Skip to content
New storyWe built Skald to count visitors. It caught an AI-key thief.
SkaldEvery visit, truly told.

Compare · Plausible

Skald vs Plausible: a light script tag beside the whole server log

Skald vs Plausible is not a fight. Plausible counts people with a small, cookie-free script. Skald reads your server's log, so it also sees the AI crawlers, assistants and scanners a script never meets.

Skald vs Plausible is a choice between two ways of counting. Plausible runs a JavaScript snippet in each page and counts the browsers that run it. Skald reads the access log your server already writes, so every request is counted, people and machines, in five groups that add up to the log. Both can run on your own server.

30-day free trial · no card · we set it up within one working day · free to self-host

one minute · samplewho saw it
  1. 09:14:02 198.51.100.7 "GET /pricing/" ref "google.com"both see it
  2. 09:14:05 198.51.100.9 "GET /docs/" · ad-blocker onlog only
  3. 09:14:06 203.0.113.24 "GET /pricing/" "GPTBot/1.2"log only
  4. 09:14:08 198.51.100.140 "GET /docs/setup/" "ChatGPT-User"log only
  5. 09:14:11 192.0.2.77 "GET /.env" 404log only

Sample lines. A script tag counts the browsers that run it. The server log has every request.

[01] question by question

Skald vs Plausible, question by question

Pick what you need to know. The table lights those questions and shows how each tool answers them. Every comparison is on Skald alternatives and comparisons.

What do you need to know? pick one

0 of 5

AI and machines: Plausible is built to count people and filters machines out, so it answers none of these 5. Skald answers all 5 from your server log.

Skald and Plausible compared on 21 questions, grouped by what you need to know
The question PlausibleJavaScript snippet Skaldyour server log
Which AI training crawlers read your pages? (GPTBot, ClaudeBot)No, filtered out, not reportedYes, page by page
Did an assistant fetch a page live, for someone? (ChatGPT-User)NoYes, kept apart from training crawls
Why did the AI come: training, a live question, or a click?NoYes, three reasons, never added together
Which search engine crawlers came?No, excludedYes, in their own group
Who is probing for .env and wp-login?NoYes, with what your server answered
Visitors and visitsYesYes, as an upper bound on people
People who clicked through from an AI assistantYes, an AI Assistants channel since June 2026Yes, by assistant; a floor, as apps strip referrers
Visitors who decline cookiesYes, no cookiesYes, the log needs no cookie
Visitors with an ad-blockerA 5 to 25% gap without its proxy, by its own estimateYes, the server logs every request
Works where you cannot read the server's logYes, one snippet on any hostPeople only, with the script tag; no AI or bots
Search queries, impressions, positionThrough its Search Console integrationSynced from Search Console and Bing every 6 hours
Each visit's pages in orderNot in the free Community EditionYes
Events inside the page (clicks, sign-ups)Yes, goals and eventsYes, with the optional script tag
FunnelsOn its Business planYes, counted every day by source and device
Leads by source and AI assistantGoals by sourceYes, first and last touch, from signed webhooks
Retention cohortsNoYes, by week, split by first source
Totals must add up, or the day is withheldNoYes
A second, separate recount of every figureNoYes, every day
Runs on your own serverCommunity Edition, with fewer featuresYes, free to self-host, every feature
Raw requests and addressesNo, it never stores IP addressesYes, if you choose to keep them
Email alertsYes, traffic spikes and dropsYes, site, AI and security alerts

What each tool is built to do, as we read their own docs in October 2026. If we have yours wrong, tell us.

[02] to be fair

Where Plausible does more than Skald

Plausible is a good tool for what it is built to do. Here is where it is the better pick, so you can choose with open eyes.

  1. Any host, one snippet

    Plausible works wherever you can paste a script: a site builder, a static host, a shop. Skald needs a server or CDN log to see machines. Without one, its script tag sees people only.

  2. Nothing to run

    Plausible's cloud needs no server at all. Skald runs on a server: yours, or ours if we host it for you.

