An electrical materials distributor receives every day Delivery Notes from dozens of different manufacturers: from the biggest, best-known brands to hundreds of minor brands. Different layouts, fragmented GTIN codes, cables arriving in reels or by the meter, promotional kits, returns from electricians and above all an internal master data set that for the same product can have 2-4 alternative codes.
The result is almost always the same: the goods receiving office becomes a bottleneck, manual data entry into the ERP is slow, errors propagate all the way to the sales counter and the B2B portal, and customers call to ask about the status of an item that the system has not yet recorded. In this guide we see how a delivery note automation software based on artificial intelligence solves the problem in a structural way for the electrical sector, which technical specificities a software must handle and what to evaluate before choosing one.
What is delivery note automation in electrical materials distribution
Delivery note automation in an electrical materials distributor is the process of automatic data extraction from incoming delivery notes and their direct entry into the ERP (and/or into the warehouse WMS) without manual data entry.
A modern delivery note automation software for the electrical sector does not just read the text of the document: it recognizes the manufacturer's item codes, matches them to the distributor's internal code (which often has 2-4 alternative codes for the same product), handles mixed units of measure (pieces, sets, full reels, linear meters of cable), distinguishes partial deliveries, and proposes recording on the corresponding purchase order. The operator only validates the exceptions.
Why the electrical materials sector has some of the highest delivery note volumes in Italian B2B
Italian electrical materials distributors operate in an operational context that amplifies the data entry problem to levels among the most severe in the market:
- Huge catalog: reference distributors manage between 300,000 and over 1 million active SKUs across cables, lighting, automation, home automation, photovoltaics, antennas
- Thousands of suppliers: from large global manufacturers to established Italian brands, to niche producers of specific components - each with its own delivery note layout
- Thousands of incoming delivery notes per week: central warehouses receive daily waves of deliveries, especially at the beginning of the week and after price list promotions
- “24h ready delivery” service: operational pressure requires recording incoming goods within the day, otherwise the sales counter and the B2B portal go out of sync
- Distributed branches: medium/large distributors have 5-50 regional branches with a sales counter, each with its own partial document flow
- Mixed deliveries: cables in full or split reels, single items, kits, promotional materials - on the same delivery note
- Frequent returns: the sector has a high percentage of returns from electricians, and return delivery notes have inverse logic to handle
- Manufacturer codes ≠ internal codes: the same Bticino switch can have the original manufacturer code, the abbreviated price list code, an internal distributor code and a commercial alias used at the counter
A back office team in a medium electrical distributor (€50-150M revenue, 5-10 branches) manually processes thousands of delivery notes per week. Even with experienced operators, the average time per delivery note with full verification ranges from 5 to 10 minutes. On a volume of 1,500 weekly delivery notes, that is 125-250 person-hours per week - the equivalent of 3-6 FTEs dedicated solely to delivery note data entry, not counting errors, disputes with manufacturers and delays in system availability.
For an electrical materials distributor the delivery note is not an administrative document: it is the document that opens availability to sell at the counter and on the B2B portal. Every hour of delay in entry is an hour of lost revenue.
The 5 technical specificities of electrical materials delivery notes
Compared to other sectors of industrial distribution, electrical materials delivery notes present five specificities that raise the bar for any automation software.
1. Structured but fragmented manufacturer codes
Manufacturers in the sector have structured GTIN/EAN codes. It sounds like good news, but in practice: the same product can be identified with 3-4 different codes on the same delivery note: GTIN code, manufacturer's commercial code, abbreviated price list code, promotional line code. An AI software must recognize them all and map them to the single internal code of the distributor.
2. Cables: reels, cut lengths, mixed units of measure
Cables (FG7, N1VV-K, H07V-K, RG58, fiber optic) are sold in full reels (e.g. 100m, 500m, 1000m), in cut lengths requested by the customer, or in predefined rolls. On the same delivery note rows in meters, in reels, in kg can appear. A classic OCR reads the number without distinguishing whether they are pieces or meters. A modern document AI must normalize the unit of measure in the context of the single item (a cable is almost always in meters or reels, a switch in pieces).
