CPQ for B2B Manufacturing in 2026: Three Incumbents and a Vibe-Coded Wildcard

We get asked some version of "just tell us which CPQ to buy" constantly, and every year the honest answer gets a little more complicated. 2026 added a real wrinkle: alongside the usual incumbents, we're now fielding questions from manufacturers whose engineering teams have quietly prototyped a configurator themselves, using AI coding assistants and off-the-shelf solver libraries, before ever calling a vendor. This piece is our opinionated read on both stories, the incumbents and the wildcard, not a neutral roundup. It's the manufacturing-specific, more argumentative companion to our broader vendor comparison.

The 2026 CPQ landscape in 30 seconds

Full disclosure before we go further: cpq.se is a Tacton reseller and implementation partner for small and mid-sized manufacturers, and we advise enterprise organizations on vendor choice more broadly. We build with Tacton and know it best. That doesn't make the other two incumbents wrong for you, it means you should weigh our Tacton opinions accordingly and demo everything on your own product, not ours.

Tacton CPQ: the one that actually gets manufacturing

Tacton's pitch has never changed much, and that's the point: model the product's real engineering constraints, and let a solver guarantee that whatever a seller configures can actually be built. For a crane, a truck chassis, or a machine with genuine interdependent options, that guarantee is the whole game, a "close enough" configuration that reaches production is the expensive kind of mistake manufacturers actually make. Tacton was named a Leader in the Gartner Magic Quadrant for CPQ for the 4th consecutive year in 2026, and shipped an AI Product Modeling Assistant aimed at the people who build and maintain the constraint model, which is, in our experience, the actual bottleneck on most CPQ programs, far more than the buyer-facing chat interface everyone demos first. Our customer HMF Cranes cut quote cycles from 11 days to 32 hours across 35 countries this way; Swift Lifts collapsed 40 separate configuration workbooks into one model that distributors now use to quote and visualize independently. Neither result came from AI, both came from doing the modeling work properly, with AI now speeding up how that model gets maintained.

Conga + PROS: the pricing king's identity crisis

Conga completed its acquisition of the PROS B2B business on February 2, 2026, worth being precise about, because it was the B2B business specifically, not all of PROS. The logic is coherent on paper: Conga's document and contract-lifecycle heritage plus PROS's price-optimization science makes for the clearest pricing-intelligence story in CPQ. PROS itself was named a Leader in the 2026 Gartner Magic Quadrant for CPQ, with Conga positioned as a Challenger. What we'd flag for manufacturers evaluating this combination right now: integrating two product lines and a rebrand simultaneously tends to produce a rocky year of roadmap churn before the combined platform settles. If your bottleneck is genuinely price realization, discount governance, margin visibility, deal analytics, rather than whether a configuration is physically valid, this is worth a serious look. If your bottleneck is product validity, this isn't the incumbent to start with.

Configit: compiled configuration, guaranteed performance

Configit plays a different game than the other two. It's a Configuration Lifecycle Management platform first, with CPQ sitting on top as one application layer, the core technology, Virtual Tabulation, pre-compiles the entire configuration space so that validity and performance are guaranteed by construction rather than checked at runtime. In 2026 Configit added Ace Prompt, a generative-AI assistant for building configuration models, mirroring the modeling-side AI bet Tacton has also made. The honest trade-off: you're not buying a quoting tool, you're buying a single configuration truth meant to span sales, engineering and manufacturing, which is the right call at genuine enterprise scale (think automotive-style product complexity across multiple systems) and more platform than most mid-sized manufacturers need on day one.

The wildcard: vibe-coding your own CPQ

Here's the part of this piece that didn't exist two years ago. General-purpose AI coding tools have gotten good enough that some manufacturers, usually with a capable in-house engineering team and a product that isn't yet at its most complex, are experimenting with stitching together a lightweight configurator themselves: an AI assistant to write the interface and glue code, and an open-source constraint solver like Google's OR-Tools to do the actual validity checking. We think this is a genuinely interesting development and not a fad worth dismissing outright. It also inherits the exact problem every home-grown rules engine has always had: it works fine right up until the product is complex enough that an unvalidated configuration slips through to production, and by then the cost of that mistake usually dwarfs whatever the DIY approach saved on license fees. [VERIFY: current data on how many manufacturers have actually shipped a production vibe-coded configurator versus prototyped one and moved on, we don't have a reliable figure for this yet.] We covered this trend in more depth, alongside the rest of the AI-in-CPQ picture, in our market trends piece.

What about Salesforce, SAP and Oracle?

We didn't forget the big suites, we're deliberately narrowing this piece to the manufacturing-specialist question, because that's where most of the confusion actually lives. Salesforce's strategic product is now Agentforce Revenue Management; it's a strong choice if your CRM gravity is already Salesforce and your complexity is mostly in the deal rather than the product. SAP CPQ makes sense as the default when SAP already runs your ERP, with solid pricing integration but real implementation weight. Oracle CPQ is one of the more mature enterprise platforms, a repeat Gartner Leader, and a sound pick for large sales organizations already on the Oracle stack. None of the three were built around engineering-constraint validity the way Tacton or Configit were, which is exactly why manufacturers with real product complexity keep ending up on this page instead. For the full picture across both shortlists, see our complete CPQ vendor comparison and our broader CPQ for manufacturing guide.

How to pick your headache

Every one of these options comes with a real trade-off, and picking the wrong one is expensive to unwind. Our shortlist logic: if your product has genuine engineering complexity and quotes routinely wait on an engineer, start with Tacton or Configit and let ERP/CRM gravity break the tie. If your pain is pricing discipline more than configuration validity, put Conga+PROS on the list and watch how the integration settles over the next few quarters. If you're tempted by the wildcard, be honest with yourself about how complex your product actually is today, and where it's headed, before you bet a sales process on code nobody outside your engineering team can maintain. None of this is independent advice in the purest sense; we sell and implement Tacton and say so plainly. It's still, we think, an honest one. If you want a second opinion on your specific shortlist, or the pricing math behind any of these options, book a meeting or read what CPQ actually costs before you sign anything.

Frequently asked questions