  3. No addresses, by design

    Plausible never stores IP addresses. Skald can keep them, so you can look a person up; you choose full, shortened or none.

[03] what the log adds

What Skald sees that a script tag never meets

Crawlers, AI fetchers and scanners never run a page's JavaScript. Your server still writes a line for each of them, and Skald files every line in one of five groups.

People, AI assistants, search engines, bots and scanners, and your own traffic. The five add up to the log, and a second count checks every figure.

access.log · sample · what Skald files5 groups
  1. people 198.51.100.23 "GET /docs/setup/" ref "chatgpt.com"from ChatGPT
  2. AI 198.51.100.140 "GET /docs/setup/" "ChatGPT-User"live fetch
  3. search 203.0.113.66 "GET /blog/" "Googlebot"verified
  4. bots 192.0.2.200 "GET /" "GPTBot/1.2"claims to be GPTBot
  5. ours 192.0.2.10 "GET /pricing/"your office

[04] use both

Keep Plausible for people, add Skald for the rest

The two never fight over a visitor. Plausible's script and Skald's log reader can run on the same site, and each answers its own questions.

  1. Keep Plausible for

    Simple, cookie-free people counts from one snippet, and sites where you cannot read the server's log.

  2. Add Skald for

    AI crawlers and live fetches, what each crawler cost your server, who is probing for secret files, and leads joined to the pages they read.

  3. Trust both

    Skald's figures add up to the log and are counted twice. A day that does not agree is withheld, never guessed.

[05] choose by the question

Plausible, Skald, or both

  • "How many people came, on a host with no log?" Plausible.
  • "Which AI crawlers and assistants read us?" Skald, from the log.
  • "Did the page ChatGPT read bring a lead?" Skald.
  • "What did people search for?" Search Console, synced into either.

What it does not do

  • Without a server or CDN log (some site builders, beacon only), Skald sees people only, through its script tag: no AI crawlers, fetchers or bots.
  • Visits are an upper bound on people, never an exact count of humans.
  • AI clicks are a floor: an app that strips the referrer sends a direct visit. Likely visits are shown apart, never added.
  • Skald runs on a server. If you want nothing to run, Plausible's cloud is simpler.

questions

Skald vs Plausible FAQ

Straight answers, including what Skald does not do.

Is Skald a Plausible alternative?

For counting people, both work, and Plausible is simpler to add to any host. Skald is the one to pick when you also need what a script never sees: AI crawlers, AI assistants fetching your pages, search engine crawlers and scanners. Many sites keep Plausible and add Skald beside it.

Does Plausible show AI crawlers like GPTBot?

No. As we read Plausible's own docs in October 2026, it filters bots and AI crawlers out and does not report them. It does show people who clicked through from an AI assistant, in an AI Assistants channel added in June 2026. Skald reads the server log, so it shows the crawlers, the live fetches and the click-throughs, each kept apart.

Do both work without cookies?

Yes. Plausible's script sets no cookies. Skald counts from the server log, which needs nothing in the browser. Skald's optional script tag sets no cookie by default either. Signing in to Skald's dashboard uses one session cookie, for your own team.

Which one counts visitors with an ad-blocker?

Skald, because the server writes a log line whatever the browser blocks. Plausible's own docs estimate a 5 to 25% gap when its script runs without its proxy.

Can I self-host both?

Yes. Plausible's Community Edition is AGPLv3 and has fewer features than its cloud. Skald is source-available under the Elastic License 2.0 and free to self-host, with every feature.

Which one should I use?

Plausible if you want simple, cookie-free people counts on any host, including hosts with no server log. Skald if you want every visitor, crawler and AI fetch from your own log, with checked figures and leads joined to their visits. Both together if you want both views.

$ tail -f access.log | skald

See what AI does with your site. Free for 30 days.

No card. We set up hosted Skald for you within one working day, or you self-host it free. Every line of your log, sorted and added up.

Start free trial Live demo