3. Return delivery notes
The electrical sector has a high return rate: electricians return what they have not used on the job site. Return delivery notes have inverse logic: the goods leave the customer's warehouse and re-enter the distributor's, and must be matched not to a purchase order but to an original sales delivery note. Few AI software solutions natively handle this logic.
4. Promotions and multiple price lists
Sector manufacturers launch monthly promotions (discounts on product families, special price list codes, kits with bonuses). Promotional delivery notes have rows with prices and codes different from the base price list. The software must recognize the promotional flag and track it in the ERP, otherwise the price list adjustment becomes an administrative nightmare.
5. Distributor's alternative internal codes
Almost all electrical materials distributors have item master data with multiple alternative codes: the “official” product code, an abbreviated code for the sales counter, possibly legacy codes from previous ERP migrations. The AI software must handle this many-to-one mapping on the fly: it recognizes the code on the supplier's delivery note, matches it to any of the alternative codes in the distributor's master data, and records on the “master” item.
A software that does not handle these 5 specificities seems to work in demo, but in production, on the real delivery note of Monday morning, it triggers exceptions on 30-40% of the rows - and the promise of automation breaks.
How a delivery note automation software works for electrical materials
A modern delivery note automation software, like TypeLens for delivery notes, follows five phases, adapted to the specificities of the electrical sector:
Between point 4 and point 5 there is the human validation phase: the operator sees a screen with the original delivery note on the left and the extracted data on the right, confirms with a click or corrects the low-confidence rows. It is a human-in-the-loop approach that combines machine speed and human control where it really matters.
Traditional OCR vs document AI for electrical materials delivery notes
Many electrical distributors have already tried classic OCR and came away disappointed. The reason is structural: traditional OCR requires a template for each supplier, and in a company with hundreds of active manufacturers this means hundreds of configurations to maintain (plus all those that break at the first layout change).
| Aspect | Traditional OCR | Modern document AI (TypeLens) |
|---|---|---|
| Initial setup | One template for each supplier (Bticino, Vimar, Gewiss…) | Works without templates, on any manufacturer layout |
| Maintenance | Breaks when a manufacturer updates its delivery note | Adapts automatically |
| GTIN/EAN codes | Extracts the literal value | Recognizes GTIN, matches to the distributor's internal code |
| Alternative codes | A single code → a single match | Many-to-one mapping on item master data |
| Cables and reels | Does not distinguish pieces/meters/reels | Normalizes the unit of measure by item category |
| Promotions | Confuses promo code with base code | Recognizes the promotional flag |
| Return delivery notes | Not handled | Inverse logic natively supported |
| Italian documents | To be configured from scratch | Native handling of delivery notes, XML invoices, PEC |
| Typical go-live time | Minimum 2-3 months + system integrator | 3-4 weeks on mature solutions |
In summary: for the electrical sector, pure OCR is structurally insufficient. The heterogeneity of manufacturers, codes and units of measure triggers exceptions in 30%+ of cases, nullifying the automation gain. We explore the differences and achievable precision levels in this OCR accuracy guide.
Features to look for in a delivery note software for the electrical sector
For an Italian electrical materials distributor, a delivery note automation software is evaluable on eight specific criteria:
- Native integration with the ERP in use: TeamSystem (Enterprise, Alyante, Polyedro), Zucchetti (Ad Hoc, Mago.net, Infinity), Dynamics 365 Business Central, Odoo, or custom ERP via API. For the large companies in the sector that use SAP enterprise, integration is to be evaluated case by case.
- Integration with the warehouse WMS (when present): many electrical distributors have a dedicated WMS for the physical management of goods. The delivery note data must land in both systems consistently.
- Management of alternative codes in the item master data: the software must handle many-to-one mapping without the distributor having to “flatten” its own master data.
