CPQ in 2026: A Configure-Price-Quote Buyer's Guide

CPQ — configure, price, quote — is the software layer that turns a sales rep's intent into an accurate, approved, deliverable quote without a spreadsheet, a pricing email chain, or a margin surprise at the deal desk. For any B2B business selling configurable products, tiered subscriptions, or usage-based pricing, it is the system that stops manual quoting errors and discount leakage before they reach a customer.

What makes 2026 an unusual year to buy CPQ is timing. The category is consolidating in public: Salesforce moved its long-dominant legacy CPQ package to End-of-Sale in March 2025, and Conga completed its acquisition of PROS's B2B business in February 2026. Thousands of enterprises are under simultaneous migration pressure, and the shortlist a buyer would have drawn up two years ago is no longer the right one.

This guide is written for the RevOps director, VP of Sales, or founder trying to make a defensible decision in that environment. It covers what CPQ genuinely solves, the migration wave reshaping the shortlist, the recognised leaders (sourced only from primary press releases), an AI capability maturity model, a scenario-based decision matrix, the real total cost of ownership, the build-vs-buy crossover, and why pricing guardrails matter more as AI takes over the quote.

Every B2B company that sells anything more complex than a flat price list eventually hits the same wall: quoting becomes a manual, error-prone bottleneck. A rep builds a spreadsheet, copies last quarter's template, emails finance for a discount sign-off, and three days later sends something that may or may not reflect current pricing, valid product combinations, or the customer's contracted rate. CPQ exists to remove every one of those failure points.

The cost of leaving it manual is well documented. According to figures cited by industry sources, Salesforce data shows roughly 43% of sales reps say generating quotes is the biggest time drain in their process, and IDC has been cited as finding that slow quoting adds around 20% to the average sales cycle. Spreadsheet-heavy environments have seen error rates reported as high as 8%, while quoting software wired into an ERP has been reported to push that below 0.5%. These figures come through secondary, vendor-adjacent sources rather than primary reports, so treat them as directional rather than audited — but the direction is consistent everywhere.

CPQ is the connective tissue between qualifying an opportunity and getting a signed deal. It sits squarely in the RevOps workflow — and it works best when the stages around it are defined. If you have not yet formalised how opportunities move from qualification into quoting, our breakdown of the discovery-to-proposal handoff is the prerequisite to getting full value from any CPQ rollout.

The real story in CPQ right now is not the technology — it is the consolidation, and the migration pressure it creates. Two events inside eighteen months reshaped the buying landscape.

First, Salesforce CPQ (the legacy SteelBrick-based managed package) entered End-of-Sale on March 19, 2025. New customers can no longer buy it; Salesforce has redirected development investment into Revenue Cloud, rebranded as Agentforce Revenue Management at Dreamforce 2025. This is End-of-Sale, not End-of-Life — existing customers can still renew their licences — but every Salesforce CPQ customer now faces a forward path decision. For the full Salesforce trajectory, see our guide to Salesforce Agentforce Revenue Management.

Second, Conga completed its acquisition of PROS Holdings' B2B business on February 2, 2026. The combined entity now covers CPQ, contract lifecycle management, document automation, and AI-powered pricing optimisation under one roof. This is an acquisition of a business line, not a merger of equals — but for buyers it means two formerly distinct shortlist entries are now one vendor relationship.

The practical implication: any CPQ shortlist drawn up before 2025 is stale. A buyer evaluating Salesforce CPQ as a new purchase is evaluating a product that can no longer be bought new. A buyer who shortlisted Conga and PROS as alternatives is now negotiating with a single combined vendor. The disruption is the context for every other decision in this guide — which is why we lead with it rather than the feature checklists most CPQ content opens on.

The 2026 Gartner Magic Quadrant for Configure, Price and Quote Applications evaluated 16 vendors. Because the Gartner report itself is paywalled, we cite only the Leaders confirmable from primary vendor press releases: PROS, Tacton, Infor, Oracle (recognised for the ninth consecutive year), and Conga. DealHub, SAP, Salesforce, Vendavo, Zilliant and others were also evaluated, but we will not assign quadrant positions we cannot source directly.

Here is the under-reported part. Analyst commentary notes that the 2026 evaluation criteria carry a structural weighting toward complex manufacturing and tangible goods — reportedly around 40% combined. That lens advantages configure-to-order vendors and can systematically under-rank SaaS-native platforms whose deal motion is subscription and usage-based. A high-growth software company that sees Oracle and Tacton at the top and concludes they are the best fit for its subscription business may be reading the wrong map.

The takeaway is not that the leaders are wrong — PROS's recognition as a Leader in The Forrester Wave: Configure, Price, Quote Solutions, Q1 2025 (with top scores across multiple criteria including pricing optimisation and AI) is real and earned. The takeaway is that vendor fit is a function of your deal motion, not a leaderboard. Picking a CPQ vendor is a structured evaluation exercise — the same discipline we lay out in our vendor selection framework applies directly here.

