Compare · Cloudflare
Skald vs Cloudflare Web Analytics: edge counts beside a sorted log
Skald vs Cloudflare Web Analytics is not either-or. Cloudflare counts people with a beacon and AI crawlers at its edge, and can charge them. Skald reads the log behind it, checks every figure twice, and joins leads to their visits.
Skald vs Cloudflare Web Analytics compares two views of the same traffic. Cloudflare Web Analytics counts people with a JavaScript beacon, and AI Crawl Control counts and rules AI crawlers at Cloudflare's edge. Skald reads your server's log, or Cloudflare's own, and sorts every request into five groups that add up to the log.
30-day free trial · no card · we set it up within one working day · free to self-host
- 09:14:02 198.51.100.7 "GET /pricing/" · cache hitbeacon and log
- 09:14:05 198.51.100.9 "GET /docs/" · beacon blockedlog only
- 09:14:06 203.0.113.24 "GET /pricing/" "GPTBot/1.2"Crawl Control and log
- 09:14:07 192.0.2.200 "GET /" "GPTBot/1.2"log: claims to be GPTBot
- 09:14:09 192.0.2.77 "GET /.env" 404log: scanner
Sample lines. Skald reads Cloudflare's own log, or the Skald Worker posts each request, cache hits included.
[01] question by question
Skald vs Cloudflare Web Analytics, question by question
Cloudflare here means two products: Web Analytics, a beacon in the page, and AI Crawl Control, at its edge. Pick what you need to know. Every comparison is on Skald alternatives and comparisons.
What do you need to know? pick one
AI and machines: Cloudflare answers 4 of these 7 in full and 3 in part, some only on higher plans, and it can charge AI crawlers. Skald answers 6 in full; it blocks but does not charge.
| The question | Cloudflarebeacon · AI Crawl Control at the edge | Skaldyour server log, or Cloudflare's |
|---|---|---|
| Which AI training crawlers read your pages? (GPTBot, ClaudeBot) | Yes, AI Crawl Control, all plans, if proxied | Yes, page by page |
| Did an assistant fetch a page live, for someone? (ChatGPT-User) | Yes, by crawler name | Yes, kept apart from training crawls |
| AI fetches per page | Yes, top paths per crawler | Yes, with what each fetch asked for |
| Block or charge AI crawlers | Yes, allow and block rules; pay per crawl | Blocks by your web server's rules; does not charge |
| Is that crawler who it claims to be? | Below Enterprise, AI crawlers are told apart by user agent | Yes, its owner's published addresses or DNS |
| Which search engine crawlers came? | Bot Analytics, Business and Enterprise plans | Yes, in their own group |
| Who is probing for .env and wp-login? | Bot Analytics, Business plan and up | Yes, with what your server answered |
| Visitors and visits | Yes | Yes, as an upper bound on people |
| Countries | Yes | Yes, and cities from a GeoIP file |
| Visitors who decline cookies | Yes, no cookies or local storage | Yes, the log needs no cookie |
| Pages a CDN served from its cache | Yes | Yes, from Cloudflare's log or the Skald Worker |
| People who clicked through from an AI assistant | AI referrers in AI Crawl Control, paid plans only | Yes, by assistant; a floor, as apps strip referrers |
| Works where you cannot read the server's log | The beacon works anywhere; Crawl Control needs traffic through Cloudflare | People only, with the script tag; no AI or bots |
| Visitors with an ad-blocker | No, its FAQ says the beacon is blocked | Yes, every request is in the log |
| Each visit's pages in order | No | Yes |
| Events inside the page (clicks, sign-ups) | No | Yes, with the optional script tag |
| Funnels | No | Yes, counted every day by source and device |
| Leads by source and AI assistant | No | Yes, first and last touch, from signed webhooks |
| 404s, and who met them | No status-code dimension | Yes, people apart from probes and crawlers |
| Server errors (5xx) by page | No | Yes |
| Page speed | Yes, measured in the browser | Yes, Google PageSpeed for top pages, and server time |
| Unsampled | Adaptive sampling; full detail for 7 days | Yes, every line counted |
| Raw requests and addresses | A GraphQL API, sampled, without the IP | Yes, if you choose to keep them |
| Totals must add up, or the day is withheld | No | Yes |
| A second, separate recount of every figure | No | Yes, every day |
| Runs on your own server | No, Cloudflare's network | Yes, free to self-host |
| Several sites in one place | Yes, a soft limit of 10 sites | Yes, with a site switcher |
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 Cloudflare does more than Skald
Cloudflare sits in front of your site, so it can act before a request reaches your server. Here is where it is the better tool.
