Gong : So Gong in der Praxis (2026)

Häufige Herausforderungen bei der Verwendung von Gong

Gong’s main drawbacks are cost and setup. It isn’t cheap and it isn’t quick to get running, and its AI works best with someone checking the output. None of that makes it a bad product, but it’s worth knowing before you sign.

Hohe Kosten und versteckte Preisgestaltung

Gong doesn’t publish prices. Its pricing page says licenses are priced per user, plus a platform fee based on how many users you have, and asks you to request a quote. Third-party reports put the license at roughly $1,200 to $1,600 per user per year, with platform fees reported between $5,000 and $50,000 a year. Those are estimates, not Gong’s figures. Our Gong pricing guide breaks down what’s been reported.

Contracts run annually. Gong’s own documentation talks about “contract years”, and credit allowances reset at the start of each one.

How Gong Credits Work

Since June 2026, Gong meters its heavier AI features through Gong Credits. Each paid core seat adds 2,000 credits a year to a pool shared across the company. Here’s what uses them, according to Gong’s documentation:

Processing costs 0.1 credits per email, 0.3 credits for a call up to 10 minutes and 1 credit for anything longer. When the pool runs out, credit-based features stop processing until you buy more, and existing data stays where it is. Admins can set monthly caps per user and per feature. Gong says there are no automatic charges, so the risk is features pausing mid-quarter, not a surprise invoice.

Komplexe Einrichtung und steile Lernkurve

Gong isn’t plug-and-play. Before it’s useful, an admin has to connect calendars and your CRM, set recording rules, and decide who gets permissions for features like Deep Mode and Enrich. Once credits are involved, someone also has to manage the pool.

The reporting takes learning too. Dashboards and trackers only pay off if managers know which signals to act on, so budget time for training as well as setup.

Grenzen der KI von Gong

Gong’s AI is capable, but it works from what’s in your calls and your CRM, and it’s only as good as that data. Gong’s own design choices reflect this. Deep Mode makes you approve a research plan before it runs and cites its sources, and custom trackers pause when credits run out rather than carrying on silently.

Coaching is the clearest example. Gong’s interaction stats and AI Call Reviewer show where a rep talks too much or asks too few questions, but turning that into better selling is still a manager’s job. If coaching is the main reason you’re looking at Gong, our guide to sales coaching covers what the software can and can’t do on its own.

There’s a bigger debate underneath this. Gong’s CEO has argued that “have AI update my CRM” is a feature that won’t last, because AI will eventually read the transcripts directly and CRM fields won’t be needed. It’s a bold pitch, and it conveniently positions Gong as the system of record rather than a tool that feeds one.

The pushback is practical, and we’re inclined to agree with it. A CRM field forces a rep to stop and commit to a number. AI summaries of a long call can miss context. And a clean overview screen is still what most people use to make decisions. Querying transcripts is a great addition. It’s not yet a safe replacement for the structured record, which is why tools that work with your existing CRM, rather than trying to become it, are the safer bet for now.

Compliance- und Datenschutz-Risiken

Gong holds the certifications you’d expect from an enterprise vendor: SOC 2 Type II (including HIPAA), ISO 27001, 27017, 27018 and 27701, and certification under the EU-U.S. Data Privacy Framework. It offers a GDPR data processing agreement with standard contractual clauses.

Two details matter for European teams. Gong stores data in US data centers by default, with an EU option, and both run on AWS. It processes data in the United States, Israel and Ireland. If EU residency matters to you, ask for it during procurement rather than assuming it.

Gong’s certifications don’t make your recording compliant. You still need a lawful basis for recording, you have to tell participants, and you need to set retention rules that fit your policies. In healthcare or other regulated sales, check your industry rules before the first call is recorded.

It Can Be Overkill for Smaller Teams

Gong’s value comes from patterns across many calls: which talk tracks win, which deals stall, where forecasts drift. With a handful of reps, there aren’t enough calls for those patterns to mean much, and you’re paying a platform fee for analytics you won’t use.

If what you mainly need is recordings, searchable transcripts and solid notes, a lighter tool will cover it for far less. Gong earns its price when you have the call volume, and someone in RevOps to act on what it finds.

Verbessert die KI von Gongtatsächlich den Umsatz?

Gong’s customers report strong results, but the published numbers come from Gong. At its September 2026 Celebrate event, Gong said AT&T Business saw a 54% improvement in rep productivity, Cisco saw 32% larger deals and 26% higher win rates among early-adopting teams, and Experian saw a 25% lift in win rates. Treat those as vendor-reported figures from the companies most invested in making Gong work.

Independent reviewers are more mixed. G2 reviewers rate Gong highly overall, but some report recordings that fail to capture and data that’s hard to export. Gong improves sales when a team uses its insights consistently. It doesn’t improve them just by recording calls.