The short answer. The license is the smallest part of what CPQ costs, and it is the only part on the price page. Implementation takes months of paid work, and an admin has to keep the pricing rules current for as long as you use it. If you build your own quote tool instead, building is now cheap. Running it is the real cost.
A sales ops team of five signs a CPQ contract for 40 sellers. The license line is easy to see: seats times a monthly price. Six months later the quote tool is still being configured, a partner is billing by the week, and one person on the team spends half of every week editing price rules. None of that was on the order form.
This post walks through the three costs of buying CPQ, then the cost of building and running your own, and ends with a worksheet you can fill in. If you are new to the term, what is CPQ covers the definition. The case for building is in why build your own CPQ.
Cost 1: licenses, charged per seat
CPQ is priced per user. Every seller who builds a quote needs a seat, and so does every approver who looks at one. Add a sales team in Q2 and the bill grows with it.
Some vendors publish list prices. Many do not. This is what each vendor’s own pricing page said on November 14, 2026:
| Vendor | Published price | Notes |
|---|---|---|
| Salesforce Revenue Cloud Growth | $150 per user per month, billed annually | Quoting and configurator, order capture, subscriptions |
| Salesforce Revenue Cloud Advanced | $200 per user per month, annual contract | Adds contracts and orders, consumption and invoicing, AI and analytics |
| Salesforce Revenue Cloud Billing | Request a quote | Sold separately from its prerequisite products |
| PandaDoc Starter and Business | $19 and $49 per seat per month, billed annually | Neither plan includes configure, price, quote |
| PandaDoc Enterprise | Not published | Configure, price, quote is an optional add-on |
| DealHub | Sales led, no published price | The pricing page offers a “Request Pricing” button |
Two details matter more than the headline numbers. First, Salesforce’s pricing page says prices are for information only and subject to change, and that detailed pricing comes from a sales representative. Its pricing FAQ also describes Revenue Cloud as a paid add-on to Sales Cloud, so budget for both. Second, the only plan on PandaDoc’s page that offers CPQ is the one with no published price.
Salesforce CPQ, the older product, is a separate case. Salesforce has announced end of sale. Its CPQ page lists no price, existing customers keep support and critical fixes, and Salesforce says no migration is forced. It has not announced an end-of-life date. The full story is in what the Salesforce CPQ end of sale means.
To put a number on this line, multiply your seats by the monthly price and by 12. Then add the seats you will need next year. If you are comparing vendors, the DealHub alternatives list covers how the main products price and where each one fits.
Cost 2: implementation, which takes months
A CPQ license does nothing on its own. Someone has to put your products, prices and rules into it, and connect it to the systems around it.
The work usually includes:
- Product catalog. Every product, bundle, option and dependency, entered and tested.
- Pricing rules. List prices, volume tiers, regional price books, and contract pricing for named accounts.
- Discount and approval rules. Who can approve what, at which discount level, for which deal size.
- Quote documents. Templates that look like your company sent them, with the right legal terms.
- Integrations. The CRM on one side and the ERP or billing system on the other, so a won quote becomes an order. See quote to order for where that handoff fails.
- Testing and training. Real quotes run through the new rules, then sellers taught to use them.
It takes months because the rules already exist, but nobody has written them down in one place. They live in a spreadsheet, a senior rep’s head and a few old Slack threads. Implementation is partly the work of finding out what your own pricing policy is.
We could not find a reliable public figure for implementation cost, so we are not going to quote one. What is public is time. Salesforce says a typical migration from Salesforce CPQ to Revenue Cloud takes 3 to 6 months, and that complex setups take longer. That is Salesforce describing its own move between two of its own products. A first implementation from scratch should not be assumed to be faster.
When you get a quote from a partner, ask four questions. What exactly is in scope? What happens when the scope grows? Who owns the rules afterward? What does a change request cost?
Cost 3: admin and change, every month
After go-live, the pricing rules need an owner. Every one of these is a ticket for that person:
- A price change, in every price book it touches.
- A new product, with its options, dependencies and quote template.
- A new discount rule, or a changed approval threshold.
- A new region or currency.
- A reorganization that changes who approves what.
Most mid-sized teams end up with a CPQ admin, often a part of someone’s job that grows into a full one. Even when the vendor’s admin screens are well designed, the system is generic. Your business is specific, so the translation work never ends.
This cost is the easiest to miss in a budget, because it is salary rather than spend, and it shows up as a queue. A seller waits two days for a new bundle to appear in the quote tool. A deal desk approves a discount by message because the rule in the system is out of date. Both are admin costs, paid in slower deals.
