What is CPQ? Configure, price, quote
Why does quoting get lost in spreadsheets? We break CPQ into its three steps — configure, price, quote — and the rules that make each one reliable.
Tuesday afternoon, and the sales floor goes quiet. A rep has to pull back a quote she sent three days ago: the mounting kit that ships mandatorily with the chosen model never made it onto the line items, and the reseller discount was applied twice on two separate rows. She had built it from a copy of last year's quote for a different account, its formulas wired to cells nobody could account for. Two questions are on the table: what do we tell the customer, and who absorbs the difference?
Any team selling configurable products or services knows this scene. CPQ — configure, price, quote — puts rules behind exactly those three steps. Below: what CPQ is, who needs it, what gets built in each stage, how to design the approval flow, and how to get there in stages rather than one leap.
What is CPQ? The meaning of the three steps
CPQ splits quote preparation into three separate questions. Configure answers "what exactly is the customer buying?" Price answers "how is this combination priced, and whose authority covers this discount?" Quote answers "how does that become a document a buyer can read and decide on?"
Keeping the three apart is not a technical nicety; it is the whole point. In a spreadsheet they collapse into one grid: product selection, discount and narrative live in the same file, guarded only by how alert one person happens to be that day. Wire them into a rule engine and each step gets its own validation, so the next step safely inherits the previous one's output. We walk through the full cycle in our piece on quote management; CPQ is the layer that puts rules behind the preparation half of it.
Who actually needs CPQ?
Not every team does. If you sell one product at one price as a single line item, a solid quote template is enough. The need appears as the multiplication of options grows and more people start writing quotes.
- High-variant products: When size, material, color, capacity and accessories multiply, holding the valid combinations in your head stops being realistic; a clean product variant management structure is the foundation CPQ sits on.
- Installation and commissioning lines: If a site survey, freight, assembly or start-up service travels with the product, the rules for adding each matter as much as the product itself.
- Tiered service packages: Starter, standard and enterprise tiers differ in scope, limits and add-ons; move a customer one tier up and half the quote is rewritten.
- Project-based work: When every job differs, a quote is really a small budget of labor days, materials and subcontractor lines, and one mistake comes straight out of project margin.
Configure: building the product with rules
The goal of the configure step is plain: only a combination that can actually be built, installed and delivered may reach the quote. Rules written over the product tree make that possible. It looks laborious at first, but the knowledge already exists inside your team — you are only moving it out of people's heads.
Compatible and incompatible combinations
Not every option works with every other. A given frame size will not carry a given motor rating, a finish is not rated for outdoor use, a software module needs a newer platform release. In CPQ this lives in the system, not in a rep's memory: an incompatible option either never appears or triggers a warning with its reason attached. A good configurator does more than block — it proposes the alternative that does work, turning the rule into a guide rather than a wall.
Mandatory items and quantity rules
The second family of rules kills forgotten lines. Choose a base product and the bracket, cable, license or training that must accompany it drops onto the quote by itself. Quantity rules follow the same logic: cable per unit, licenses per user, material per square meter, all bound to a formula. Scope then rests on the product definition, not on who happens to be quoting — and a new hire can produce a correct quote on day one.
Price: putting rules behind the number
The price step turns a validated configuration into a commercial figure. Four questions live here: which list applies, which tier applies, which added lines move the number, and who owns discount authority. The first three are largely solved with data — customer segment, contract term, volume tier, currency and delivery terms all feed the pricing rule. Lists that differ by account, and the matrices behind them, are a subject of their own, covered in customer-specific price lists and discount matrices.
Discount authority levels and approval thresholds
The fourth question is the delicate one. Leaving discounts open quietly erodes margin; routing every discount to the executive team stalls selling. The balance comes from authority levels: a band the rep can grant alone, a second that opens with the sales manager's approval, a third that goes to commercial leadership. The system reads the discount a quote produces and routes it to the right level automatically, so the rep never has to work out whom to ask. You set each band from your own margin structure; what matters is not where the line falls but that it is written down and configured.
Quote: turning it into a document people read
The third step converts a valid configuration and price into something a buyer can decide on. A quote is not a price list, it is an argument: scope, explicit exclusions, lead time, payment terms and the next step, all stated plainly. For the structure see our guide to the sales proposal template, and for the language inside it, how to write a winning price quote.
