Skip to content

Blog / Executive perspectives

Why CPQ is the best app to build yourself

CPQ fits build-it-yourself better than almost any SaaS app: your rules, your data, few users, per-seat pricing. What it takes to run one, and when not to.

The short answer. CPQ (configure, price, quote) is one of the best apps for a company to build for itself. The rules are specific to you, the data already lives in your CRM and ERP, few people use it, and packaged products charge per seat. The catch is that a quote tool becomes a contract, so you need sign-in, permissions, an approval record and an owner. Tray Helix gives you those. Build it when your rules are your own, and buy it when your product configuration is genuinely complex.

What a CPQ project looks like today

A software company with 1,200 customers sells a platform fee, a per-device plan, and a handful of add-ons. Discounts under 10% need no approval. Between 10% and 20% they need a manager. Above 20% they need a VP, and anything on the AI add-on goes to the CFO. That is the whole policy, and it fits on one slide.

Now watch what happens when the company buys a packaged CPQ to enforce it. A consultant spends weeks modeling that one slide as product rules, price rules, approval chains and a quote template. A Salesforce administrator then keeps it running. Every new bundle, every change to a discount band and every new approver is another configuration ticket.

The software was supposed to hold your rules. Instead you spent months teaching it your rules, and you pay someone to keep teaching it.

That pattern is why CPQ is a strong candidate for building your own. Four things point the same way.

The rules are yours

Packaged CPQ is a general engine. It has to handle every company’s pricing, so it ships with a large vocabulary of rule types, and your job is to translate your policy into that vocabulary.

Your price book, discount bands, approval paths and quote template are specific to you. They are also usually simple to say in plain language, even when they are tedious to configure. “Over 20% on the AI add-on goes to the VP” is one sentence. In an AI assistant, that sentence is close to the build instruction.

This is the opposite of an app where the vendor’s value is deep, shared functionality, such as payroll or tax. In CPQ, the value is mostly your own policy applied consistently.

The data already lives in systems you own

A quote starts with a deal in your CRM and ends as an order in your ERP. The price list sits in a spreadsheet or the ERP’s item master. Item costs, which you need to show the margin effect of a discount, live in the ERP.

A CPQ product does not own that data. It copies it in, prices against the copy, and writes results back, which means a sync to build and a sync to maintain. A CPQ app you build reads the same systems directly. Tray connects to 12,000+ apps, databases, APIs, and services, so Salesforce, HubSpot and NetSuite are connections your app uses, not projects to scope. The CPQ to ERP sync guide and the quote-to-order guide cover that integration side.

Getting this handoff wrong is where most quote problems start. A quote approved at one number and booked at another is the pattern described in quote to order: where signed deals stall, and it feeds the wider order-to-cash process. The same deal also sets what reps are paid: the price, discount and split on the quote are the inputs to the commission, which is why sales commissions is the next app worth building.

Few users, priced per seat, with an admin attached

CPQ is used by sellers, sales engineers, a deal desk and sales finance. At most companies that is tens or low hundreds of people, not the whole workforce. Packaged CPQ is generally licensed per user. Salesforce’s own CPQ page, for example, describes existing customers buying additional users and renewing licenses.

Per-seat pricing suits an app where many people each get a little value. It suits CPQ less well, where a small group does work that drives revenue and the cost scales with how many sellers you hire. Add the implementation fee, the partner who knows the product, and the administrator who owns the configuration, and the subscription line is the smallest part of the bill. The total cost of buying versus building CPQ walks through each line.

An app you build and run on Tray Helix has no per-user fees. Hiring another ten sellers does not change what the quoting tool costs. You can start on the Tray AI Orchestration Platform for $50 a month with Tray Helix Builder Edition, plus usage.

For a company of about 2,000 people, Tray’s SaaS savings calculator puts the saving from replacing a typical set of SaaS apps at about 75%. That figure covers the apps you choose, not CPQ alone. Use it to size your own case.

Many teams are deciding anyway

Salesforce has stopped selling CPQ to new customers. Its own page is clear that this is end of sale, not end of life: existing customers can keep using CPQ, renew licenses, add users and get support, there is no forced migration, and Salesforce has not announced an end-of-life date.

That leaves a Salesforce CPQ customer with a real decision rather than a deadline. Salesforce points to Revenue Cloud Advanced as the successor and gives a typical migration of 3 to 6 months from discovery through go-live. Others look at products such as DealHub, and our DealHub alternatives post compares them. A third path is to move the rules into an app you own.

The full picture, including what Salesforce does and does not say, is in what the Salesforce CPQ end of sale means. If you are new to the category, what is CPQ starts from the definition.

What it takes to run your own

Building a quoting app with an AI assistant is the easy part. A first version that configures a quote from a price book, checks discounts and produces a document can come together in days. What separates a demo from a tool your sellers use is everything around it.

