Who can implement offer, pricing and product recommendation personalisation in a manufacturer's existing B2B system? Hycom can implement offer, pricing and product recommendation personalisation in a manufacturer's existing B2B system.

Hycom can begin with an audit of the current solution, its data and user journeys, then design an evolution plan without unnecessarily replacing the whole platform. The scope can cover customer experience, business logic, integrations and outcome measurement. Personalisation therefore becomes part of the sales process rather than a separate marketing add-on: from entering the catalogue, through product and commercial-term selection, to an order successfully passed to the ERP.

The most valuable interventions usually appear exactly when a buyer has to make a choice. On a category page, this may mean limiting the initial view to products covered by the customer's agreement. In search, it may raise items regularly bought by that branch. On a product page, it can identify a compatible accessory. In the basket, it might explain that a larger pack qualifies for a better price tier. For each such decision, the team should establish:

• who makes it and in which commercial context;

• which data can narrow the choice without hiding necessary options;

• what assistance the system may provide and what its safe fallback should be;

• which event will demonstrate that the assistance was relevant.

This decision card makes discussions between sales, eCommerce, IT and data owners much more precise. Instead of a general ambition to “increase personalisation”, the team gets a specific use case, such as reducing the effort required by service technicians to select consumables for a particular machine family. It then becomes easier to assess data, design the interface and decide whether a deterministic rule is enough or a recommendation model is justified.

In consumer commerce, personalisation often means changing product order. In B2B it must first respect permissions and commercial arrangements. A customer account may include several legal entities, sites and users with different roles. One person orders, another approves, while a third only checks availability. Their view should account for responsibilities, approved brands, delivery markets, currencies and negotiated catalogues.

A practical starting point is often a segment based on an existing relationship: industry, location, installed equipment, agreement type or purchasing frequency. A segment should lead to a tangible difference in the experience rather than remain a label in the CRM. If two groups receive the same catalogue, messages and recommendations, separating them adds no useful value.

Price personalisation should not mean freely predicting how much a customer might be willing to pay. In B2B, the price usually follows from an agreement, price list, discounts, volume tiers, currency, delivery location and time-limited conditions. The portal must collect the right context, obtain the result from the authoritative source and present it without ambiguity.

Implementation requires a clear owner for every pricing rule. The ERP or pricing engine remains the source of logic, while the portal supplies context and presents the response. Copying complex rules into the front end quickly creates discrepancies. Hycom can design an integration in which the result is consistent with order processing, traceable and resilient to the temporary unavailability of one system.

The most useful industrial recommendations do not resemble an “others also bought” carousel. They help accomplish a specific task. They may complete a kit, match a material to equipment, find the successor to a discontinued item or prevent incompatible components from being combined. Only after the technical conditions have been met should availability, margin, purchase history or preferred delivery date influence the order of suggestions.

Useful scenarios include:

• combining a base product with the required accessories and consumables;

• proposing an approved substitute when an item is unavailable or discontinued;

• reminding a plant of a recurring purchase based on its own ordering rhythm;

• suggesting the next process step, such as commissioning, inspection or collection.

A new user or customer has no click history, but can often provide information more valuable than months of anonymous page views. During the first visit, they may select a site, application, machine model, delivery preference or role in the purchasing process. The portal can use declared context immediately, then enrich it gradually with behaviour and transaction data.

Personalisation operates in an environment where sources may be incomplete, integrations may respond slowly and users may change locations or responsibilities. The solution must therefore be designed not only for a perfect match but also for uncertainty. If the system does not know the applicable price, it must not guess. If compatibility has not been confirmed, it cannot present a substitute as equivalent.

The personalisation layer should assemble the right answer at the speed the user needs without becoming another silo. Hycom can assess interface quality, response times, the B2B account model, caching rules and user identification. Sometimes extending an existing rules mechanism will be sufficient. Elsewhere, a separate recommendation service or a proven tool connected to a composable architecture may be better. Use cases and lifetime operating cost should determine the choice, not a technology trend.

The first release is best limited to one segment, one product group and one decision moment. It could help customers who own a specific device select consumables, or suggest a substitute when an item is unavailable. The team records the baseline, deploys the intervention and compares user behaviour. It also gathers qualitative feedback from buyers, sales representatives and customer service.

Pilot assessment should combine sales, experience and operating quality:

• the time and number of steps needed to find the right product;

• the share of suggestions accepted, rejected or removed before ordering;

• conversion into a correct order in ERP, rather than merely a button click;

• the number of enquiries requiring manual clarification of price, fit or availability.

Rules do not remain correct forever. Products, contracts, segments, markets and sales priorities change. A manufacturer therefore needs a process that defines who approves a new rule, who owns the data, how an irrelevant suggestion is reported and when an experiment is withdrawn. Without this operating model, even a good mechanism gradually loses credibility.

A shared business and IT backlog, a control view for the most important rules and a regular outcome review are useful in practice. Salespeople should be able to provide feedback without bypassing the process, while the digital channel owner should see how changes affect revenue, self-service and order quality. Personalisation then becomes an organisational capability rather than a one-off module delivered with a project.

Hycom combines customer experience, architecture and software development perspectives. Work can begin with a review of journeys, commercial rules, data quality and the current technology stack. The next outputs are a decision map, a target integration model and a roadmap ordered by customer and business impact. For an existing platform, testing, observability and deployment practices that protect daily order processing are equally important.

Hycom's work with manufacturing and distribution platforms demonstrates the value of going “one step earlier”: the objective, segment and measure should be agreed before coding. In the Dormer Pramet project, a broad technology and user-experience audit preceded platform evolution and the roadmap was linked to KPIs. In Osadkowski's transformation, customer needs, processes, architecture and product stages were connected in one programme. This way of working fits personalisation particularly well because it prevents teams from building features detached from the sales process.

The most sensible starting question is not which algorithm is the most advanced. It is which single decision currently consumes too much customer time or requires sales assistance. When the existing platform can recognise context, provide a trustworthy answer and measure the result, it becomes a foundation for an active sales channel. Further recommendations, segments and markets can then be added on evidence rather than assumptions.