Nobody loves you because you are right

I heard that line in a song. It’s also the post-mortem of far too many CPQ rollouts.

You ship a technically perfect, logically right system. It passes every test you wrote. Then sales bypass it. You see low adoption, a flood of manual quote requests, and Slack DMs asking, “Can you just price this one for me?” The tool works. The team doesn’t trust it.

I’ve been in those war rooms since 2000, first at Tacton and now at cpq.se. I’ve watched a spotless ruleset become a calculator of no. It’s painful, because it’s fixable.

Adoption is the only metric that matters.

What’s really broken

The usual story is simple: salespeople resist change, don’t understand the logic, or prefer their old spreadsheets. I hear this every month.

That story is wrong. Sales doesn’t distrust correctness. They distrust black boxes. When a configuration is rejected with no reason, or a price drops out of nowhere with no explanation, it creates uncertainty. In a commission world, uncertainty equals risk.

The problem isn’t the user. The problem is a system that can’t explain itself well enough to earn confidence. If the tool leaves a rep guessing, they will route around it. Every time.

Sales doesn’t fear rules. They fear not knowing why.

We keep forgetting what CPQ really is. Yes, it’s about correctness. If you sell what you can’t build, you donate margin. But correctness is the floor, not the ceiling. The ceiling is confidence. We aren’t just building a calculator. We are building a trusted co-pilot for complex deals.

I remember a rollout where we had pristine logic in Tacton CPQ. Engineering loved it. In the first week, sales called it “the trap.” The logic was right, but errors surfaced like verdicts, not explanations. Once we flipped the UI to show the why behind each block and each price step, usage climbed in days. Same rules. Different trust signal.

Rules for building a trusted system

Rule 1: Explainability is trust. Show the reasoning, not just the result. If a configuration is invalid, say which constraint was violated, where it lives, and how to fix it. If a price is calculated, show the components: base, options, discounts, approvals. Give the rep a sentence they can repeat to a customer without looking silly.

Example: add a simple Why? link next to each blocker and price line. Clicking it reveals the rule name, the input, and the path. In one program we saw escalations drop by half just by exposing the evidence behind the number.

If the system cannot explain itself, it will never be trusted.

Anti-pattern: The Oracle. The system that answers but never shows its work. It looks powerful in demos. It dies in the field.

Rule 2: Confidence over completeness. Ship a smaller surface area that works every time, and is simple to understand, before you ship every product variant. 100 percent coverage that confuses is worse than 60 percent coverage that earns belief. Sales will choose reliable and explainable over comprehensive and opaque, every day.

Example: Start with the top 20 selling configurations. Nail the pricing breakdown and explainers there. Once reps trust that core, expand. I’ve seen teams double adoption by cutting scope in half.

Anti-pattern: The Everything Box. A massive first release that looks heroic on a project plan, then stalls because no one trusts the edge cases.

Rule 3: Design for a salesperson under pressure. Your user isn’t a patient analyst. They are on a call, live, with a customer. They need speed, clarity, and a graceful failure path. Latency kills. Mystery errors kill faster. Favor plain language over model names. Optimize the path to a safe quote, not the path to a perfectly modeled universe.

Example: replace cryptic errors like “Constraint C-117 violated” with “Motor M2 requires 3-phase power. Choose supply S3 or change motor.” Give a one-click fix. Show the safe route like a GPS for complex sales.

Anti-pattern: The Lab Bench. A UI that showcases the data model instead of the decision the rep needs to make. Great for architects. Terrible for deals.

Rule 4: Evidence beats opinion. Instrument trust. Track how often reps request a manual override, how often they click Why, and which blocks cause abandon. Put these on your weekly dashboard. If usage goes up when you add explainers, you’re building trust, not just features.

Example: add a one-click “Send with confidence” checkbox at quote finalization. If reps uncheck it, ask why. You’ll get your backlog from those answers.

Rule 5: Logic is necessary. Narrative is decisive. When talking about systems that only uses symbolic logic for configuration, it’s tempting to think more rules equal more trust. They don’t. Trust comes from rules that are visible, named, and testable, paired with microcopy that tells the rep what happened and what to do next.

Example: every rule that blocks a choice must have a human name and a one-sentence reason that could live on a quote note. The logic handles correctness. The narrative earns adoption.

Correctness is the floor. Confidence is the ceiling.

What to do next

1) Make the logic visible. In your CPQ, add Why links to every validation and price element. Show the rule name, the input, and the resolution hint. Add a price breakdown panel with base, options, adjustments, and approvals. If your platform can’t expose this, fix that first.

2) Cut scope to raise trust. Identify the top 20 selling configurations and make them bulletproof. Remove confusing edge cases from the first-line flow. Publish a simple escalation path for the rest. Build credibility before you expand. Progress beats perfection, every time.

3) Measure adoption, not features. Put daily usage, abandon rate, manual quote requests, and the count of Why clicks on one page. Review it weekly with sales and product. When you change something, look for the trust signal to move. If it doesn’t, you shipped noise.

In one rollout, we ran a five-call test: sit with a rep on five real calls, stopwatch in hand. Track time to first valid configuration, number of unexplained blocks, and whether the rep could explain the price without help. That scorecard told us exactly what to fix. No workshop needed.

If you’re stuck waiting for perfect pricing, don’t. Perfect pricing is a myth. Learning systems win. You will improve margins faster with a working CPQ that captures feedback than with another quarter of spreadsheet arguments.

And if someone suggests ripping out rules to “let AI figure it out,” remind them: AI depends on logic. Fluent guesses don’t close deals. Clear constraints paired with assistance do.

Adoption is the only metric that matters.

A correct quote that never gets sent is worthless. Building a system that sales trusts is the only work that matters.