Fair Use vs Metered Billing: How Analytics Pricing Should Handle Traffic Spikes
Here's a scenario every site owner eventually hits. Your launch post lands on Hacker News, a tweet gets picked up, or a newsletter features you. Traffic goes 20x for three days. It's the best week your site has ever had.
What does your analytics tool do?
Depending on the vendor, one of three things: it locks your dashboard until you upgrade, it bills you automatically for the overage, or it sends you an email and keeps working. Those three responses reflect three genuinely different pricing philosophies, and it's worth choosing your tool with eyes open.
Key idea: The moment you exceed your plan is the moment you can least afford your analytics disappearing — it's when you most need to know where the traffic came from and what it did.
Option 1: The lockout
Some tools restrict dashboard access when you exceed your tier — sometimes immediately, sometimes after consecutive over-limit months. The logic is straightforward: limits mean limits.
The problem is timing. A lockout hides your data during the single most interesting traffic event of your year. You can't see which referrer drove the spike, which pages held attention, or whether any of it converted — until you've completed a billing flow you didn't plan for, priced at the exact moment your leverage is lowest.
Option 2: Metered overages
Other tools charge per extra thousand events, automatically. It's arguably fairer than a lockout — you get your data and pay for what you used.
But metered billing changes your relationship with success. A traffic spike becomes a number you check nervously. You start estimating what virality costs. Some teams respond by sampling their traffic or removing the tracker from high-traffic pages — degrading their own data to manage a bill. When a pricing model incentivises customers to collect less data in a data product, something is off.
Metered billing exists because it maps cleanly to infrastructure cost. But analytics events are tiny — a pageview is a small row in a database. The marginal cost of a viral week is real but modest, and it's the vendor who is best positioned to absorb variance across customers, not each individual customer.
Option 3: Fair use — soft limits, no surprises
The third model: plans include a generous allowance, and exceeding it triggers communication instead of enforcement.
This is how Antlytics works:
- Starter includes 500,000 pageviews a month; Pro includes 2 million — measured across all your sites, reset each calendar month.
- A usage meter in the dashboard shows where you stand.
- At ~80% of your allowance you get an email heads-up. At 100%, another one, plus an upgrade suggestion in the dashboard.
- And that's it. No lockout, no metered charges, no dropped events. Your data keeps flowing and your dashboard keeps working.
If your traffic has permanently outgrown your plan, the ask to move up a tier comes by email, with notice, like a conversation between adults. Sustained extreme overuse — the agency routing fifty clients through one Starter account — is a different thing from a viral month, and gets handled as such.
Why this works economically
Fair use isn't charity; it's a bet that most overages are temporary. The site that spikes to 700K pageviews in launch month usually settles back to 80K. Billing them for the spike (or locking them out during it) extracts a few dollars once and poisons the relationship. Letting the spike through costs cents in storage and earns the kind of trust that makes people stay for years — and upgrade when their baseline grows, which is the honest signal that they need a bigger plan.
It also keeps the published promise simple. "No metered surprises" only means something if there's genuinely no meter attached to your card.
Questions to ask any analytics vendor
- What exactly happens at 100% of my plan — to my dashboard, and to my data?
- Is anything charged automatically without my action?
- Are events past the limit still ingested, or silently dropped?
- How am I warned before I hit the limit?
You want written answers, because the difference only shows up in your worst — that is, best — week.
FAQ
What is fair-use analytics pricing? Generous included allowances, email alerts as you approach them, and no automatic overage charges or lockouts.
What happens if I exceed my pageview limit on Antlytics? Nothing breaks — data keeps flowing, the dashboard keeps working, and you get email alerts at ~80% and 100%.
Why not just bill per pageview over the limit? Because it punishes success and incentivises customers to degrade their own data collection.
The details are in the docs: Usage & fair use · Pricing