- Recognition of the codes of the main manufacturers in the sector: Bticino, Vimar, Gewiss, Legrand, ABB, Schneider Electric, Philips, Osram, Prysmian, Hager, Eaton - the system must learn the patterns of each (ideally from the customer's historical delivery notes).
- Handling of mixed units of measure: pieces, sets, kits, reels, linear meters, kg, on the same document.
- Handling of return delivery notes and matching to sales delivery notes: inverse logic natively supported, not as an “add-on module”.
- Realistic go-live time: 3-4 weeks to be operational on the first delivery note flows, not a 4-6 month project.
- Italian team with sector onboarding: understanding of the dynamics of the sales counter, of returns from electricians, of manufacturer promotions.
The first 4 are the real filters: without them automation stops at the demo and does not withstand the impact of real delivery notes in production.
How TypeLens addresses these specificities
TypeLens is the document automation software of Mastranet AI (headquartered in Dalmine, BG). On the delivery note case it works on the flow described above: it receives documents from any channel (supplier email, PEC, manufacturer portal, upload, API), extracts data even from low-quality scans, recognizes the GTIN codes of manufacturers and matches them to the distributor's internal catalog - including mapping on alternative codes. It normalizes the unit of measure by item category, distinguishes promotional rows, handles partial deliveries and return delivery notes, and writes the data directly into the ERP and - when available - into the WMS via native integration.
Sector specificities (GTIN codes, alternative codes, reels/cut lengths, return delivery notes, promotions) are modeled on the customer's historical document flow during onboarding, not through templates for each supplier.
ERP and WMS integrations for Italian electrical distributors
TypeLens is natively integrated with:
- TeamSystem: Enterprise, Alyante, Polyedro (the most widespread ERPs among medium Italian electrical distributors)
- Zucchetti : Ad Hoc, Mago.net, Infinity (present in many companies in the sector)
- Microsoft Dynamics 365 Business Central : the typical choice of sector companies between €50M and €300M
- Odoo : adopted by rapidly growing distributors and new entrants
- Documented REST APIs : for integration with dedicated WMS (e.g. custom or niche solutions), legacy ERPs, or enterprise ERPs such as SAP in the case of larger distributors
For distributors operating with a WMS separate from the ERP (a very common model in the electrical sector, where the physical warehouse has peculiar complexity), TypeLens can write the delivery note data into both systems consistently, with automatic reconciliation.
Stop typing. Let the delivery notes of your electrical warehouse load themselves.
In 30 minutes we look together at a real delivery note from your flow - Bticino, Vimar, Gewiss, ABB, Schneider, Prysmian or a minor brand - and we show you TypeLens in action. No slides, just a live demo on your document flow.
Book a demo →Timing, costs, ROI for an electrical materials distributor
Typical implementation time: 3 weeks from project start to operational go-live on the first delivery note flows.
- Week 1: connection with the ERP (and any WMS), mapping of the alternative codes of the item master data, access to the incoming document flow
- Week 2: AI training on the set of historical delivery notes (ideally with samples from the main manufacturers and about a dozen of the most frequent brands)
- Week 3: operator-assisted validation, refinement on residual exceptions, go to production
Cost model: TypeLens applies a SaaS model sized on document volumes. For an electrical materials distributor with realistic delivery note volumes (1,000-10,000 weekly between central warehouse and branches), the cost typically pays back in 6-12 months considering only the time recovered by the back office. Specific conditions are evaluated case by case during the demo phase.
Realistic ROI: for a back office team of 4-8 FTEs dedicated to delivery note data entry (including branches), automation typically frees 2-4 FTEs, reallocable to value-added activities - supplier management, quality control on returns, sales counter support, B2B portal development. Residual error drops by 80-90%, reducing disputes with manufacturers (invoice adjustments, credits, credit notes) and administrative hours lost fixing them.
The realistic goal is not to zero out human intervention: it is to shift it from typing to supervision. The operator stops copying Bticino and Schneider codes, and starts checking exceptions.