CoreDash vs RUMvision

Two similar tools

RUMvision and CoreDash have more in common with each other than with most tools on the comparison list. Both do real user monitoring for Core Web Vitals and nothing else, built on Google's web-vitals library. Both are Dutch, keep the data in the EU and don't need a cookie banner because they don't collect personal data. If you're choosing between these two you've already made the important decision: you've stopped waiting for a 28 day CrUX average and want to see what your visitors get today.

I built CoreDash, so keep that in mind. I'll still try to be fair to RUMvision. It's good at what it does, and a comparison that pretends otherwise isn't much use to you.

What RUMvision does well

Attribution goes deep and it's on by default. LCP is split into TTFB, resource load delay, load duration and render delay, INP into input delay, processing and presentation delay, and there are breakdowns for TTFB and FCP as well. You get element selectors for LCP, INP and CLS that you can paste straight into DevTools, and Long Animation Frames data with the script, the function and the invoker type. The RUMvision team has done real research on LoAF and INP (their work was mentioned at Google I/O 2024).

rumvision.com, October 2026.

My favourite part is the Third Parties dashboard. It turns the LoAF data into a Good, Moderate or Poor rating for every third party host, plus a first party versus third party score. That's the screen you want open when marketing asks for one more tag.

Health Checks look for common problems by themselves: lazy loaded LCP images, missing fetchpriority, query parameters that bust your cache, low bfcache usage, layout shifts per element and slow interactions per element or host. Each one becomes a to-do that closes itself once the fix is live.

There are 45+ dimensions, some quite unusual (battery level, ad blocker, in-app browser, CDN location). Custom timings are well covered, with Server-Timing guides for the big CDNs plus Element Timing and User Timing. Single page apps get soft navigation support. And there's a remote MCP server with 13 tools and access control per role.

Every plan also comes with a setup call and monthly support and consultancy. If your team doesn't have its own performance engineer, having the people behind that LoAF research look at your numbers is worth a lot.

Happy with RUMvision? Stay

If you run RUMvision today and you're happy with it, stay. You have field data per page, per element and per release, so you're ahead of most sites I look at, where all anyone has seen is a Lighthouse score and a Search Console report that lumps pages into groups. Switching costs you your history and a few weeks of getting used to new dashboards, and if the tool does what you need that's not worth it. The comparison below is more useful if you don't have RUM yet or your contract is up for renewal.

CoreDash and RUMvision side by side

Pricing and features checked September 2026.

On attribution it's pretty much a tie. RUMvision has more dimensions, and both give you the phases and elements you need to figure out why an LCP or INP is slow. Most of the differences only show up after you've used the tool for a year or two.

Where they differ

From field data to a diff

I think this one will matter most a couple of years from now. More and more performance fixes will start in an AI coding agent, and the agent can only work with the field data it can read. Both tools have an MCP server, and to be fair RUMvision's has more tools than mine. What CoreDash adds is CWV Superpowers, a free, MIT licensed agent for Claude Code, Cursor, Windsurf, VS Code and Gemini CLI. It pulls your CoreDash field data, finds the page and the phase that cost the most time on mobile, checks it with a throttled Chrome trace and writes the fix as a diff with file and line in your own codebase. Or an HTML report if you'd rather hand it to someone. I wrote about how that works for LCP and for INP.

The MCP server is also the API. Same endpoint, same three tools, so a curl command, a CI job, a reporting script and Claude all get the same numbers, at any percentile from p50 to p95, grouped by any dimension. AI Auto Alerts use the same data. You turn them on with one toggle, and after that they check hundreds of segments every day and only flag regressions that hit 2% or more of your traffic.

Agencies get the Agency API: one master key that creates and manages client projects and reads any client's data, on public agency plans with pageviews pooled across clients. RUMvision works with agencies through a partner programme (there's no public agency plan), and its per-domain user roles work well for client work. In the pipeline, CoreDash runs a pre-deploy check against your performance budgets and tags every pageview with the release, so you see the p75 of each release next to the previous one.

You pay for the traffic you have

RUMvision starts at €150 a month and CoreDash at $16. That's not a fair comparison by itself, because their Growth plan includes 1.5 million pageviews and CoreDash Starter 75K. At several million pageviews the difference gets a lot smaller. Where you notice it is on sites that are still growing, and most sites don't start at 1.5 million pageviews a month. A shop doing 300K pays $31 a month for Standard (billed yearly), gets every dashboard, and moves up a tier when traffic does.

Domains add up as well. RUMvision counts subdomains, ccTLDs, checkout domains and staging as separate domains, at €20 a month each on top of the two in Growth (with a reduced fee for the same stack in another language). CoreDash Studio covers up to 10 sites with 1 million pooled pageviews for $71 a month billed yearly. So a brand with a .nl, .de and .be shop plus a checkout subdomain doesn't have to count domains.

Whatever tool you pick, ask what happens when you go over your quota. RUMvision stops tracking by default, so you have a gap for the rest of the month unless you turn on autoscaling. That protects you from a surprise invoice, but the month you go over is usually the month with the sale in it.

How long you keep history

RUMvision states 13 months of retention on Scale and doesn't list a number for Growth anymore. CoreDash keeps 1 year on Standard, 3 years on Studio and up to 8 on Enterprise. With thirteen months you get one year-over-year comparison. You won't see whether INP has been slowly creeping up for two years while more tags got added, or how this Black Friday compares to the last three. That kind of slow drift only shows if the older data is still there.

Every page, no setup

Out of the box RUMvision tracks the homepage URL you registered. Everything else needs URL or template rules (a fixed path, a regex or a CSS selector) or a catch-all. That conditional loader is also why their CSP guidance asks for unsafe-inline and unsafe-eval unless you track every page, so check that with your security team first. CoreDash measures every URL from the first pageview, nothing to configure. The landing page marketing launches next spring is in your data on day one.

RUMvision help center on URL rules, October 2026.

When to pick RUMvision instead

• You want people as well as a tool. Every plan comes with monthly consultancy from a team that really knows LoAF. CoreDash plans include support, but no consultancy slot.
• You need Server-Timing from your CDN, Element Timing or dimensions like battery level and ad blocker.
• Your alerts have to go to Teams, Discord or Zapier.
• The people reading it are mostly SEO, marketing or management. Weekly Monday reports, CrUX competitor benchmarks and Health Checks as a to-do list work well for them.
• You want the bigger MCP toolset: access per role, and marking issues as fixed from your AI client.

Try both on your own site

The easiest way to decide is to run CoreDash next to what you have for two weeks. The 14 day trial needs no credit card and the script is under 2KB, so running both doesn't hurt. Both use the web-vitals library, so the numbers should match as long as you compare the same percentile. RUMvision defaults to p80 and CoreDash to p75, and at p75 every number looks a bit better than at p80.

Prefer a walkthrough first? Book a 20 minute demo.

Sources: RUMvision pricing, RUMvision features, RUMvision MCP, RUMvision URL rules, RUMvision pageview quota, RUMvision snippet and CSP.