MarTech · 27 พ.ค. 2026 · 6 min read

Data Stack สำหรับ B2B ในไทย ปี 2026 — CRM, CDP, Reverse ETL

คู่มือ marketing data stack สำหรับ B2B ในไทย ปี 2026 — CRM (HubSpot/Salesforce), CDP (Segment/RudderStack), Reverse ETL (Hightouch/Census), warehouse (BigQuery/Snowflake), activation

TL;DR: B2B data stack ปี 2026 = Warehouse-first + CDP-as-pipeline + Reverse ETL activation. Modern stack: BigQuery/Snowflake เป็น single source of truth, ingest จาก HubSpot/Salesforce + Pixel events + 3rd-party tools, push activation data กลับไป Meta/Google/LinkedIn ผ่าน Reverse ETL (Hightouch/Census). คู่มือนี้ครอบคลุม: stack architecture, CRM selection (HubSpot vs Salesforce), CDP role (Segment/RudderStack), warehouse setup, Reverse ETL use cases, identity resolution, attribution stack, และ TH-specific considerations

ทำไม B2B ต้องการ Data Stack ที่ซับซ้อนกว่า B2C

B2B-specific challenges:
– Long sales cycle (30-180 วัน) — ต้อง track ทุก touchpoint
– Multi-stakeholder (CEO + Manager + Procurement) — multiple identities per company
– Account-Based Marketing (ABM) — target company, not user
– High-value transactions — accuracy matters more than scale
– Offline events critical (demos, meetings, calls) — ต้อง import กลับ
– Sales + Marketing tightly coupled — share data needed

B2C ไม่ต้องการแบบนี้ — short cycle, fast ROAS, less identity stitching needed

Modern Data Stack — Architecture

Data Stack สำหรับ B2B ในไทย ปี 2026 — CRM, CDP, Reverse ETL — แผนภาพที่ 1

Layer Responsibilities

Layer	Role	Tools
Sources	Generate raw events/data	Pixel, CRM, App, Tools
CDP / EL	Ingest + standardize	Segment, RudderStack, Fivetran
Warehouse	Single source of truth	BigQuery, Snowflake, Redshift
Transform	Build models, identity stitch	dbt, SQL views
Reverse ETL	Send back to tools	Hightouch, Census
Activation	Use data in ad/CRM/BI	Meta CAPI, CRM, LinkedIn

CRM — Foundation Choice

HubSpot

Best for: SMB-Mid B2B, marketing-led companies, ≤500 employees

✅ Pros:
– Marketing + Sales + Service in one
– Free tier viable for small
– API rich, integrations many
– Workflow automation friendly
– TH user community growing

❌ Cons:
– Pricing scales steeply (Marketing Hub Pro 800+ USD/month)
– Less customizable for complex sales process
– Reporting limitations at scale

TH pricing 2026: Starter 18 USD/seat, Pro 800/month, Enterprise 3,600/month

Salesforce

Best for: Mid-Enterprise, complex sales, ≥100 sales seats

✅ Pros:
– Most customizable
– AppExchange ecosystem largest
– Enterprise-grade governance
– Marketing Cloud + Pardot integration

❌ Cons:
– Steep learning curve
– Expensive consultants required
– Slower implementation (3-12 months)
– License cost 25-300 USD/seat/month

TH partners: Bluebik, Accenture, Salesforce TH direct

Pipedrive / Zoho

Best for: Sales-led SMB, lighter marketing needs
– Cheaper alternative
– Less marketing automation

TH-specific considerations

LINE OA integration — HubSpot via Zapier/Make, Salesforce via custom
Thai language support — both ดีพอใช้
PDPA compliance — both มี data residency/retention features

CDP — Customer Data Platform

Why CDP exists

Old model: data ใน marketing tool → ใช้ใน marketing tool. Modern: data → warehouse → ใช้ทุกที่

CDP Tools

Segment (Twilio)

Industry standard
300+ integrations
1,000 MTU free tier
Paid: starts $120/month
Best for: SaaS, e-com, B2B mid-market

RudderStack

Open-source friendly
Self-host option
Cheaper at scale
Good for: dev-led teams

CustomerIO / Bloomreach (light CDP)

Marketing-focused
Less technical

CDP Functions

Event collection — Pixel, mobile SDK, server-side
Identity resolution — match anonymous → known user
Data standardization — common schema across sources
Audience building — segment customers
Destination sync — push to ad platforms, email, etc.

