The website served a UK-focused business. To reduce irrelevant submissions, the client requested that enquiries originating outside the UK to be treated as spam.
In August 2026, an enquiry arrived from an Australian internet connection and triggered the country rule. The filter had done what it was configured to do, but the visitor’s location did not tell the complete story.
Rather than permanently discarding the submission, BoselaWeb retained it for manual review.
The enquiry contained several signs of genuine commercial intent:
No individual signal proved that the enquiry was genuine. Considered together, however, they provided enough context to recover and deliver it to the client.
We compared reviewed lead outcomes across two equal 54-day periods before and after the country filter was introduced.
Between 1 May and 24 June 2026, the website received 291 lead submissions: 280 confirmed spam submissions and 11 valid enquiries. Suspicious activity accounted for 96.79% of confirmed spam, representing 271 of 280 spam submissions. None of the valid enquiries appeared in the suspicious queue.
Between 1 July and 24 August 2026, the website received 235 lead submissions: 227 confirmed spam submissions and 8 valid enquiries. Suspicious activity accounted for 99.56% of confirmed spam, representing 226 of 227 spam submissions. Spam outside the suspicious classification fell from 3.21% to 0.44%.
The post-change suspicious queue also contained one genuine enquiry: the Australian submission described here. Manual review recovered it and ensured that it reached the client.
Google Search accounted for 6 of 11 valid enquiries during the earlier period. During the post-change period, it accounted for 6 of 8 valid enquiries when the recovered Australian submission is included.
Although both periods covered the same number of days, their traffic sources differed. Google Ads generated 6 of the 8 human-classified spam leads during the earlier period but did not appear in the post-change traffic-source data. The results therefore describe the observed outcomes and do not establish that country filtering alone caused every difference.
Country information can indicate where an internet connection appears to originate. It cannot reliably determine where a service is required or whether the visitor has genuine commercial intent.
A decision-maker may be travelling. A parent company may be located overseas. A business may be preparing to enter the UK market. Corporate networks, mobile providers and privacy services can also make a visitor’s apparent location different from their actual business context.
Country filtering can therefore be useful for identifying submissions that need closer attention. It becomes risky when it permanently rejects enquiries without considering any other evidence.
BoselaWeb uses a layered process with three possible outcomes:
This approach helps reduce inbox spam while preserving the opportunity to recover genuine enquiries that do not fit a simple automated rule.
The country filter correctly identified that the submission originated outside the client’s expected market. Manual review then identified the UK commercial requirement, recovered the enquiry and delivered it to the client.
The lesson is not that country filtering is ineffective. It is that country filtering works best as one part of a managed lead-protection process.
For businesses that depend on website enquiries, reducing spam is only half the job. The other half is ensuring that legitimate customers are not filtered out with it.
These figures come from one BoselaWeb-managed website and show what occurred during two 54-day periods. Traffic sources differed between the periods, so the results should not be interpreted as proof that country filtering caused every change or as a result representative of every BoselaWeb website.