How to Build an NPS Survey That Actually Drives Action

Net Promoter Score is the most widely used loyalty metric in business. Over two-thirds of Fortune 1000 companies track it. But here's the uncomfortable truth: most NPS programs fail to drive meaningful change. They collect a number, put it in a dashboard, and never understand why customers gave that score.

The problem isn't NPS itself. It's that a 0-10 scale without context is a vanity metric. The real value is in the follow-up, understanding the reasoning, emotions, and experiences behind the number. That's where Koji transforms NPS from a metric into a growth engine.

What Is NPS and Why It Matters

Net Promoter Score measures customer loyalty with a single question: "On a scale of 0 to 10, how likely are you to recommend [company/product] to a friend or colleague?"

Respondents are categorized into three groups:

• Promoters (9-10): Loyal enthusiasts who fuel growth through referrals
• Passives (7-8): Satisfied but unenthusiastic, vulnerable to competitors
• Detractors (0-6): Unhappy customers who can damage your brand

NPS = % Promoters - % Detractors

Scores range from -100 to +100. Above 0 is acceptable. Above 30 is good. Above 50 is excellent. Above 70 is world-class. But these benchmarks vary dramatically by industry.

Why Traditional NPS Surveys Fail

1. They capture the "what" but not the "why"

A score of 6 tells you someone is a detractor. It doesn't tell you whether they're frustrated with your pricing, your product quality, your support experience, or something else entirely. The standard follow-up, an open text box asking "Why did you give this score?", gets either one-word answers or nothing at all.

2. Response rates are declining

Email NPS surveys now average 10-15% response rates. In-app surveys do better (25-40%) but still miss the majority of users. Worse, the people who respond tend to be either very happy or very angry, creating selection bias.

3. The data sits in dashboards, not decisions

Even when companies collect NPS data, it rarely reaches the teams who can act on it. Product managers don't see verbatim feedback. Support teams don't know which issues drive detractors. The score becomes a quarterly metric to report, not a tool for improvement.

Building a Better NPS Survey with Koji

Koji solves all three problems by replacing the static survey with a conversational interview. Here's how to build one:

Step 1: Set Up Your Study

Create a new study in Koji and select the "Customer Feedback" objective. Name it something like "Q1 2026 NPS Study" or "Ongoing NPS Program."

Step 2: Design Your Questions

Start with the standard NPS question, then add conversational follow-ups that Koji's AI will explore naturally:

Core NPS Question (Scale type, 0-10): "On a scale of 0 to 10, how likely are you to recommend [product] to a friend or colleague?"

Configure this as a Scale question in Koji with:

• Scale min: 0, max: 10
• Labels: 0 = "Not at all likely", 10 = "Extremely likely"
• Probing: Enable anchor follow-up

Qualitative Follow-Up Questions (Open-ended):

• "What's the primary reason for your score?"
• "What's the single biggest improvement we could make?"
• "Is there anything we do particularly well that you'd like to highlight?"

Segmentation Questions (Single Choice):

• "Which of our products/features do you use most frequently?"
• "How long have you been a customer?"

Step 3: Configure AI Probing

This is where Koji shines. Set probing depth to 2-3 follow-ups per question. The AI will:

• For Detractors (0-6): Probe deeply into specific pain points. "You mentioned the onboarding was confusing, can you walk me through what happened?" The AI follows the thread until it uncovers actionable specifics.

• For Passives (7-8): Explore what's missing. "It sounds like you're generally satisfied but not enthusiastic, what would make you a true advocate?"

• For Promoters (9-10): Understand what drives loyalty. "You clearly love the product, what specifically makes it indispensable for your workflow?" The AI captures testimonial-quality quotes.

Step 4: Distribute and Collect

Share your NPS interview via:

• Email: Send personalized links to your customer base
• In-app embed: Trigger after key milestones (30 days, first project complete, renewal)
• Website widget: Capture feedback from active users
• Import contacts: Upload your CRM list and let Koji manage outreach

Step 5: Analyze and Act

Koji automatically generates:

• NPS score with distribution charts showing your promoter/passive/detractor breakdown
• Thematic analysis of qualitative responses, grouped by score segment
• Key driver analysis identifying which themes most strongly correlate with high or low scores
• Actionable quotes attributed to specific segments
• Trend tracking when you run NPS quarterly

NPS Best Practices

Timing matters

• Relationship NPS: Survey your entire customer base quarterly. This measures overall brand loyalty.
• Transactional NPS: Trigger after specific interactions (support ticket resolved, feature launched). This measures experience quality.
• Don't survey too frequently. Quarterly is the maximum for relationship NPS.

Segmentation is essential

Always analyze NPS by segment:

• Customer tenure (new vs. long-term)
• Plan/tier (free vs. paid vs. enterprise)
• Use case or department
• Geographic region
• Account size

Close the loop

Companies that follow up with detractors within 48 hours recover 50-70% of them. Koji makes this easy by surfacing specific, actionable feedback in each report.

Avoid common mistakes

• Don't incentivize responses. Gift cards for NPS surveys create biased data.
• Don't ask more than 5 questions. Keep the core survey short. Koji's conversational follow-ups replace the need for a long questionnaire.
• Don't average NPS. NPS is a spread metric (% Promoters minus % Detractors), not an average.
• Don't compare across industries. A SaaS NPS of 30 is good. A hospitality NPS of 30 is below average.

Why Koji Is the Best Tool for NPS

Traditional NPS tools (Delighted, Medallia, Qualtrics) collect the score and a text box. Koji collects the score AND has a natural conversation about why. This means:

• 3-5x more qualitative depth per response compared to open text boxes
• Higher completion rates because conversations are more engaging than forms
• Automatic theme detection across hundreds of responses
• Score-segmented insights that connect quantitative metrics to qualitative reasoning
• Both text and voice options so customers can explain in whatever mode they prefer

Sample NPS Interview Flow

Here's what a Koji NPS conversation looks like:

Koji: "Thanks for taking a moment to chat! On a scale of 0 to 10, how likely are you to recommend us to a colleague?"

[Scale widget appears, user selects 7]

Koji: "Thanks for the 7! That puts you in the satisfied-but-not-thrilled range. I'm curious, what would need to happen for that to become a 9 or 10?"

User: "The product is great but your reporting is really limited compared to what I had before."

Koji: "That's really helpful feedback. When you say the reporting is limited, what specifically are you missing? Is it certain chart types, export options, or something else?"

User: "Mainly I can't create custom dashboards. I need to show different views to different stakeholders."

This conversational depth is impossible with a traditional survey form. And Koji does it at scale, hundreds of these conversations running simultaneously, all automatically analyzed and synthesized into actionable insights.