A new paradigm: the best CPQ might be the one you build yourself

For the past 10 years, the main argument when selecting a Configure-Price-Quote (CPQ) tool was pretty obvious – let’s not reinvent the wheel, go best-of-breed, standard, following “configuration over customisation” principle.

I heard many times: “if our products or commercial models can’t be modelled with a standard, out-of-the-box configuration – let us know. We’ll revisit them and adhere to industry standard” (don’t even dare to guess how many times it actually happened).

So people selected CPQ tools promising to be 80% ready out-of-the-box, catalog-driven, fully configurable – and then few months down the implementation with dozens of consultants performing acrobatic feats trying to fit unique business rules into the available configuration framework (and billable hours crunching along) it suddenly became more difficult than originally anticipated.

Some of these projects made to production. Many failed spectacularly. Only a small fraction actually generated positive business value and reached a solid adoption rate justifying initial investment.

Yet I still remember a conference-floor conversation with a (prospective) customer who once said to me something like – “You know… I think the best CPQ that we ever had was the one we’ve built for ourselves. It worked exactly how we wanted”.

And I think there’s a piece of wisdom in that. Every business is slightly different, yet for years we’ve been trying to fit these little nuances into a common framework. Mainly because building your own was simply too expensive.

But I do think that something is shifting here.

With AI-assisted engineering you can build a quoting tool that is 100% fit for your business. Cost of development? 1/5 if not 1/10th of what it used to be. Licensing – zero. Hosting, support & maintenance – let’s say that doesn’t change. TCO – I’ll let you do the math yourself.

I know this sounds just like yet another “AI-chad” post, but I literally spent past 15 months playing with various AI-assisted engineering approaches and I firmly believe there is something in that. In this new reality it’s easier (and cheaper) to hardcode your complex, non-standard, everchanging business rules than to actually make them configurable 😮.

Curious to hear if others have been thinking about this? Or if you’re at the edge of deciding on CPQ direction – drop a DM via LinkedIn and let’s have a chat.

Disclosure: written by a human (myself). AI was involved. Mostly ignored, occasionally adhered to especially when it comes to spell-check and grammar.