When you don’t need CDP

< 100k MTU
Use only 1-2 marketing tools
All data already in one tool
→ Skip CDP, use direct integrations + warehouse

Data Warehouse

BigQuery (Google Cloud)

✅ Free tier 10GB storage + 1TB query/month
✅ Easy GA4 integration (native)
✅ Good Pricing for variable workload
❌ Cost can spike if poorly indexed

Best for: Most TH B2B startups + mid-market, especially if using Google Workspace

Snowflake

Pay-per-use compute
Best for: Large data, multi-cloud, enterprise
Expensive for small workload

Other

Redshift (AWS native)
Databricks (analytics + ML)
ClickHouse (open-source, fast)

TH consideration

BigQuery มี Singapore region — low latency
Snowflake มี AWS Singapore
Data residency requirement (PDPA) — verify with vendor

Reverse ETL — Activation Layer

What is Reverse ETL

ส่ง data จาก warehouse → operational tools (CRM, ad platforms, support)

Use Cases B2B

1. Lead Scoring → CRM

Calculate lead score ใน warehouse → push back to HubSpot/Salesforce property → trigger sales workflow

2. Custom Audience → Meta/Google/LinkedIn

Sync high-value customer list → Custom Audience
Sync churn-risk customers → win-back campaign
Sync stage-based segments → ABM ads

3. Account Enrichment → CRM

Append firmographic data (industry, revenue, employees)
Append intent signals (visited pricing page 3x)
Append technographic (uses Salesforce vs HubSpot)

4. Product Qualified Lead (PQL) → Sales

User reached “aha moment” feature
→ Push notification to AE in Slack/CRM

Tools

Tool	Best for	Pricing
Hightouch	Most popular, easy UI	Free 5M syncs, $350+/month paid
Census	Enterprise, robust	Custom
RudderStack	Open-source option	Free self-host
Polytomic	Visual modeling	$300+/month

Identity Resolution

Why important สำหรับ B2B

User journeys span:
– Anonymous web visit (cookie)
– Email subscribe (email ID)
– Login (user ID)
– Phone call (phone)
– LINE chat (LINE ID)
– In-person event (badge scan)

ต้องรู้ว่า “นี่คือคนเดียวกัน” ตลอด journey

Resolution Patterns

Deterministic — exact match (email, phone, login ID)
Probabilistic — fingerprint matching (IP + device + behavior)
Hybrid — both layered

Tools

CDP (Segment, RudderStack) built-in
Tealium (enterprise)
Custom in warehouse (dbt models)

TH-specific

LINE ID can serve as primary persistent identifier
Phone-based identity common (high mobile-first)
Hash PII before sending to ad platforms (PDPA + ad platform req)

Account-Based Marketing (ABM) Data Stack

ABM-specific tools

6sense / Demandbase — intent data + ABM platform
Clearbit / ZoomInfo — firmographic enrichment
Bombora — 3rd-party intent
G2 Buyer Intent — review-platform intent

Data Flow

Anonymous web traffic
  → IP reverse lookup (company)
  → Firmographic enrichment (Clearbit)
  → Intent score (6sense)
  → Push to LinkedIn Matched Audience
  → Sales alert if in ICP + high intent

TH ABM challenges

Smaller B2B market = less 3rd-party intent data
LinkedIn penetration lower than US
Many companies don’t have web presence

Email + Marketing Automation Integration

B2B Common

Marketo — enterprise, $1k+/month
Pardot — Salesforce-native
HubSpot Marketing Hub — bundled
ActiveCampaign — mid-market
Customer.io — product-led

Integration with warehouse

Email engagement → warehouse → lead score
Lead score → CRM → sales workflow
Send conversion data → ad platform CAPI for optimization

Privacy + Compliance Layer

Consent Management

OneTrust / Cookiebot / Termly — consent UI
Consent Mode v2 (Google) — pass consent state
Tealium Consent Hub

Server-side Tracking

GTM Server Container
Reduce client exposure
Compliant Conv API
Better data fidelity

PDPA-specific

Document data flow
Retention policy (e.g. delete after 2 years inactive)
Right to access + delete
DPO (Data Protection Officer) if processing > 50k Thai residents

Tool Stack Examples by Company Stage

Startup (≤20 employees, <100k MTU)

CRM: HubSpot Starter
Ad tracking: Meta Pixel + GA4
Email: Mailchimp / ActiveCampaign
Warehouse: BigQuery free tier
No CDP, no Reverse ETL — direct integrations
Cost: $0-200/month

Growth (20-100 employees, 100k-1M MTU)

CRM: HubSpot Pro หรือ Salesforce Essentials
CDP: Segment Team
Warehouse: BigQuery / Snowflake (small)
Reverse ETL: Hightouch starter
Email: HubSpot / Customer.io
Cost: $1.5-5k/month

