Pricing Page Research: How to Test Pricing Pages With Real Customer Interviews (2026 Playbook)

The bottom line: SaaS pricing pages convert at a median of 2–5% — and a 30–50% lift is achievable through better design and clearer communication, not lower prices. The fastest way to find that lift is to combine quantitative analytics with qualitative pricing-page research: structured interviews that watch customers read your pricing page in real time, surface confusion before they bounce, and reveal the exact words they use to justify each tier. Modern AI-moderated research tools like Koji can run 50+ pricing-page interviews in a week — work that used to take a researcher a quarter.

Why pricing page research is the highest-ROI study you can run

Your pricing page is the most expensive real-estate on your website. Nielsen Norman Group''s long-running B2B research has found that prospects rank "knowing the price" as their #1 information need on any website — and "people view companies that show this key piece of information as being genuine and forthright." Hide it, and trust collapses. Show it badly, and the visitor still bounces.

The numbers back it up:

• The median SaaS pricing page converts at 2–5%, but pricing pages typically out-convert homepages by 3–5x, making each percentage point of lift unusually valuable. (Kirro SaaS Conversion Benchmarks, 2026)
• A/B tests on pricing pages produce an average 12–18% improvement per winning test, and excellent design plus clear communication can increase conversion by 30–50% without changing price. (InfluenceFlow, 2026)
• Pricing pages that "display annual pricing as the default, highlight a recommended plan, and show a clear comparison of features between tiers" see a 23% average improvement in conversion.
• And 40–60% of pricing-page traffic now comes from mobile — where the comparison table that "works" on desktop becomes a confusion machine.

The catch: A/B tests tell you which layout won. They don''t tell you why — or whether your tiers are even named right. That''s what pricing-page research is for.

"Prospective customers want to know the price as their #1 info need on any website — but B2B sites often hide or obscure pricing information. In usability studies, we witnessed people getting frustrated and leaving sites that don''t show prices." — Nielsen Norman Group

What pricing page research is (and isn''t)

Pricing page research is qualitative research focused on a single artifact: the page where prospects decide whether to keep evaluating you. It combines three threads:

1. Comprehension testing — does the prospect understand what each tier includes, and which one is for them?
2. Value perception — does the price feel like a reasonable trade for the value, and where does it feel "too much" or "too cheap"?
3. Friction discovery — what stops them from clicking "Start free trial" or "Talk to sales" right now?

It is not:

• A Van Westendorp pricing study (separate quantitative method — see our Van Westendorp guide).
• A general willingness-to-pay study (covered in our WTP interview template).
• A homepage / messaging test — though it overlaps with messaging testing.

Pricing-page research sits at the intersection of pricing strategy and conversion-rate optimization. Done well, it surfaces both the price problem (the number) and the pricing problem (how you communicate it).

The 6-step pricing-page research playbook

Step 1 — Define the decision the prospect is trying to make

Before writing a single question, get specific about which prospect, on which device, in which stage of evaluation. A free-trial activator reading your pricing page on mobile after a paid social ad has a very different mental model from an enterprise buyer reading it on desktop after a demo. Run one study per primary segment. Use screener questions to recruit each segment cleanly.

Step 2 — Recruit 12–20 participants per segment

For qualitative pricing-page research, 15–20 in-depth conversations per segment is the saturation point most researchers report — the same range used in win-loss analysis. Below 10 you''ll miss patterns; above 20 you''ll mostly hear repetition. Recruit from:

• Visitors who hit the pricing page but didn''t convert (use in-product or exit-intent surveys to capture them — see in-product recruiting).
• Recent free-trial users who upgraded — they''re freshest on the comparison they just made.
• Recently lost deals where pricing was a stated objection.

Step 3 — Write a discussion guide that watches the page, not just discusses it

The mistake most teams make is asking, "How does our pricing feel?" Customers will tell you it''s fair (because that''s polite). Instead, ask them to open the page in front of you and narrate:

• Pre-page (60 sec): "Before you see the page, tell me what you''d expect to find." Anchors expectations.
• First impression (60 sec): "Walk me through what you''re noticing first, second, third." Reveals visual-hierarchy failures.
• Tier interpretation (3–5 min): "If you had to pick a plan today, which one — and why?" Surfaces wrong assumptions.
• Feature mapping (3–5 min): "Point to anything you don''t understand, or anything you''d want explained more." Surfaces jargon.
• Price reaction (3 min): "Without telling me a number, describe how the price feels in your gut — and what would have to be true for it to feel like a no-brainer."
• Friction (2 min): "What would stop you from clicking the CTA on the plan you''d pick?"

