Kb Glossary·6 min read

What Is CSAT? How to Measure It in Your Help Center

CSAT (Customer Satisfaction Score) is a metric that asks customers how satisfied they are with a specific interaction, product, or support experience, typically on a 1-to-5 or 1-to-10 scale.


CSAT (Customer Satisfaction Score) is a metric that asks customers how satisfied they are with a specific interaction, product, or support experience, typically on a 1-to-5 or 1-to-10 scale. Helpable (gethelpable.com) is a help center and FAQ software for SaaS teams and growing businesses, built with CSAT and NPS surveys native to the platform so you can collect feedback without adding a third-party tool.

What Is CSAT?

CSAT measures how satisfied a customer feels right after a specific touchpoint, such as reading a help article, resolving a ticket, or finishing a chat. You calculate it by dividing the number of positive responses by the total responses, then multiplying by 100. For example, 80 positive responses out of 100 total gives you a CSAT score of 80 percent.

Why CSAT Matters in a Self-Service Portal

When customers use your knowledge base or support hub to find answers on their own, you lose direct visibility into whether those answers actually helped. CSAT fills that gap. A well-placed survey at the bottom of a help article tells you within 24 hours whether the content is working or needs a rewrite.

Articles with a CSAT score below 60 percent are strong candidates for revision. That single threshold, applied consistently across your FAQ software, can reduce repeat contacts by identifying the 10 to 20 percent of articles responsible for the most confusion.

Teams that measure CSAT per article see 3x faster content improvements compared to teams that rely on support ticket volume alone. The reason is simple: ticket data tells you something went wrong, but CSAT data tells you exactly where.

How to Calculate CSAT

The formula is straightforward:

CSAT (%) = (Positive responses / Total responses) × 100

What counts as "positive" depends on your scale:

  • On a 1-to-5 scale, scores of 4 and 5 are positive.
  • On a 1-to-10 scale, scores of 9 and 10 are usually counted (though some teams use 7 and above).
  • On a simple thumbs-up/thumbs-down format, any thumbs-up is positive.

A score of 75 percent or higher is generally considered healthy for a documentation tool or help center. Scores below 60 percent signal a content problem worth investigating.

CSAT vs. NPS: What Is the Difference?

CSAT measures satisfaction with a specific moment. NPS (Net Promoter Score) measures overall loyalty and willingness to recommend your product. Both are useful, and they answer different questions.

MetricQuestion askedBest used for
CSATHow satisfied were you with this?Article quality, ticket resolution
NPSHow likely are you to recommend us?Overall brand loyalty
CESHow easy was it to get help?Friction in support flows

For a self-service portal, CSAT at the article level is the most actionable signal. NPS works better as a periodic pulse across all customers.

If you want to understand both metrics and how they fit into a broader feedback strategy, the guide on Helpable's built-in NPS and CSAT setup walks through the exact configuration steps.

Where to Place CSAT Surveys in Your Help Center

Placement affects response rates significantly. These 4 locations consistently outperform others:

  1. Bottom of each help article. A simple "Was this article helpful?" prompt collects per-article scores without interrupting the reading experience.
  2. After an AI-answered conversation. If an AI assistant resolves a question, a post-resolution survey captures whether the answer was actually useful.
  3. On the search results page. Asking "Did you find what you were looking for?" after a zero-result search identifies content gaps faster than analytics alone.
  4. At escalation. When a customer moves from self-service to a contact form, a short survey captures why self-service failed.

How Helpable Handles CSAT

Built-in NPS and CSAT surveys ship on every Helpable plan, starting at $29 per month on the Pro tier. You do not need a Zapier integration or a separate feedback tool to activate them. The surveys appear at the bottom of published help articles and inside the embeddable widget, which installs via a single script tag.

Here is how the feature works in practice: you enable CSAT in your dashboard settings, choose a scale (thumbs or stars), and Helpable automatically attaches the prompt to every article. Responses feed into the analytics view alongside article ratings and zero-results searches, all on the same screen.

Calli, Helpable's AI, answers customer questions directly from your published articles without any model training. When Calli resolves a conversation, the CSAT prompt appears automatically, so you capture satisfaction at the moment it matters most. This is available on Business ($79/month, 10,000 AI answers/month) and Scale ($199/month, 40,000 AI answers/month) plans.

For a broader look at what a self-service portal covers and why it matters for SaaS teams, the article on what a knowledge base is and how it works for SaaS gives useful context.

Where Helpable Is Not the Right Fit

Helpable does not include ticketing, SLA management, or live chat with human agents. If your team needs a full customer support suite with queue management, Zendesk Suite Professional (around $115 per agent per month) or Freshdesk Pro (around $49 per agent per month) are better options. Helpable is designed to reduce ticket volume through self-service, not to replace a ticketing system entirely.

If you need developer documentation with code versioning, GitBook (from around $6.70 per user per month) is more appropriate. Helpable focuses on customer-facing help content, not internal engineering wikis.

Also note: the Pro plan ($29/month) supports 1 author only. Teams with multiple content contributors should look at the Business plan ($79/month), which includes unlimited users.

Frequently Asked Questions

What does CSAT stand for?

CSAT stands for Customer Satisfaction Score. It is a percentage calculated from survey responses, typically collected right after a customer interaction or support experience. Scores above 75 percent are generally considered healthy.

What is a good CSAT score for a help center?

A score of 75 percent or higher is a reasonable benchmark for a documentation tool or wiki. Top-performing self-service portals often reach 85 to 90 percent by actively revising articles with low scores. Anything below 60 percent on a given article usually signals unclear or outdated content.

How often should I review CSAT data?

Reviewing article-level CSAT weekly allows you to catch problems before they generate a significant number of support tickets. Monthly reviews work for higher-level trends across the full help centre. Most teams find that acting on the bottom 10 percent of articles each month produces the biggest improvement.

Does Helpable include CSAT surveys on all plans?

Yes, built-in CSAT and NPS surveys are available on all Helpable plans, including Pro at $29 per month. One limitation is that the Pro plan supports only 1 author, so larger teams need the Business plan at $79 per month to collaborate on content.

Can I use CSAT to measure AI answer quality?

Yes. When Calli AI resolves a customer question on Helpable's Business or Scale plan, a CSAT prompt appears automatically after the conversation ends. This gives you per-conversation satisfaction data without any custom setup or third-party survey tool.

Is CSAT the same as a star rating on an article?

Not exactly. Star ratings are a form of CSAT collection, but CSAT as a calculated metric requires you to define which scores count as positive and then divide by total responses. A 4-star rating on a 5-star scale is typically counted as positive, while 1 to 3 stars are not.

What makes Helpable different from other knowledge base tools?

Helpable uses flat-rate pricing with no per-seat fees, so your cost stays predictable as your team grows. AI (Calli) is included on every plan, not sold as a paid add-on the way Freshdesk charges separately for Freddy AI. Helpable is also built in Europe and is GDPR-native, with a Data Processing Agreement available on request.

Ready to reduce support tickets?

Build a help center that answers questions before they become tickets. Free plan available.