The same pattern shows up with discounts. If the rules in the tool lag the rules in people’s heads, approvals move back to Slack, and the record you bought CPQ to keep goes with them. The deal desk discount approval guide walks through a model where the rule check runs as the discount is typed.
The worksheet at the end of this post adds the three buy costs up next to the cost of building.
The cost of building your own
Here is what has changed. A person who knows your pricing can now build a working quote tool with an AI assistant. A price book, a quote builder, discount rules with an approval path and a printable quote document are a few days of work, not a few months of configuration. The four-part model, and the prompts to build it, are in Tray Helix’s guide to building a CPQ tool.
So building is cheap. If that were the whole story, every company would have done it already. The cost sits in running the app after the demo works.
A quote tool holds prices and discounts. Quotes become contracts. That means the app needs:
- Sign-in. Only your people get in, and a leaver loses access without anyone remembering to remove it.
- Permissions. A seller sees their own quotes. An approver sees the ones waiting on them. Nobody edits the price book by accident.
- Approval records. Who approved which discount, at what level, and when. This is the reason most teams buy CPQ in the first place.
- Hosting. Somewhere to run it, a database, backups and someone to answer when it goes down.
- An owner. A named person who fixes it and keeps the rules current. If nobody wants this job, do not build.
What it costs to replace a SaaS app argues the same point at more length: the subscription is the visible number, and the people who keep the replacement alive are the one that decides whether it is worth it.
This is what Tray Helix is for. Tray Helix deploys, runs and governs apps that people build with AI. You build in your AI assistant, deploy with one command, and the app runs on a managed runtime with sign-in, access control, execution logs and a named owner, and SSO and an audit trail on Helix Enterprise Edition. There are no per-user fees, which matters most for CPQ, because the usual cost driver is the number of sellers and approvers. You can start on the Tray AI Orchestration Platform for $50 a month with Tray Helix Builder Edition, plus usage.
A CPQ app also needs your systems. It reads deals from Salesforce or HubSpot and sends orders to NetSuite or another ERP. That is the integration side of Tray, which connects to 12,000+ apps, databases, APIs, and services. See CPQ to ERP sync and quote to order for how it works.
If you want a rough sense of the gap for your size of company, the SaaS savings calculator is public and asks for no email. For a company of about 2,000 people, it estimates a saving of about 75% on the apps it models. That is the calculator’s estimate, not a quote, and it covers more than CPQ.
When building is the wrong answer
Be honest with yourself on three things.
- Deep configurators. If you sell products with hundreds of valid combinations, such as engineered or made-to-order equipment, a specialist configurator has years of work behind it. Keep it.
- Very large catalogs with complex bundling. Thousands of products with rules between them are a hard thing to rebuild and a harder thing to test.
- No owner. If nobody on your team will keep the pricing rules current, any quote tool will go stale, and an app you built will go stale faster than a product someone else maintains.
Trade-offs on the Tray side are plain too. Tray Helix is a managed cloud runtime, and the app is yours to run, so the owner is still you. Buying CPQ moves some of that to the vendor, at a price per seat.
A worksheet: buy versus build
Fill in one column for the CPQ product you are considering and one for building on Tray Helix. Use your own numbers. Leave a line blank if it does not apply, then compare the totals over three years.
| Cost line | Buy a CPQ product | Build and run on Tray Helix |
|---|---|---|
| Seats today (sellers plus approvers) | No per-user fees | |
| Price per seat per month | Not applicable | |
| License per year | Tray Helix plan, from your quote | |
| Seats added over three years | Not applicable | |
| Implementation partner or services | ||
| Internal time to implement or build | ||
| Integrations to CRM and ERP | ||
| Admin salary share (price changes, new products, rules) | ||
| Hosting and database | Metered as usage, one number on the bill | |
| Sign-in, permissions, audit record | Built in. SSO and audit history come with Helix Enterprise Edition | |
| Named owner, share of salary | ||
| Training for sellers | ||
| Renewal increase assumed | ||
| Total, year one | ||
| Total, three years |
Two lines usually decide it: the share of an admin’s salary and the number of seats in year three. If the buy column wins on both, buy. If the build column wins on both and you have an owner, build.
Sources
Read November 14, 2026.
- Salesforce, Revenue Cloud pricing and Sales add-on features and pricing
- Salesforce, Salesforce CPQ end of sale
- PandaDoc, pricing
- DealHub, pricing