Skip to main content

Trends, tech news, and tool comparisons.

How to set up error monitoring for an indie SaaS (4 tools compared)

· 8 min read
Dealpuffin
Proof points and live coupons for early adopters

To set up error monitoring for an indie SaaS, pick one tool with a free tier, install its SDK in both your browser code and your server code, upload source maps so stack traces are readable, and send alerts to one channel you actually check. Sentry, GlitchTip, PostHog and BugSnag all have free plans, and the main differences are how many users, events and days of history you get.

Limits and prices below come from each vendor's pricing page, read on 15 November 2026. They change often, so check the page before you commit.

What error monitoring does for a small product​

Without it, you learn about a crash when a customer emails you, and many customers never do. An error monitor catches uncaught exceptions in your app, groups repeats into one issue, and shows the stack trace and context. PostHog describes the same job as turning exceptions into issues you can prioritize, assign and resolve.

This is different from uptime monitoring, which only tells you whether a URL answers. A page can return 200 and still fail for every user whose card is declined, and only an error monitor will show you that.

The four tools at a glance​

ToolFree planPaid entry pointWhat happens past the limit
SentryDeveloper: $0, 1 user, 5k errors, 30-day lookbackTeam: $26/mo billed annually, 50k errors, unlimited usersPay-as-you-go rates above quota
GlitchTip1,000 events/mo, 30 days retentionSmall: $15/mo, up to 100k events, 90 daysThrottles by default; optional on-demand billing with a spend cap
PostHog100K exceptions/mo free, 1 project, 1-year retentionPay-as-you-go, usage-basedFree plan drops extra events; billing limit stops spend
BugSnagFree: 1 user, 7.5K events/mo, 7 days retentionSelect from $20/mo (50k events)Covered in their quota FAQ; check it before relying on the plan

Sources: Sentry pricing, GlitchTip pricing, PostHog pricing, BugSnag pricing.

Sentry​

Sentry's Developer plan is free and limited to one user, with 5k errors a month and a 30-day lookback. The Team plan is $26 a month when billed annually with default pre-paid data, adds unlimited users and includes 50k errors. Above the quota you pay per error, starting at $0.0003625 per error between 50K and 100K on Team.

For a Next.js app, the docs give one command, npx @sentry/wizard@latest -i nextjs, which installs the SDK and creates separate init files for the browser, the Node.js runtime and the Edge runtime. It also wraps your Next.js config so source maps upload at build time.

The catch for a two-person team is the single user on the free plan. The moment a cofounder needs a login, you are on a paid plan.

GlitchTip​

GlitchTip calls itself an open source, Sentry API compatible error tracking platform, and it offers a hosted version and a self-hosted one. The hosted free plan covers 1,000 events a month, where each error and each performance transaction counts as one event. That is small, so it suits a very early product or a side project.

The Small plan is $15 a month for up to 100k events, with unlimited projects and team members. If you go past your limit, GlitchTip throttles by default, starting at dropping about 10% of events and blocking fully at twice your limit. You can turn on on-demand billing at $0.17 per 1,000 events for the first million over, with a monthly spend cap you set.

Because it is Sentry API compatible, it is also the tool to try if you want your data in the EU. GlitchTip lists hosting in Germany as available on all plans.

PostHog​

If you already use PostHog for product analytics, error tracking is the lowest-effort option. The free tier includes 100K exceptions a month, unlimited team members, one project and one year of retention. The practical benefit is context: PostHog says each issue comes with the affected user's session replay, events and properties attached.

Two limits matter. Billing is organization-wide, so extra projects share one free allowance. And on the free plan, usage above the limit is permanently dropped, so a bad deploy that throws a million exceptions will cost you visibility for the rest of the month. Add a card and set a billing limit if you want to keep capturing.

BugSnag​

BugSnag's free plan covers one user with 7.5K events and 1M spans a month, and its pricing page notes the free tier keeps 7 days of data. The paid Select plan starts at $20 a month, with event packs starting at 50k monthly events. Its feature list leans toward stability tracking, such as release health and stability scores. That helps once you ship mobile apps or have several releases in the field. For a single web app it is more than you need on day one.

A decision rule​

Answer these in order and stop at the first yes.

  1. Do you already run PostHog for analytics? Use its error tracking. One SDK and one bill is worth more than a feature gap at this stage.
  2. Do you need a second person to log in today? Skip Sentry's free plan and compare GlitchTip Small at $15 a month against Sentry Team at $26 a month.
  3. Do you need to self-host or keep data in the EU? Look at GlitchTip first.
  4. Otherwise start with Sentry's free plan. You can move later, because GlitchTip says it is compatible with the Sentry API.

A worked example​

Say you run a Next.js app and expect a few hundred errors a month while you have under a hundred users. That fits inside the Sentry Developer plan's 5k errors, inside PostHog's 100K exceptions, and inside BugSnag's 7.5K events. It can overflow GlitchTip's free 1,000 events if you also send performance transactions, because those count as events too. Your real constraint is how fast a bug loop can burn quota. One retry loop that logs an error every second produces 86,400 events a day. Rate-limit or sample your reporting before that happens, and see how to cap AI API costs in a SaaS for the same idea applied to spend.

Setup checklist​

  1. Create the project and install the SDK for both client and server code.
  2. Add the framework hook that catches server render errors. For Next.js, Sentry's guide uses instrumentation.ts with an onRequestError export and a global-error.tsx for React rendering errors.
  3. Upload source maps in your build. In CI, set SENTRY_AUTH_TOKEN in your build system instead of committing it. If you deploy Next.js on a single VM with Docker, pass it as a build argument or build secret.
  4. Throw a test error from a button and confirm it appears. Sentry's guide notes that errors triggered from the browser console are sandboxed and will not show up.
  5. Set one alert, sent to the place you already read. A loud channel you mute is worse than none.
  6. Tag each error with the release, so you can tell whether a deploy caused it.
  7. Review what the SDK sends. Sentry's docs say it can include IP address, user ID and request data by default and lets you turn categories off, which matters for privacy policies.

Frequently asked questions​

Which error monitoring tool is best for a solo founder?​

Sentry's free Developer plan is a common starting point and has a one-command setup for Next.js. If you already use PostHog, its error tracking avoids adding another vendor. Both are free for one person at small volume.

Is Sentry free?​

Sentry has a free Developer plan limited to one user, with 5k errors a month and a 30-day lookback. Paid plans start at $26 a month on the Team plan when billed annually.

Do I need error monitoring before I have users?​

You need it as soon as a stranger can use your app, because they will not report bugs for you. Install it before launch so the first crash is already recorded.

What happens if I go over my free error limit?​

It depends on the tool. PostHog's free plan drops events above the limit, GlitchTip throttles by default, and Sentry bills pay-as-you-go rates on paid plans. Check the overage rule before you rely on a plan.