Charge AI crawlers
Pay per crawl lets a crawler pay for access. Skald can only allow or refuse a crawler; it never charges one.
Rules at the edge
AI Crawl Control's allow and block rules act at Cloudflare's edge for proxied traffic, before your server answers.
A beacon with no storage
Web Analytics uses no cookies or local storage, and the beacon works on any host.
[03] already on Cloudflare
Skald reads Cloudflare's log, cache hits included
A page Cloudflare serves from its cache never reaches your server, so it is not in your server's log. Skald reads Cloudflare's own log instead, or the Skald Worker posts each request after the page goes out.
The Worker works on any Cloudflare plan and deploys with one button. Then every request is sorted into the same five groups, and counted twice.
- cache hit 198.51.100.7 "GET /pricing/" 200people
- cache hit 198.51.100.140 "GET /docs/setup/" "Perplexity-User"AI · live fetch
- origin 203.0.113.66 "GET /blog/" "Googlebot"search · verified
- origin 192.0.2.77 "GET /wp-login.php" 404scanner
[04] use both
Keep Cloudflare at the edge, add Skald for the answers
If your site is behind Cloudflare, you need not choose. Cloudflare acts on requests. Skald explains them, and ties them to leads.
Keep Cloudflare for
Edge rules for AI crawlers, pay per crawl, and a beacon with no storage on any host.
Add Skald for
AI traffic split by why it came, crawler names checked, scanners and attack attempts, 404s and errors, leads and their journeys.
Check the figures
Every Skald figure is unsampled, adds up to the log and is counted twice. A day that does not agree is withheld.
[05] choose by the question
Cloudflare, Skald, or both
- "Make AI crawlers pay to read us." Cloudflare's pay per crawl.
- "Is that GPTBot real, or a fake?" Skald.
- "Did a page an assistant read bring a lead?" Skald.
- "Who probed for secret files, and did they get one?" Skald's Security view.
What it does not do
- Skald never charges crawlers, and its rules act at your web server, not at Cloudflare's edge.
- Crawler rules go by the name a crawler sends. A scraper behind a browser name is not stopped.
- Skald shows who probed your server; it blocks nothing itself. Its block list is for you to review and apply.
- Without a server or CDN log, Skald sees people only, through its script tag.
keep reading
More on Skald vs Cloudflare Web Analytics
questions
Skald vs Cloudflare Web Analytics FAQ
Straight answers, including what Skald does not do.
Does Cloudflare Web Analytics show AI crawlers?
Its Web Analytics beacon does not, because crawlers never run it. Cloudflare's AI Crawl Control does: as we read Cloudflare's own docs in October 2026, it shows AI crawlers on all plans when traffic passes through Cloudflare, with top paths per crawler. Skald shows them from the log, checked against their owners' published addresses.
Can Skald charge AI crawlers like Cloudflare's pay per crawl?
No. Skald can block AI crawlers, per AI company or only their training crawlers, through your web server's rules. It does not charge them. If you want to charge crawlers, use Cloudflare.
I'm on Cloudflare. Can Skald read its traffic?
Yes. Skald reads Cloudflare's Logpush logs, cache hits included, or you deploy the Skald Worker, which works on any Cloudflare plan and posts each request after the page goes out. Cloudflare log analytics with Skald.
Does Cloudflare Web Analytics use cookies?
No. Cloudflare says its beacon uses no cookies or local storage. Its own FAQ also says the beacon is blocked by ad-blockers. Skald counts from the log, so it needs no cookie and an ad-blocker does not hide a visit.
Is Cloudflare's data sampled?
Cloudflare Web Analytics uses adaptive sampling and keeps full detail for 7 days, by its own FAQ. Skald counts every log line, unsampled, and withholds a day whose two counts disagree.
Should I use Cloudflare or Skald?
Both, if your site is behind Cloudflare. Keep Cloudflare for rules at its edge and pay per crawl. Add Skald for checked figures, AI traffic split by why it came, security from the log, and leads.
$ 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.