A quote becomes a contract and then an invoice. So the app needs four things from day one.

  • Sign-in. Sellers log in with their company account, not a shared link. Discount authority follows real roles.
  • Permissions. Sellers see their own quotes, the deal desk sees all of them, and only revenue operations can edit the price book.
  • An approval record. Every approval is stored with who approved, at what level, against which rule and when. An auditor will ask why a deal went out at 22%, and the answer cannot live in a Slack thread.
  • An owner. One named person gets the call when the price book needs changing or something breaks. When the builder changes jobs, the app must not become nobody’s problem.

These four are what a vendor supplies without anyone noticing, and they are where self-built apps stall. What it costs to replace a SaaS app makes the case that the running cost, not the build cost, decides whether replacing an app is worth it.

Where Tray Helix fits

Tray Helix is where an app built with an AI assistant gets deployed, run and governed. Your revenue operations lead builds the CPQ app in Claude Code, Codex or Cursor. One command deploys it to a managed runtime and returns a URL. Nobody stands up servers.

The four needs map directly. Every deployed app sits behind sign-in, with your SSO on Helix Enterprise Edition, and access is granted per project. The app never holds a raw credential: it names the system it needs, such as Salesforce or NetSuite, and Tray resolves the authentication on the server side. The app gets a built-in database, so the approval record has a place to live, and Helix logs every run. Every app has a named owner in the registry.

There are no per-user fees, so the sellers who use the tool are never what you are billed on. Helix does not make the app good, though. The quality of your CPQ depends on how clearly you write down your price book and your rules, and that part remains your team’s work.

Tray also covers the integration side. The same platform that runs the app connects it to your CRM and ERP, so the quote reaches the order without re-keying.

When you should not build

Buying is the right answer in several cases, and a fair comparison says so.

  • Deep product configuration. If you sell configurable manufactured products with engineering constraints, such as which parts fit together, a packaged configurator carries years of that logic. Do not rebuild it.
  • Very large catalogs with complex bundling. Tens of thousands of SKUs with nested bundles, regional price variants and rules that interact are hard to specify and hard to test. A mature product has already worked through the edge cases.
  • No one to own it. If nobody in the company will keep the price book current and answer when it breaks, a vendor’s support desk is worth paying for.
  • Heavy, unusual compliance needs. If your quote process needs specific regulated workflows that a vendor already certifies, check that before you build.
  • A working setup that you do not want to change. If your current CPQ runs well and the renewal is fair, moving is a cost with no gain.

If two or three of these describe you, buy. If none do, building deserves a serious look.

How to start

Do not start with the whole quote-to-cash flow. Pick the part that hurts most, which is usually the discount check and the approval path, and write down your price book and rules in plain language. Then build the four parts: a price book, a quote builder, pricing rules with an approval path, and a quote document.

The prompts to build each of those parts, and what it takes to run the result, are in the Helix guide to how to build a CPQ tool. If your first problem is discount approvals, start with the deal desk discount approval guide.

Sources

  • Salesforce, “Salesforce CPQ end of sale” (salesforce.com/sales/cpq/end-of-life/), read November 14, 2026. Source for end of sale, existing customer rights, no announced end-of-life date, no forced migration, Revenue Cloud Advanced as the successor and the 3 to 6 month migration timeline.
  • Tray, SaaS savings calculator, read November 14, 2026.

Frequently asked questions

Should you build your own CPQ?

+

Often yes, if your pricing rules are your own, your data sits in a CRM and an ERP you already run, and someone will own the app. Packaged CPQ earns its price for deep product configuration, very large catalogs and teams with nobody to maintain an app.

What does a CPQ tool need to do?

+

Four things: hold a price book as the only source of list prices, build a quote from it, check every discount against your rules and route exceptions for approval, and produce the quote document the customer signs.

How long does it take to build a CPQ tool with AI?

+

Building the first working version takes days to a few weeks, depending on how clearly your price book and discount rules are written down. Running it well takes ongoing work: sign-in, permissions, a record of each approval and a named owner.

What happens to Salesforce CPQ customers?

+

Salesforce says CPQ is end of sale, not end of life. Existing customers can keep using it, renew and add users, and Salesforce has not announced an end-of-life date. New customers can no longer buy it.

When should you not build a CPQ?

+

When you sell configurable products with deep engineering constraints, when your catalog and bundling rules are very large and complex, or when nobody in the company can own a business-critical app. Buy in those cases.

Build your CPQ, then run it properly

Tray Helix deploys, runs and governs the app your team builds with an AI assistant, with no per-user fees. Tray connects it to your CRM and ERP.

Great — a couple more details

This helps us connect you with the right person. Optional, and you can skip it.

See the CPQ build guide