Validity period, revisions and versioning
Every quote needs an expiration date; it protects you against cost movement and nudges the decision along. Revisions should be numbered: when scope changes, issue a new version instead of overwriting the old file, so both sides know which one is under discussion. In Rocketly you can send a quote as a shareable link and see from the deal record when the customer opens it. Rather than waiting on wet ink, move to signing electronically — the mechanics are in e-signature contract workflows.
The approval flow: keeping quotes moving
The overlooked half of CPQ is approval. Rules get the content right, but quotes rarely stall because nobody could send them — they stall waiting for sign-off. A healthy flow makes three things explicit: what triggers approval, who receives it, and what happens if no answer comes. Requests should land in a system record, not a personal chat thread, and every approval stored with who accepted what and on what grounds.
Cover matters too: if the approver is away, the request reroutes to a deputy, and one left unanswered past a set window escalates a level. How to build that is in approval workflow automation. The principle is simple — an approval is a process step, not a favor.
CPQ and CRM: the quote lives inside the deal
None of the rules pay off if quotes are produced in a separate world. The right setup has the quote born inside the deal record: which opportunity, which version, which discount level it passed, who approved it, whether the customer opened it. As the deal advances, win probability updates with it. Methods for carrying a sent quote to a win are in quote follow-up and win rate.
What follows acceptance belongs to the same chain: order, invoice, collection. We take that chain apart in the quote-to-cash process; CPQ is about getting its first link right. If the first link is loose, every step after it carries the same defect.
What a wrong quote really costs
The cost of a bad quote is usually assumed to be the missing margin alone. The real cost is scattered: hours of corrections, credibility lost in front of the buyer, the scope argument that surfaces after the order, unplanned work pushed into the install schedule. If a line is missing, either you absorb it or you reopen a negotiation, and neither is free.
A mistake found in a quote is a mistake; the same mistake found after signature is a commitment.
| Aspect | Spreadsheet quote | Rule-driven CPQ |
|---|---|---|
| Configuration | Depends on who built it | Compatibility rules validate |
| Mandatory lines | Added if remembered | Added automatically |
| Discount | Verbal agreement | Authority level and approval |
| Revisions | Tracked in the file name | Version history retained |
| Visibility | On one laptop | Inside the deal record |
A staged roadmap to CPQ
Teams that try to build CPQ in one push usually get stuck writing rules. The healthier route is to start narrow and widen.
- Clean the product data: If codes, names, units and variant structures are inconsistent, no rule behaves; the catalog comes first.
- Start with your best-selling family: Rule one product family that carries a large share of revenue, not the whole catalog, and test it against real quotes.
- Attach the mandatory items: Bind the lines that get forgotten most often; the first visible win usually comes from here.
- Define authority levels: Write the discount bands and their approvers down on paper, then move them into the system.
- Connect the template and signature: Collapse the document to one template, then speed up closing with a shareable link and electronic signing.
- Widen the scope: Once the first family runs stably, take the next, and keep the rule set a living document.
What to measure
Whether CPQ is working is a question of a few plain indicators, not gut feel. Record the baseline for each before you switch, so improvement is visible afterward.
- Quote turnaround time: The span from request arriving to quote reaching the customer, tracked as an average and a worst case.
- Quote error rate: The share of quotes corrected after they were sent, which reads directly as the maturity of your rule set.
- Discount drift: How often a granted discount lands outside its band, telling you whether authority levels are honored in practice.
- Revision count: How many versions a quote goes through before closing, measuring how close the first attempt gets.
- Quote-to-order conversion: The share of sent quotes that become orders and the average time to close — the final exam for the whole chain.
Common mistakes
The first is writing rules disconnected from reality. A constraint nobody honors in the field gets worked around, and quoting quietly moves back into private files. Sales, production and service should write rules together, and every rule should carry a real reason.
The second is overbuilding. A configurator that opens with hundreds of exceptions is unmaintainable and exhausting to use. Most rules touch only a sliver of volume, so codify common combinations first and leave exceptions manual. The third is postponing approval: a flawless rule engine gains nothing if sign-offs sit in an inbox. The fourth is treating rules as finished — products and supply terms change, and the rule set must change with them.
One closing note: CPQ is a discipline decision more than a software purchase. Teams willing to write down their product knowledge, pricing logic and approval authority win with whatever tool they pick. If you want quoting to live inside your deal records, with shareable quote links moving the close along, create your Rocketly account and rule-map your first product family this week.