Scale (100-500 employees, 1M+ MTU)

CRM: Salesforce Enterprise + Pardot/Marketo
CDP: Segment Business
Warehouse: Snowflake / BigQuery production
Reverse ETL: Hightouch Pro / Census
ABM: 6sense + LinkedIn integration
BI: Looker / Tableau
Cost: $20-100k/month

Common Pitfalls

Buy CDP before need it — < 100k MTU = overkill
Treat warehouse as backup instead of source of truth
Skip identity resolution → fragmented user view
No reverse ETL → data in warehouse but tools don’t get it
Tools without ownership — buy MarTech, no one runs it
CRM cluttered with bad data — manual entry, no enrichment
Forget offline events — demos, calls, in-person not tracked back
No data dictionary — engineers + marketers use diff names
One-time setup, no maintenance — stack rots in 6 months
Optimize for features, not outcome — choose tool by checkbox not need

Quick Start Checklist

Phase 1 (Month 1-2): Foundation

[ ] Choose CRM (HubSpot Pro = SMB default)
[ ] Setup GA4 + GTM Server-side
[ ] BigQuery account + GA4 integration
[ ] Define key events + data dictionary
[ ] Document customer lifecycle stages

Phase 2 (Month 2-3): Pipeline

[ ] Setup CDP (Segment) if MTU > 100k
[ ] Integrate top 5 marketing tools to warehouse
[ ] Build identity resolution model
[ ] First Reverse ETL sync (Custom Audience to Meta)

Phase 3 (Month 4-6): Activation

[ ] Lead scoring model in warehouse
[ ] Sync to CRM → trigger sales workflow
[ ] Offline conversion upload to Meta/Google
[ ] ABM segment activation
[ ] BI dashboard for revenue attribution

Phase 4 (Month 6+): Optimization

[ ] MMM model (see Attribution guide)
[ ] Incrementality testing
[ ] Quarterly stack review

บทความที่เกี่ยวข้อง

MarTech Analytics & Tracking Guide — GA4 + GTM Server foundation
Attribution & Modeling Guide — MMM + Incrementality
Data Stack for E-commerce — companion for B2C/D2C
Meta Ads สำหรับ B2B — Meta + CAPI integration

TL;DR ย้ำ: B2B data stack 2026 = Warehouse-first + CDP-as-pipeline + Reverse ETL activation. Start: HubSpot/Salesforce + GA4 + BigQuery. Add Segment + Hightouch when MTU > 100k. Identity resolution + offline conversion = unlock long-cycle B2B attribution. PDPA-compliant ตั้งแต่ day 1

Data Stack สำหรับ B2B ในไทย ปี 2026 — CRM, CDP, Reverse ETL

ทำไม B2B ต้องการ Data Stack ที่ซับซ้อนกว่า B2C

Modern Data Stack — Architecture

Layer Responsibilities

CRM — Foundation Choice

HubSpot

Salesforce

Pipedrive / Zoho

TH-specific considerations

CDP — Customer Data Platform

Why CDP exists

CDP Tools

Segment (Twilio)

RudderStack

CustomerIO / Bloomreach (light CDP)

CDP Functions

When you don’t need CDP

Data Warehouse

BigQuery (Google Cloud)

Snowflake

Other

TH consideration

Reverse ETL — Activation Layer

What is Reverse ETL

Use Cases B2B

1. Lead Scoring → CRM

2. Custom Audience → Meta/Google/LinkedIn

3. Account Enrichment → CRM

4. Product Qualified Lead (PQL) → Sales

Tools

Identity Resolution

Why important สำหรับ B2B

Resolution Patterns

Tools

TH-specific

Account-Based Marketing (ABM) Data Stack

ABM-specific tools

Data Flow

TH ABM challenges

Email + Marketing Automation Integration

B2B Common

Integration with warehouse

Privacy + Compliance Layer

Consent Management

Server-side Tracking

PDPA-specific

Tool Stack Examples by Company Stage

Startup (≤20 employees, <100k MTU)

Growth (20-100 employees, 100k-1M MTU)

Scale (100-500 employees, 1M+ MTU)

Common Pitfalls

Quick Start Checklist

Phase 1 (Month 1-2): Foundation

Phase 2 (Month 2-3): Pipeline

Phase 3 (Month 4-6): Activation

Phase 4 (Month 6+): Optimization

บทความที่เกี่ยวข้อง

อ่านเพิ่มเติม — Pillar Guides ที่เกี่ยวข้อง