A head of sales sent me a screenshot in April. It was a quote one of his reps had emailed to a 200-seat prospect. The line item math was wrong by $14,000, the discount stacked on top of an already-discounted annual rate, and the contract term said 12 months in one place and 24 in another. The prospect's procurement team caught all three before the rep did.
That deal still closed. But it closed slower, at a worse price, and with a buyer who now reads every number twice. The rep was not careless. He was quoting a five-SKU product with three pricing models out of a spreadsheet that four people edit, and he was doing it at 6pm between two other deals.
This is the moment most founders start Googling CPQ. And it is also the moment most of them make a $150K mistake by buying the wrong thing too fast. I have set up revenue operations at SMB and Series A and B companies for the last ten years, and quoting is one of the few areas where I see teams err in both directions at once. Some buy enterprise CPQ years too early and spend four months configuring software they do not need. Others limp along on a broken spreadsheet until a deal like the one above costs them a logo.
This guide is about finding the line between those two mistakes, and what changed in 2025 that moved the line.
What CPQ actually is
CPQ stands for configure, price, quote. The software does three jobs that a spreadsheet does badly once your pricing gets complicated.
Configure: it stops a rep from building a combination of products that you do not actually sell, or that breaks at renewal. Price: it applies the right rate, the right discount within the rep's authority, and the right currency without anyone retyping a number. Quote: it produces a clean document and routes it for approval when the deal needs a sign-off.
That is the whole function. Everything else a CPQ vendor sells you is an add-on to those three jobs. If your reps quote one product at one price with one discount tier, you do not have a configure problem or a price problem. You have a quote template problem, and HubSpot or a Google Doc solves it for free.
CPQ earns its cost when the number of valid ways to price a deal is larger than a rep can hold in their head.
One product, one tier, one discount band does not need it. Five SKUs, usage and seat pricing, multi-year ramps, and partner margins do. The trigger is complexity, not revenue.
The two ways B2B teams get this wrong
The published advice says you need CPQ at $50M to $100M ARR. I think that number is useless, because it answers the wrong question. ARR tells you how much you sell. It tells you nothing about how hard each quote is to build. I have seen $4M ARR companies that genuinely needed quoting software and $30M companies that did fine with a locked spreadsheet and good discipline.
Here are the two failure modes I see.
Buying too early. A Series A team closes a Salesforce deal, an admin watches a demo, and someone decides the company needs Salesforce CPQ before it has 15 reps. Four months and a six-figure implementation later, they have a configuration engine governing four products and two discount tiers. The reps route around it because it is slower than the old way. The tool becomes shelfware that finance still pays for. I have walked into two of these in the last year.
Buying too late. The other failure mode is the screenshot I opened with. The team keeps quoting out of a shared sheet long after the product line and pricing models outgrew it. Manual quoting still produces errors in close to 30% of transactions, and every one of those errors either slows a deal or trains a buyer to distrust your numbers. By the time someone counts the cost, the company has lost deals it never logged as quoting failures.
The real trigger: complexity, not ARR
Forget the revenue threshold. Run your quoting through this checklist instead. If three or more of these are true, you have crossed the line where quoting software starts paying for itself.
- More than three products or SKUs that combine in different ways
- Two or more pricing models in play, for example per seat plus usage plus a platform fee
- Discount approvals that already route through Slack DMs or the CRO's inbox
- Multi-year deals with ramped pricing, where year one and year three are different numbers
- Reps quoting in more than one currency
- Partner or reseller margins layered on top of customer pricing
- A finance team that has caught more than one pricing error this quarter
The reason this list beats the ARR number is that it measures the thing CPQ actually fixes. A 12-rep team selling a usage-based platform with add-on modules hits five of these by Series A. A 40-rep team selling one flat-rate seat license might never hit three. Sell the second company CPQ and you have created the too-early failure mode on purpose.
This is also where a clean pricing structure matters more than any tool. If your own team cannot explain the price book on a whiteboard, no software will save the quote. I wrote about where that structure leaks in B2B pricing strategy, and it is worth fixing the price book before you automate it. Automating a mess just produces wrong quotes faster.
What changed in 2025
If you researched CPQ two years ago, the default answer was Salesforce CPQ. That default is gone. In 2025 Salesforce moved CPQ into what it calls an "end of sale" phase. New customers can no longer buy it. Existing customers can keep running it, but no new features are coming, and Salesforce is steering everyone toward Revenue Cloud Advanced, a heavier and more expensive product built around full revenue lifecycle management.
End of sale is not the same as end of life. Nobody is ripping CPQ out of your org tomorrow. But the signal is clear: the most common CPQ on the market is now a legacy product, and anyone starting fresh is choosing between Revenue Cloud, an independent platform like DealHub or Conga, or staying inside their CRM with native quoting plus a pricing layer.
For most Series A and B teams reading this, that last option is the one to look at first. The build-versus-buy math shifted the moment the default buy answer became a deprecated product.
What to use before you buy CPQ
Most teams that think they need CPQ actually need three smaller things, in this order.
