How do you spot fake engagement on TikTok?

The single most useful signal is whether the engagement rate fits the account's follower tier. Smaller accounts naturally engage harder, larger ones regress toward the platform mean. Using HypeAuditor / Influencer Marketing Hub 2025 ranges as a guide: nano creators under 10k followers run around 8% or higher, 10k-100k around 5%-8%, accounts above 100k around 4%-6%, and the platform median sits near 2.6%.

Now apply the test in both directions. An account with 500k followers showing 15% engagement is almost certainly inflated by purchased likes or pods, because that figure is wildly above the 4%-6% you'd expect at that tier. The same 500k account showing 0.3% engagement is the opposite red flag: the follower count is likely padded with bots that never interact. Either extreme means the follower number is lying about the real audience.

Real engagement keeps consistent relationships between metrics. Compare like-to-view and comment-to-like ratios across several videos, not just the best one. A pattern of lots of likes but almost no comments is suspicious, because genuine fans who like a video also reply, tag friends, and ask questions. Comments are harder to fake than likes, so they are the more honest signal.

Then read the comments themselves. A wall of generic one-word replies, repeated emojis, or vague praise like "nice" and "love it" with no reference to the actual content is a classic bot or engagement-farm fingerprint. Authentic comment sections reference specifics from the video, disagree, ask follow-ups, and vary in length and tone.

Sudden follower spikes with no viral video to explain them are a strong tell. Healthy growth tracks breakout content: a post that overperformed, a trend the creator caught, a collaboration. If followers jumped tens of thousands in a few days while view counts stayed flat and no video went viral, the new followers were probably bought rather than earned.

Geography matters just as much. If you are running a US brand deal but 70% of the creator's audience sits in unrelated low-cost regions often used by follower farms, that audience will not convert for you no matter how clean the engagement looks. Match the audience location to where your product actually sells, and treat a large mismatch as a disqualifier rather than a discount.

View counts disconnected from follower count cut both ways. A 500k-follower account whose videos consistently pull only a few thousand views is showing that the follower base is dead or fake, because TikTok's feed would surface content to a real audience that size. Conversely, modest follower counts with consistently high views can be perfectly healthy, since TikTok distributes by interest, not just by who follows.

The practical move is to compare against real video performance, not a profile's claimed average. KOLens computes engagement rate, average and median views from the videos actually returned by a live search, so you are measuring how the creator performs in practice rather than trusting a self-reported headline number that bought engagement can inflate.

No single signal is conclusive on its own; fake engagement shows up when several break at once. Before committing budget, run the handle through a quick authenticity pass. KOLens offers a free TikTok authenticity audit at /tiktok-audit that surfaces an authenticity / fake-follower signal and an audience snapshot for any handle, and a free engagement rate calculator at /tiktok-engagement-rate-calculator to sanity-check a creator against the tier benchmarks above.

Ultimately the question is not how many followers a creator has but whether their audience is real, fits your brand, and produces trackable results. Use authenticity signals to filter the obvious frauds out, then judge the rest on conversions, sales and retention rather than on a vanity headline number.