Digital transformation by industry: priorities for 12 sectors

In short: Digital transformation by industry means choosing what to digitise first based on where your sector typically loses money and time. A factory starts with raw-material stock and work orders, a retail chain with stock across channels, a clinic with medical records and appointments. The map below covers 12 sectors, each with two or three first moves and one AI use worth trying, in the same format so they are easy to compare.

Why the starting point depends on the industry

Every business has one main flow that drives its margins: materials in a factory, goods at a distributor, patients at a clinic, billable hours at a consultancy. Where that flow gets stuck is where digitising pays off first, because the improvement shows up directly in the P&L.

The overall method is the same everywhere (assess, plan, implement in parts, measure), as explained in our digital transformation guide. What differs is the starting point. A 40-person garment workshop and a 40-person language school will not begin with the same software. Treat this map as a starting hypothesis, then confirm it with your own assessment.

Priority map for 12 industries

Industry	Typical pain	Digitise first	AI use worth trying
Manufacturing	Unknown material stock, late product costing	Material stock tied to bills of materials; work orders and output per shift; batch traceability	Camera-based visual defect detection
Distribution, wholesale	Slow quotes, mismatched receivables, several warehouses	Customer price lists; order to shipment to receivable flow; field sales app	Draft quotes from customer requests
Retail	Stock mismatch between stores and online	Omnichannel sales with central stock; e-invoicing; loyalty data	Chatbot on messaging channels
Logistics	Dispatch by phone, paper delivery notes	Shipment orders and status; electronic proof of delivery; cost per trip	Reading and extracting transport documents
Construction	Site material losses, slow payment files	Materials per project; site diary with photos; acceptance and payment files	Turn daily site logs into weekly reports
Technical services	Paper job sheets, no equipment history	Mobile job sheets; equipment records by serial number; preventive maintenance schedule	Assistant that searches technical manuals
Healthcare (clinics, pharmacies)	Paper records, messy appointments, near-expiry stock	Electronic medical records; appointments and recall reminders; drug stock by batch and expiry	Draft record summaries for doctors to review
Education, training	Enrolment from many channels, hard-to-track fees	Enrolment pipeline; classes and tuition; online learning platform	Assistant answering parents and learners
Hospitality, F&B	Ingredient losses, bookings from many channels	POS linked to kitchen; recipe costing and stock; central booking management	Replies to reviews and booking messages
Agriculture	Paper farm logs, weak traceability	Digital farm logs; batch codes for traceability; purchasing and payables per grower	Pest and disease detection from photos
Financial services	Paper-heavy client files, strict compliance	Digital client files and approval flow; automatic payment reminders; access logs	Reading and cross-checking documents in a file
Professional services	Hours not tracked, scattered documents	Projects and timesheets; permissioned document store; clients and contracts	Search over internal knowledge

Sector notes

Manufacturing

Factories lose most money when nobody knows exactly how much material is left or what each batch really cost. Start with material stock linked to bills of materials, then work orders and output reported by shift leaders on the floor. Traceability comes once stock and production data agree. Garment makers usually start with order management by style, colour and size, since one order can have hundreds of variants. Visual defect detection is worth piloting only with enough defect images and a human inspector still in the loop.

Distribution and wholesale

Distributors live on quote speed and accurate receivables. Put items and customer-specific price lists into the system first, then link orders, shipments and receivables into one flow, then give field sales an ordering app. AI-drafted quotes from customer emails make sense once prices are in the system. See AI in operations.

Retail

Retail starts with stock: selling online an item the store just sold out loses the order and some trust. Omnichannel sales software with a central stock view fixes that. E-invoicing is mandatory too: according to VCCI (2025), from 1 June 2025 Decree 70/2025/ND-CP requires household and individual businesses with annual revenue of VND 1 billion or more to issue e-invoices generated from cash registers. Loyalty data and a chat assistant for price, size and order questions come next. See AI in sales and marketing.

Logistics

Small carriers often dispatch through phone calls and group chats, so every "where is my shipment?" means calling a driver. Digitise shipment orders and status first, with electronic proof of delivery (photos and the recipient's signature on the driver's phone). Track cost per trip next to see which routes make money. For a fleet of a few dozen trucks, AI that reads waybills and invoices is more practical than route optimisation.

Construction

Contractors lose money on site materials and slow payment claims. Track materials per project (request, purchase, receipt, use) and keep a photo site diary filled in by the site manager. Once diaries and quantities are digital, acceptance and payment files come together much faster. Summarising daily logs into a weekly owner report is an easy AI pilot.

Technical services

Installation and maintenance firms run on technicians' schedules. Mobile job sheets (accept, check in, before and after photos, customer signature) come first, then equipment records by serial number with warranty and repair history, plus preventive maintenance schedules. An assistant that answers from manufacturer manuals and repair history helps new technicians find answers without calling senior staff.

