Integrating CPQ

Getting the right data into CPQ, and the right data out

A CPQ system is only as good as the data behind it. Prices might come from ERP, from a price management system, or be maintained directly in CPQ. Wherever they live, they need to be correct. Product texts and images usually come from PIM. Configuration rules need to reflect what engineering actually builds. And when a quote is done, the result needs to flow back into CRM, into order management, into production.

We have spent 25 years connecting CPQ to the systems around it. The lesson we keep coming back to: start with the integrations that protect quality, and save the time-saving ones for later. If you are still working out what the software itself does, start with our guide to CPQ and come back here.

The broken telephone problem

In most manufacturing companies, product data lives in multiple systems. PLM, ERP, PIM, sometimes spreadsheets that one person maintains. When that data needs to reach the CPQ system, it often passes through several hands. Supply chain updates a part. Engineering makes a note. Someone exports to Excel. By the time the information reaches sales, it is either wrong, outdated, or missing entirely.

This is the broken telephone problem. The more steps between the source and CPQ, the more likely something gets lost. Good integration solves it by shortening the chain, ideally to zero manual steps.

But automation is not without risk. Remove a beacon from the product range because of a sourcing issue, and you might accidentally block sales in a country where that beacon is a legal requirement. The domino effects of automated data changes can be hard to predict.

That is why we split integrations into two categories.

Quality first, time savings second

Quality integrations fix things that would otherwise lead to wrong quotes. Pricing from ERP. Product rules from PLM. Country availability. If this data is not synced, sales will ship errors to customers. These go into the first release.

Time-saving integrations reduce manual work but do not break anything if they are missing. Auto-populating customer data from CRM. Syncing images from PIM. Generating order confirmations automatically. These can wait for phase two.

This split keeps the first release focused and gets you live faster, without compromising what matters. It is the same logic we apply to scoping the wider programme, described in our CPQ implementation process.

What we typically connect

ERP. Pricing, cost data, part availability, customer master data. This is usually the most critical integration. If CPQ prices do not match ERP, you will hear about it fast.

PIM. Product texts, images, technical specifications, marketing descriptions. Connected to CPQ so quote documents pull the right content in the right language without anyone copying and pasting.

CRM. Customer information, contact details, opportunity data. Flows both ways: CRM data into CPQ to pre-fill quotes, and quote data back into CRM so sales managers can see the pipeline.

PLM. Product structures, engineering BOMs, module variants, technical rules. When PLM is well maintained, it can drive parts of the configuration logic. But do not try to mirror the full PLM structure into CPQ. The sales view is always simpler than the engineering view.

CAD and visualisation. For manufacturers where seeing the product matters during configuration. We work with partners like Dynamaker to connect 3D visualisation and drawing generation directly to the configurator.

Workflow and automation. Approval flows, notification triggers, automated handoffs between systems. Useful for connecting CPQ to processes that are not covered by standard integrations.

Keep it debuggable

One thing we have learned the hard way: do not build integration as a black box. When something breaks, and it will, you need to understand why.

We recommend using a human-readable intermediate format, typically Excel, between the source system and CPQ. It sounds rudimentary, but it works. You cannot test every combination in a product configurator, you easily end up with billions of possibilities. It is better to inspect the data as close to the source as possible, before it goes into CPQ.

If a beacon suddenly disappears from the configurator, you want to see exactly where and why. Not dig through code to figure out what happened.

Modelbot, automated product modelling

For manufacturers who need to keep their CPQ product model in sync with master data from ERP or PIM, we work with Modelbot.

Modelbot reads product master data from your source systems, converts it into a structured format, and generates a ready-to-use product model for Tacton CPQ, including part structures, configuration interface and constraints.

The typical use case: you have hundreds or thousands of product variants maintained in ERP. Updating the CPQ model manually every time something changes is slow, error-prone, and does not scale. Modelbot automates this. It retrieves the data, processes it, builds the model, and uploads it, with full logging so you can see exactly what changed and why.

It is not a replacement for understanding your product structure. You still need a base model that defines the sales logic, the question flow, and the rules that do not come from master data. Modelbot handles the parts that change frequently and can be derived from your source systems.

HMF Cranes uses Modelbot to keep their CPQ models current as their crane range evolves. As their product manager put it, it makes it easier to implement new products and approach new markets.

Learn more about Modelbot

How we approach integration projects

We start by mapping what data exists, where it lives, and what quality it is in. Not everything that could be integrated should be, at least not right away. We help you sort out what is high-value and low-risk versus what needs more careful planning.

From there, we build integrations iteratively. Get the critical data flows working first, validate them with real configurations, then expand. Every integration gets an intermediate format you can inspect, and we make sure your team understands how the data moves, not just that it works.

The same team does the modelling and the delivery. See how we run Tacton CPQ implementations, or read why integration decides whether a quote becomes a clean order.

Ready to talk about integration?

Book a meeting and we will talk through your situation. No prep needed.

Frequently asked questions