Monitor and protect modes
Monitor mode scores and reports every paid click, protect mode also keeps invalid ones away from your ads, and sensitivity decides how high a score must be before a click counts as invalid.
The two modes#
| Monitor | Protect | |
|---|---|---|
| Scores every paid click | Yes | Yes |
| Shows reasons, reports and alerts | Yes | Yes |
| Adds IPs to Google Ads exclusion lists | No | Yes |
| Adds flagged visitors to the Microsoft Ads exclusion audience | No | Yes |
| Adds flagged visitors to the Meta Ads exclusion audience | No | Yes, on Growth and up |
Mode is set per site, in Sites, then your site, then the Rules tab. New sites start in protect.
Automatic blocking is a paid feature. On the Free plan nothing is pushed to your ad accounts whichever mode the site is in; blocking in Google and Microsoft starts on Starter, and Meta on Growth.
What protect mode does with an invalid click#
- The click's verdict becomes final: when the visitor leaves, or three minutes after landing.
- Spurn creates an exclusion for the IP address. When two or more data-centre IPs from the same /24 block misbehave, it excludes the whole block instead.
- Spurn pushes the exclusion to every Google Ads account linked to the site: the account-level list first, then campaign lists.
- If Microsoft Ads or Meta Ads is linked, the visitor is queued for that platform's exclusion audience.
An exclusion lasts 30 days by default and is extended each time the network is flagged again. Everything lands in the change log under Blocking, and any of it can be undone. See Google Ads exclusion lists.
Suspicious clicks are reported, never blocked. Only invalid clicks lead to exclusions.
Turning blocking off for one ad account#
Each connected ad account also has its own blocking switch, under Ad accounts. With it off, Spurn still imports that account's clicks and reports on them, but pushes nothing to it. Use it when a client wants to approve changes to their account, or while you check that a newly linked account belongs to the right site.
Sensitivity#
Sensitivity maps the 0 to 100 score to a verdict:
| Sensitivity | Invalid at | Suspicious at |
|---|---|---|
| Relaxed | 75 or more | 45 or more |
| Balanced (default) | 65 or more | 35 or more |
| Strict | 55 or more | 25 or more |
A score below the suspicious cut-off is valid. How the score is built is in How detection works.
What one reason does on its own#
A click with a single reason scores that reason's weight. This table shows the verdict each reason gives on its own, before any purchase or engagement halves the score:
| Reason | Weight | Relaxed | Balanced | Strict |
|---|---|---|---|---|
| Known bot | 95 | Invalid | Invalid | Invalid |
| Automated browser | 90 | Invalid | Invalid | Invalid |
| Tor exit node | 85 | Invalid | Invalid | Invalid |
| Data centre network | 75 | Invalid | Invalid | Invalid |
| Repeat clicks | 65 | Suspicious | Invalid | Invalid |
| Repeat device | 65 | Suspicious | Invalid | Invalid |
| Already blocked | 60 | Suspicious | Suspicious | Invalid |
| Outside target countries | 60 | Suspicious | Suspicious | Invalid |
| Click ID replayed | 55 | Suspicious | Suspicious | Invalid |
| Network burst | 50 | Suspicious | Suspicious | Suspicious |
| VPN or proxy | 45 | Suspicious | Suspicious | Suspicious |
| Inconsistent device | 35 | Valid | Suspicious | Suspicious |
| Page never loaded | 35 | Valid | Suspicious | Suspicious |
| Timezone mismatch | 25 | Valid | Valid | Suspicious |
| No interaction | 25 | Valid | Valid | Suspicious |
Most invalid clicks carry more than one reason, and each extra reason pushes the score up. A VPN with a timezone mismatch scores 59: suspicious at Relaxed and Balanced, invalid at Strict.
A purchase, or at least 30 seconds on the page with 5 or more interactions, halves the score. The first three reasons in the table are hard, so nothing halves them.
Which mode to start in#
Protect from day one suits most stores. Hard reasons and data-centre traffic are invalid at every sensitivity, and they are not shoppers. Add your office and agency addresses to the allowlist first, so your own ad checks are never blocked.
Start in monitor when you want to see verdicts before anything changes in your ad accounts:
- You sell to businesses, and buyers may browse from corporate VPNs or cloud desktops, which come from data-centre networks.
- You are an agency onboarding a client who wants to sign off on blocking.
Clicks are scored the same way in both modes, so monitor shows exactly what protect would have blocked.
When to switch from monitor to protect#
After a few days of traffic, open Clicks and look at the invalid ones:
- Read the reasons on a sample of them. Data centre, automation and repeat clicks are what you want gone.
- Check whether any invalid clicks converted. Spurn records orders against clicks through the Shopify pixel, the WooCommerce plugin or the conversion API.
- If real customers show up as invalid, fix the cause first: allowlist the address, add the country to your target countries, raise the repeat click limits or turn off the VPN check. See Allowlist and rules.
- Then switch the site to protect.
When to change sensitivity#
Start at Balanced. Change it when your conversion data says to:
- Move to Strict when suspicious clicks rarely or never buy, your cost per click is high, or you see a run of repeat clicks from one network. At Strict, Outside target countries, Already blocked and Click ID replayed are each enough on their own to block.
- Move to Relaxed when real customers keep landing in invalid. At Relaxed, only the hard reasons and data-centre traffic are invalid on their own, and repeat clicks need a second reason.
Prefer a specific rule change over a sensitivity change when one kind of traffic is the problem. If a B2B audience trips the VPN check, turning that check off fixes the cause without loosening everything else.
A sensitivity change applies to clicks scored from then on. Verdicts that are already final stay as they are.
What no mode can do#
Spurn blocks a network after it has clicked. The first click from a new IP has already been counted by the ad platform by the time Spurn sees it. Protect mode stops the next click from that network. For clicks you have already paid for, use refund claim packs.