Honestly, I've seen this pattern a lot and it's rarely just simple bot traffic. People keep assuming that when CTRs look healthy but conversions fall off a cliff, it's bots - but that's usually a symptom, not the root cause.
the US traffic showing up while targeting India with Hindi creatives is the biggest red flag to me. That is absolutely not normal. i'd want to know whether that US traffic is registering inside GA4, Shopify, or Meta itself. Is it hitting add-to-carts and checkouts, or is it just session-level noise? That distinction matters hugely.
What placements are you running? If you're using Audience Network or broad automatic placements, I'd turn off Audience Network immediately and see what changes. I've seen that alone clean up a lot of garbage.
Session recordings are another tell. Are people scrolling, engaging, spending real time? Or is it the usual crap - super short sessions, erratic mouse movements, no scrolling? that's a dead giveaway for low-quality traffic.
Also, check your tracking setup. I know everyone thinks their pixel is "working perfectly," but CAPI mismatches, broken deduplication, malformed purchase events, checkout domain issues, or attribution drift can completely throw Meta's optimisation off. You end up with cheap clicks, decent CPCs, and solid CTRs while conversion quality disintegrates.
A lot of what people call "bot traffic" is actually Meta drifting toward low-intent users. Sometimes it's real bots or proxies. Sometimes it's VPN users, click farms, low-quality placements, or incentivised clicks. But more often than not, it's technically real humans with zero purchase intent. Meta still counts those as valid traffic even if they never buy.
That's also why click fraud software only helps in specific cases. Those tools are decent at detecting repeated IPs, datacenter traffic, VPNs, and obvious bot patterns. They're much weaker against real people who simply have no intention of purchasing. If Meta's optimisation loses strong purchase signal quality, it starts finding cheap clickers instead of buyers. Fraud software can't fix that because the traffic is technically real.
I'd also dig into the Shopify funnel and payment logs. Hidden checkout or payment friction can look exactly like "bad traffic." Everything upstream might look healthy while payment issues quietly kill conversions.
One last thing: did your CPMs suddenly get cheaper before this started? And did the drop happen gradually or almost overnight? That usually tells you whether it's optimisation drift, tracking issues, or traffic quality problems. Absolutely infuriating when it happens, but there's usually a clear thread to pull once you start mapping the site architecture and funnel.