A locked price book. One source of truth for list prices, discount bands by deal size, and who can approve what. This lives in a sheet or a CRM property, and it is read-only for reps. Half the "we need CPQ" conversations I have end here, because the real problem was that everyone had their own copy of the prices.
Native CRM quoting. Both HubSpot and Salesforce produce quotes from CRM deal data. If your products map to line items and your discounts map to deal size, native quoting plus an approval workflow handles a surprising amount of complexity. We build these for clients as part of CRM and RevOps work, and a well-wired HubSpot Quotes setup carries a lot of teams through Series B.
An approval workflow that routes by rule. The thing reps actually hate is waiting on a Slack reply for a sign-off. You can fix that without CPQ. A simple rule that routes any discount over the rep's band to the right approver, with a logged decision, removes most of the chaos people blame on the lack of CPQ. I covered the governance side of this in the deal desk process, and it pairs with native quoting better than most people expect.
The point of doing it in this order is that each step makes the next decision clearer. By the time native quoting genuinely cannot handle your configuration rules, you will know exactly which rules broke it. That spec is what you hand a CPQ vendor, instead of buying a platform and reverse-engineering your process to fit it.
What CPQ actually buys you
When the complexity is real, the payback is real too. The numbers vendors cite are rosy, so treat them as the ceiling rather than the expectation. Across published 2025 CPQ data, teams report quotes produced roughly 10 times faster, approval times cut by up to 95%, quote turnaround down about 28%, and configuration errors down 42 to 45%. Some report larger average deal sizes from guided selling, because the tool surfaces the add-on the rep would have forgotten.
What I see in practice is narrower and more useful than the marketing numbers. Quote errors stop reaching buyers, which protects price and trust. Reps stop losing an hour a day to quote admin. And the sales cycle loses the dead days where a deal sat waiting for someone to approve a discount over Slack. Those three outcomes are the whole reason to buy. If a vendor cannot tie their demo to those, you are looking at the wrong tool.
The cost side is where teams underestimate. An enterprise CPQ rollout runs 8 to 16 weeks with a dedicated project team, and all-in costs commonly land in the high five to low six figures once you count licenses, the platform stack underneath, implementation, and the admin who maintains it. That is a real number to put against a real problem, not a tool to buy on a hunch.
Not sure if you have a CPQ problem or a price book problem?
Book a free 30-minute audit and we will tell you which of the three smaller fixes you need first, and whether CPQ is actually one of them.
Book an audit →How we would roll it out
When a client does have a genuine CPQ case, the rollout looks a lot like the deal desk work that usually precedes it. Map the products and the valid combinations first, on paper. Write the pricing and discount rules as plain language before anyone touches software. Pick the tool against that spec, not against a demo. Build it in a sandbox, test it on your ten most common deal shapes, and only then move it live.
The teams that get burned skip straight to tool selection. They buy the platform, then try to discover their own pricing logic by configuring it, which is the most expensive possible place to figure out how you sell. The software is the last decision, not the first. If you want the same discipline applied to the rest of the revenue stack, the go-to-market and AI automation work follows the same order: process on paper, then the tool.
FAQ
When does a B2B SaaS company actually need CPQ?
When the number of valid ways to price a deal outgrows what a rep can hold in their head. In practice that means three or more SKUs that combine, two or more pricing models, ramped multi-year deals, or partner margins. ARR is a poor trigger because it measures how much you sell, not how hard each quote is to build.
Is Salesforce CPQ still worth buying in 2026?
For new customers it is not an option. Salesforce moved CPQ to end of sale in 2025, so you cannot buy it fresh, and existing customers get no new features. Anyone starting now is choosing between Revenue Cloud Advanced, an independent platform like DealHub or Conga, or native CRM quoting plus a pricing layer.
Can HubSpot replace a CPQ tool?
For many Series A and B teams, yes. HubSpot Quotes builds quotes from deal data, and paired with a locked price book and a rule-based approval workflow it handles real complexity. It stops being enough when your configuration rules get too intricate for a workflow to express, which is the point where a dedicated CPQ earns its cost.
How much does CPQ cost to implement?
Enterprise rollouts run 8 to 16 weeks with a dedicated team, and all-in first-year costs commonly reach the high five to low six figures once you count licenses, the platform underneath, implementation services, and ongoing admin. The native-quoting path costs a fraction of that and carries most teams further than they expect.
What is the most common CPQ mistake?
Buying the tool before fixing the price book. If your own team cannot explain the pricing on a whiteboard, the software just produces wrong quotes faster. Lock the prices, move quoting into the CRM, and route approvals by rule first. If that genuinely breaks, you will have a clean spec to buy CPQ against.
Get the quoting fixed before it costs a deal
Most teams asking about CPQ have a price book problem, a quoting problem, and an approval problem stacked on top of each other, and only one of those needs software. We untangle which is which, fix the cheap ones first, and only put you on CPQ when the complexity is real.
If your reps are quoting out of a spreadsheet and you have already caught one bad number reach a buyer, book a free audit and we will show you the three fixes we would make first.