Healthcare

For clinics and pharmacies, this means records, appointments and drug stock in one system. Electronic medical records are also a legal requirement: according to the Can Tho legal information portal, Circular 13/2025/TT-BYT (in force from 21 July 2025) required hospitals to finish by 30 September 2025 and other healthcare facilities by 31 December 2026. Health data is generally treated as sensitive personal data, so keep AI at the drafting stage with a doctor reviewing, and check the sensitive-data list in the implementing decree before any patient data touches an AI tool.

Education and training

Training centres and private schools recruit through many channels, and leads slip between advisers. An enrolment pipeline that gathers every channel comes first, then classes, attendance and tuition, then an online learning platform if the model needs it. An assistant answering parents about schedules and fees is a good pilot. AI grading should stay under teacher review, with extra care for minors' data.

Hospitality and F&B

Restaurants lose money on ingredients: without recipe quantities nobody knows how much the kitchen wastes. Start with POS orders going straight to the kitchen, then recipe costing tied to ingredient stock. Small hotels should centralise bookings from all channels to avoid double bookings. Cash-register e-invoicing under Decree 70/2025/ND-CP also applies to food businesses above the revenue threshold. AI replies to reviews and booking messages are low-risk to try.

Agriculture

Farms and cooperatives are often asked by buyers, especially exporters, where a batch came from. Digital farm logs (fertiliser, spraying, harvest dates) and batch codes answer that directly. Firms buying from many growers should then track purchases and payables per grower. Pest detection from leaf photos is worth trying, but an agronomist should confirm before treatment.

Financial services

Small financial service firms (insurance agencies, brokers, consumer finance) handle document-heavy client files under strict compliance. Digitise client files and approval flows with clear permissions and access logs first, then automatic payment reminders. AI that reads and cross-checks documents is a sensible pilot. Any AI that decides on applications affecting customers' interests should be checked against the risk classification in Vietnam's AI law, with legal advice.

Professional services

Accounting, legal, architecture, consulting firms and agencies sell expert hours, so the first thing to measure is which project those hours go to. Projects with timesheets come first, then a document store permissioned by client, then contract management. Internal knowledge search and drafting help right away, as long as client information stays out of public AI tools (see AI governance and data security).

What every industry needs

Three foundations appear in almost every first phase: accounting and e-invoicing in software connected to sales or order management; one customer list instead of contacts scattered across salespeople's phones; and central control of accounts, permissions and backups, with a named owner. Our guide to business management software covers how to choose tools for these.

Rules to know (checked November 2026)

Regulation	Who it affects	What it covers	In force from
Law on Digital Transformation (No. 148/2025/QH15)	All sectors	General framework, support for SMEs, cooperatives and household businesses	1 July 2026
Law on Personal Data Protection (No. 91/2025/QH15)	Anyone holding customer or staff data; heavier for healthcare, education, finance	Duties when collecting and processing personal data	1 January 2026
Law on Artificial Intelligence (No. 134/2025/QH15)	Firms developing or deploying AI systems	Risk classification, transparency for AI-generated content	1 March 2026
Circular 13/2025/TT-BYT	Healthcare	Electronic medical records	21 July 2025
Decree 70/2025/ND-CP	Retail, F&B and other household businesses above the threshold	Cash-register e-invoices	1 June 2025

Sources: law number from LuatVietnam; effective date and SME support from the Ha Tinh provincial portal. Under the AI law, according to the Can Tho legal information portal, AI systems already running before it took effect have 18 months to comply in healthcare, education and finance, and 12 months elsewhere. Always check the original text or ask a lawyer before acting.

Using the map well

Take your sector's row and ask department heads whether the listed pain really is your worst bottleneck. Pick one item from "digitise first" and record today's number (quote time, stock discrepancies, days to prepare a payment claim). Pilot it in one team, branch or project for two to three months, following the digital transformation roadmap. Add the AI use only once the data from the earlier columns is in the system. That is why AI sits in the last column.

Do not copy another company's model just because it shares your industry label: a make-to-order workshop and a mass-production plant need very different software, and a single restaurant does not need a 20-store chain's system. And if you are restructuring or merging, wait for processes to settle before configuring anything.

FAQ

What does digital transformation mean in practice? Using software and data to replace manual work so the business runs faster, with fewer errors and decisions based on numbers. The first step differs by industry because the bottleneck differs.

What is the Law on Digital Transformation? Law No. 148/2025/QH15, passed in December 2025 and in force from 1 July 2026, replacing the 2006 Law on Information Technology. It sets the general framework and includes support policies for SMEs; check the original text and implementing decrees for specific duties.

My industry is not on the list. Where do I start? Find the sector whose main flow looks most like yours (goods, customers, hours or projects), start there, then test it against your own assessment.

Where should HR and marketing start? Marketing starts by putting customer data in one place so you can see which channel produces orders. HR starts with attendance, payroll and staff records. See AI in HR.

---

HAT Solutions assesses your current state with your industry in mind and helps you choose the two or three things to digitise first, with measurable KPIs. Tell us your industry and the process that hurts most, and we will suggest a starting point. Read next: Where should digital transformation start?