CoreDash vs Request Metrics

Who makes Request Metrics

Request Metrics is made by the people behind TrackJS. Todd Gardner and Eric Brandes have been building monitoring tools for a long time. The product is clean and easy to learn, and if you email support you get the engineers who built it. Their customers bring that up a lot, and from what I've seen of the product I believe it.

I built CoreDash, so keep that in mind. Request Metrics does some things better than CoreDash and I'll point those out. I still think CoreDash is the better choice once you're past the first few months, but for some teams Request Metrics is the right tool.

What Request Metrics does well

Long Animation Frames, mainly. There's a dedicated LoAF dashboard where you can separate frames caused by interactions from frames caused by script evaluation, and break it down per page and per script, all the way to the event listener or setTimeout callback that ran. LoAFs are also drawn into the page load waterfall, which is where they make the most sense. There's even a "Show Code" button that pulls up the inline script so you don't have to open DevTools (it fetches the script as it's live now, so right after a deploy it can be different from the code that actually ran). If you want to know what LoAF is and why it matters for INP, I wrote about the Long Animation Frames API on corewebvitals.io.

requestmetrics.com, October 2026.

It also does more than RUM. Since January 2025 it runs weekly Lighthouse scans on desktop and mobile for your busiest pages, and there's a CrUX dashboard where you can add any public competitor origin. The scans start before you install anything, so there's something to look at on day one. On the Professional plan you also get JavaScript error monitoring, API monitoring, traffic analytics, funnels and session timelines, all from the same script.

Attribution is good. The LCP and CLS elements come as CSS selectors you can paste straight into DevTools, and the agent collects the LCP subparts (resource load delay, load time and render delay) next to TTFB.

Privacy is handled well: no cookies, IP addresses are hashed with a salt and then thrown away, data is stored in Canada, and Enterprise customers can self-host or get a dedicated instance. Installing is easy with a CDN snippet or an NPM module, it's safe for server side rendering, and you can route the collector through your own domain so ad blockers don't eat your beacons. Not many tools help you with that, and lost beacons are a real problem.

Pricing is predictable. There's a free plan, Essentials starts at $34 a month billed annually with unlimited users, and there's never an overage bill.

If you already run Request Metrics

Then you're already measuring your real visitors, and most sites I audit aren't. They work from Lighthouse scores and a Search Console report that only shows URL groups. Any real user monitoring is better than none, and Request Metrics is a good one.

If it answers the questions you ask it, stay. Switching RUM tools costs you your history and a week of attention, so it's only worth it once your current tool stops giving answers. For a lot of teams that happens later, after the easy wins are done and the questions get harder.

CoreDash vs Request Metrics at a glance

Pricing and features checked September 2026.

Where CoreDash pulls ahead later

How long you keep your data

Request Metrics keeps 30 days on Essentials and 90 days on Professional. That's enough to catch a regression from last week's deploy, but not to answer "are we faster than last November?", which someone always asks after Black Friday. CoreDash keeps 6 months on Starter, 1 year on Standard, 3 years on Studio and 8 years on Enterprise.

Traffic limits work differently too. Request Metrics meters your capacity per minute and drops pageviews that go over the limit. They're open about it, it's how they can promise you'll never get a surprise bill, and their docs recommend buying 1.5 to 2 times your expected traffic. Most days you won't notice. On a big sale day the peak gets dropped, and that's the hour you wanted to see.

Request Metrics docs on traffic limits, October 2026.

And the price difference gets bigger with traffic. At around 500,000 pageviews a month Request Metrics Professional is $88 a month billed annually with 90 days of history. CoreDash Standard is $31 with a year of history. To be fair, Professional also gives you error monitoring, funnels and analytics, which CoreDash doesn't do, and unlimited users where Standard has 10. On CoreDash you pay for pageviews and retention, and every plan has every dashboard.

INP phases

Request Metrics tells you which element was the INP target, whether it was a keyboard or pointer event, and whether the page was still loading. Useful, but it stops there. CoreDash also splits every INP into input delay, processing and presentation, and links it to the LoAF script URL, the input type and the load state. The phases tell you what to fix. A long input delay means something else was blocking the main thread when the visitor clicked, usually a tag manager or consent script. Long processing means your own handler is slow. Long presentation means the render afterwards is too big. Two pages with the same INP can need a completely different fix, and often a different team.

I couldn't find an INP phase breakdown in the Request Metrics docs (they don't have an INP guide yet). If INP is your main problem, check that during their trial.

For LCP the difference is smaller, Request Metrics collects the subparts too. CoreDash adds the element type and whether the LCP image was preloaded or lazy loaded, which is one of the first things I check on a slow LCP. The Diagnosis dashboard then gives you the fix for your stack.

Your AI agent can read your field data

For me this is the biggest difference. I couldn't find a public API in Request Metrics to read your performance data. The API in their docs is the browser agent API, for sending data in, and data comes out as CSV exports. They were early with AI though: back in 2023 they launched a ChatGPT-based assistant called Professor Sloth, which lives inside their dashboard.

CoreDash has a read-only API and an MCP server on the same endpoint. The MCP server is included in every plan, Starter too. Add it to Claude as a custom connector, or to Claude Code, Cursor, Windsurf, VS Code or Gemini CLI, and your agent can query your field data itself. Something like: give me p75 LCP on mobile for the checkout, grouped by LCP element, against last week.

I also built CWV Superpowers, a free, MIT-licensed agent. It finds the worst bottleneck in your CoreDash data, runs a Chrome trace on that page and comes back with a code fix (file, line and diff) or an HTML report. It never looks at a Lighthouse score. I wrote about how that works for fixing INP with an AI agent.

A lot of my own debugging starts with an agent asking the field data something. If the data only comes out as CSV, the agent can't ask it anything, no matter how good the dashboards are.

Agencies

Request Metrics gives agencies white-label reports on Professional, which is nice if you want to send clients reports under your own name. CoreDash isn't white label. Instead the Agency API gives you a master key that creates, updates and deletes client projects and reads data across all of them. Pageviews are pooled across clients and there's one invoice. You build your own reporting and CoreDash supplies the data. See agency pricing.

When to pick Request Metrics instead

• You want RUM, JavaScript errors, API monitoring and funnels in one tool with one script. The error monitoring comes from the TrackJS team, who've been doing that for years.
• You spend most of your time on Long Animation Frames and like having a dedicated LoAF dashboard.
• You need white-label reports for clients.
• Your data has to stay in Canada, or you need a self-hosted or dedicated instance.
• Your site is small enough for the free plan, or your team needs unlimited users on the entry plan.
• You want Lighthouse scans and competitor CrUX data from day one, before you install anything.

Try it on your own site

Easiest way to compare is to run both for two weeks. Both scripts load async and don't get in each other's way. The CoreDash trial is 14 days and needs no credit card, and you can install it with one script tag or the WordPress plugin. See pricing, or book a 20 minute demo and I'll go through your own data with you.

For other tools, see all CoreDash comparisons.

Sources: Request Metrics pricing, pricing docs, traffic limits, LoAF guide, LCP guide, CrUX guide, browser agent API reference, end user data, Unified Web Performance launch, Professor Sloth announcement. Agent size measured on rm.js v2.4.1.