- **CRM / lifecycle** — prioritise early-stage retention (sessions 1–6); re-engage new users quickly after the first visit to shorten the return gap; personalise flows by RFM segment.
- **Performance marketing** — build re-engagement programmes by recency and frequency with different CPA levels, message urgency and offer strength; concentrate pressure right after the first and second visits.
- **Analytics / BI** — analyse repeat purchase behaviour continuously to find the best timing, audience and incentive, then feed insights back into CRM and media.
- **Management** — align on retention quality, for example increasing customer retention rate by 20% while improving repeat value from high-potential segments.

---

## 8. Consolidated recommendations

| Team | Action |
|---|---|
| **SEO** | Recover desktop SEO on high-impact pages and keywords to rebalance acquisition across devices |
| **Performance marketing** | Scale Google Paid Search; strengthen mobile new-user acquisition; reduce Google Display if revisit value stays low |
| **E-commerce / merchandising** | Build visibility around 25–34, mobile explorers, desktop converters and key brand groups; use high-browsing categories as discovery hubs |
| **Content / creative** | Personalise by segment, brand preference and device role |
| **UX / CRO** | Improve the mid-funnel — Product Page and Cart Page experience |
| **CRM / lifecycle** | Prioritise sessions 1–6 with onboarding, follow-up and RFM-based re-engagement |
| **Analytics / BI** | Track value-based KPIs, not traffic volume |
| **Management / growth** | Align teams on acquisition quality, mid-funnel conversion and retention of high-value users |

---

## Limits

- The window is 30 days during a seasonal demand peak, so year-round behaviour may differ.
- Revenue impact figures are expected ranges from the recommendation set, not measured outcomes.
- The analysis is observational — no experiment established that the recommended changes cause the modelled lift.