AI in CPQ is not one thing — it is a ladder. Reportedly about 54% of CPQ providers launched new AI-enhanced functionality in 2024, from predictive analytics to intelligent discounting, and the gap between a vendor saying "AI-powered" and what it actually does is wide. The maturity model below maps where a platform sits, so you can buy the level you need rather than the level on the slide deck.

Across these levels, organisations have reported quote cycle times falling 30-50%, pricing errors dropping by over 90%, and average deal size lifting 5-15% — figures that recur across the vendor-adjacent ecosystem and should be read as commonly-reported benchmarks rather than independently audited results. The honest planning posture is to buy the lowest maturity level that solves your actual bottleneck, then climb only when your data quality and pricing governance can support the next rung.

Most CPQ comparison tables are feature checklists across vendors. That is the wrong axis for a buyer. The matrix below is organised by buyer scenario — find the row that matches your business and read across to the platform class, realistic implementation window, relative cost tier, the AI capability that matters most, and the risk to watch. Cost tiers are deliberately qualitative ($ = lowest relative cost, $$$$ = highest) because real pricing is multi-layered and changes often; verify current numbers with each vendor.

The most under-considered row is the last one. A Salesforce CPQ customer facing End-of-Sale has a genuine fork: follow the platform to Revenue Cloud, or treat the forced migration as the moment to re-evaluate the whole category. Either way the cost is real, so the decision deserves the same rigour as a greenfield purchase rather than a default upgrade.

The single most common CPQ budgeting mistake is anchoring on the per-seat licence fee. Across pricing analyses, licence fees typically represent only 30-50% of total CPQ cost — implementation, training, ongoing administration, and integration development make up the rest. A 3-year total cost of ownership for a 50-user mid-market deployment has been reported to reach $600,000-$900,000 once professional services are included.

Salesforce CPQ is the clearest example of why the headline number misleads. The legacy package has been listed starting around $75 per user per month, but a CPQ seat requires Sales Cloud underneath it, so the fully-loaded starting cost has been reported at roughly $240 per user per month for a new customer. Enterprise-scale Revenue Cloud implementations have been reported at $150,000-$500,000+ in professional services on top of licensing. These are vendor-indirect figures — verify current pricing directly with the vendor before you model a budget.

For some businesses, the right answer is neither a per-seat SaaS CPQ nor a spreadsheet — it is a custom-built quoting layer. The build-vs-buy decision often hinges on quoter headcount: the more users paying a perpetual per-seat fee, the more attractive a one-time custom build becomes against years of subscription cost. It is not a universal answer, but it is a real crossover that off-the-shelf comparisons rarely surface.

Here is how we frame the decision for clients. Buy when your configuration logic matches what a mature vendor already encodes, when you need it live quickly, and when seat counts are modest. Build when your pricing logic is genuinely idiosyncratic, when seat counts are high enough that per-seat fees dominate, and when you have the engineering capacity to own it — increasingly realistic given how much AI-assisted development has compressed custom-build timelines.

This is precisely the kind of decision where a strong CRM & automation partner earns its keep — modelling the real TCO of each path against your seat count and config complexity, rather than defaulting to whichever vendor demoed best. Our view, consistent with how we build for our own clients, is that AI-assisted custom development has moved the crossover point meaningfully toward build for teams with distinctive pricing and the appetite to own their stack.

Most CPQ marketing celebrates AI autonomy. The more important engineering reality is the opposite: as AI takes more of the quote, the deterministic guardrails around it matter more, not less. The emerging best practice is a CPQ that pairs AI recommendations with hard-coded constraints the model cannot override.

Revenue leakage from disconnected quote-to-cash systems has been estimated at 1-5% of EBITDA, and the primary recovery mechanism is governance: mandatory approval workflows, hard-coded margin floors, and AI-enforced discount guardrails. Finance teams can set real-time margin guardrails the AI enforces automatically, only escalating to a human when a threshold is breached — autonomy inside the fence, never outside it.

There is a forward-looking risk worth naming. Industry analysts have warned about the governance exposure of increasingly autonomous pricing AI — one widely-cited Gartner prediction holds that by 2028, AI regulatory violations will result in a 30% increase in legal disputes. Whether or not that specific figure lands, the direction is clear: the organisations that win with agentic CPQ will be the ones that built the deterministic floor first and let the AI optimise above it — not the ones that handed pricing to a model and hoped.

This is the through-line of the whole 2026 CPQ story. The category is consolidating, the AI capability is racing up the maturity ladder, and the buyers who do well will be the ones who buy the level they can govern — pairing automation with hard constraints, modelling the full cost, and matching the vendor to their own deal motion rather than a leaderboard.