The Five Signals We Analyze
Engagement rate, follower authenticity, comment quality, growth pattern, and username entropy. Here is exactly what each one catches and why AI weighs them together.
EngageCheck detects fake followers, bot comments, and engagement pods using five signals: engagement rate, follower authenticity, comment quality, growth pattern, and username entropy. AI plus statistical analysis weighs how those signals interact, since a faked account usually trips several at once. The combined result is an Audience Quality Score from 0 to 100, an evidence-based probability rather than a guarantee.
How the five signals become one score
Collect public data
EngageCheck reads public profile, post, and comment data. No login, no password, and no Instagram connection.
Measure each signal
Every signal is calculated and compared against accounts of similar size so the numbers have real context.
AI weighs the interactions
A model trained on real and fake accounts judges how the signals combine, since fakery usually shows up across several at once.
Score it 0 to 100
The signals roll up into one Audience Quality Score with a clear breakdown of what drove the result.
The five signals in depth
Engagement rate
Likes and comments as a share of followers, benchmarked against accounts of similar size. Real audiences engage in a predictable range. Far below it suggests followers who never see or react to posts, a classic sign of bought followers.
Follower authenticity
The make-up of the follower base itself: dormant accounts, no profile photo, zero posts, or follow-to-follower ratios that point to bots and bulk-purchased followers rather than real people.
Comment quality
Whether comments read like real people. Generic one-word praise, repeated emoji, or identical phrases across posts point to bots or paid engagement instead of a genuine audience.
Growth pattern
The shape of follower growth over time. Organic accounts grow in waves tied to their content. Vertical spikes with no matching engagement bump usually mean followers were bought in a batch.
Username entropy
The randomness of follower usernames. Clusters of handles built from random letters and numbers, or templated patterns, betray automatically generated bot accounts farmed at scale.
Where the signals matter most
Vetting an influencer before payment
Read every signal before money changes hands so an inflated audience does not cost you a campaign.
Spotting engagement pods
Comment quality and growth pattern together expose pods where the same accounts like and comment in lockstep.
Auditing your own account
See which signals are strong and which need work, and check whether past growth services left bot followers behind.
Common questions
What are the five signals EngageCheck uses?
Engagement rate, follower authenticity, comment quality, growth pattern, and username entropy. Each measures a different angle of audience quality, and AI weighs them together because fake activity rarely shows up in just one.
Does EngageCheck actually use AI, or just math?
Both. Statistics measure each signal and benchmark it against accounts of similar size. AI then weighs how the signals interact and recognizes patterns of fake followers, bot comments, pods, and abnormal growth that no single metric captures on its own. It is pattern recognition, not magic.
Which signal best detects bought followers?
Usually two together: a sudden growth spike with no matching engagement, plus low follower authenticity. Bought followers arrive in batches and stay silent, so the growth pattern and authenticity signals tend to fire at the same time.
How do the signals catch engagement pods?
Pods leave a fingerprint across comment quality and growth. The same accounts comment on every post with similar phrasing, and engagement looks high relative to real reach. AI flags that coordinated, repetitive pattern.
Can one signal be wrong?
Yes, any single signal can mislead. A real account can have a quiet week or a viral spike. That is why EngageCheck combines all five into one probability score instead of judging on a single metric, and why it does not claim 100 percent accuracy.
What is username entropy?
It is a measure of how random or templated follower usernames are. Real audiences have varied, human-chosen handles. Bot farms generate names from patterns and random characters, so low-variety clusters point to automated accounts.
Do the signals work on any account?
They work on any public Instagram account. Private accounts cannot be analyzed because their post and follower data is not public, and EngageCheck never needs a login to read what is already public.
Is the audit free?
Your first audit is free. You add a card to unlock it so each account gets one free audit and bots cannot farm the tool. You are not charged for the free audit, and your card goes only to Stripe.
Related tools and guides
See the five signals scored on any account
Run your first audit free and get the full signal-by-signal breakdown in under a minute.