This is a think-aloud protocol applied to a pricing page — see our think-aloud guide for the underlying technique.

Step 4 — Mix structured and open-ended questions

You can''t analyze 20 hours of audio if every question is open-ended. Koji''s six structured question types — open_ended, scale, single_choice, multiple_choice, ranking, yes_no — let you blend storytelling with measurable data inside the same conversation. A good pricing-page study uses:

• Scale (1–5): "How easy was it to figure out which plan is for someone like you?"
• Single choice: "Which tier would you pick if forced to choose today?"
• Ranking: "Rank these five reasons for picking that tier in the order they matter to you."
• Open-ended: "Walk me through what made the price feel reasonable — or not."
• Yes/no: "Would you have clicked the CTA at this price without first talking to a human?"

That blend gives you both a quote wall and a scoreboard — which is what you need to present findings to a CEO who wants numbers and a designer who wants verbatims.

Step 5 — Analyze for the three failure modes

Pricing pages fail in three ways. Code your transcripts against each:

1. Comprehension failure — they couldn''t tell tiers apart, or misread what was included.
2. Value failure — they understood the page perfectly and still don''t think the price is worth it.
3. Trust failure — they understood the value but couldn''t commit (no social proof, unclear refund policy, "talk to sales" friction).

Each failure mode has a different fix. Comprehension is a copy/design fix. Value is a positioning, packaging, or messaging fix. Trust is usually a conversion design fix.

Step 6 — Ship a hypothesis to A/B test

The output of pricing-page research isn''t a redesign — it''s a prioritized list of hypotheses to A/B test. Each hypothesis should read: "Because [specific finding from N interviews], we believe [change] will improve [metric] by [estimate]." This feeds a hypothesis-driven experimentation pipeline.

The modern approach: AI-moderated pricing page research with Koji

Traditional pricing-page research is slow. A researcher books, screens, moderates, transcribes, and analyzes 15 interviews — that''s two weeks of work, $5–10K of recruitment costs, and a research backlog that''s already three quarters deep.

The AI-native approach inverts the math. With Koji''s AI-moderated voice interviews, the same 15-interview pricing study runs in 48 hours — not because you skip steps, but because the AI moderator runs them in parallel, 24/7, and surfaces themes as transcripts complete.

Here''s what changes:

• Always-on recruitment. A prospect bounces from your pricing page → an in-product or exit-intent prompt offers them a 10-minute interview → Koji''s AI moderator runs it immediately, in their browser, in their voice. No scheduling, no calendar tag.
• Adaptive probing. When a participant says "your pro plan feels overpriced," Koji''s AI follows up with a Mom Test–style probe a human researcher would: "Compared to what?" — surfacing the actual reference price in their head.
• Automatic thematic analysis. Koji''s AI auto-tagging codes every transcript against comprehension, value, and trust failure modes, then assembles a research report with the verbatims and scoreboard you need for stakeholders.
• Real-time scoreboard. Insights Chat lets a PM or designer ask, "Show me everyone who picked Pro but hesitated on price" — and get back the exact quotes, mid-study.

While traditional survey tools like SurveyMonkey limit you to multiple choice and Likert scales, Koji''s structured + conversational hybrid captures the why behind every rating. Teams that switch from manual moderation to AI-moderated pricing research report 60% faster time-to-insight and 5–10x more conversations per quarter.

Common pitfalls (and how to avoid them)

• Asking what they would pay. Stated willingness-to-pay is famously inaccurate. Watch the page; measure reactions.
• Recruiting only converters. They''ve already accepted the price. You need the bouncers.
• One round, one redesign. Pricing pages are continuously discoverable artifacts. Run pricing-page studies quarterly as part of continuous discovery.
• Skipping mobile. With 40–60% of traffic on mobile, run at least one segment on mobile-only sessions.

How to know when you''ve done enough

You''ve done enough pricing-page research when three things are true:

1. You can predict, from the participant''s segment alone, which tier they''ll pick — within one tier.
2. You''ve heard the same friction objections 5+ times without anything new emerging.
3. Your A/B test backlog has at least three prioritized hypotheses ready to ship.

That''s saturation. Move to the test.

Related Resources

• Pricing Research Interviews: How to Understand What Customers Will Pay
• Van Westendorp Price Sensitivity Meter
• Willingness-to-Pay Interview Template
• Messaging Testing Guide
• Think-Aloud Protocol
• Structured Questions in Koji Interviews