Stop paying for clicks that were never going to convert.
Maddet scores every click on your ad accounts in real time, catches bots and click farms before they drain your budget, and gives you the evidence to dispute the charge.
The blind spot
Your dashboard counts clicks. It does not judge them.
Ad platforms filter the obvious cases and bill you for the rest. What slips through looks exactly like a customer until you ask each click where it came from and how it behaved.
- Verified 1,104
- Invalid 180
- Billed for invalid ₹8,640
Invalid clicks look identical to real ones in your ad dashboard.
A click from a script in a data center and a click from a buyer on their phone show up as the same row, at the same price.
By the time a pattern is obvious, the budget is already spent.
Weekly reviews catch a spike after it has cleared your daily cap. Protection has to happen at the speed the clicks arrive.
Disputing a charge needs evidence you don't have.
Platforms review refund requests on specifics: which clicks, when, from where, and why they are invalid. A hunch is not a case.
How Maddet works
Four steps. One continuous loop.
Connect once. From then on every paid click is scored, sorted and documented without anyone on your team pulling a report.
- 01
Connect
Link Google Ads, Meta Ads and display accounts with read-level access. No code required.
- 02
Score
Each click is scored against velocity, IP reputation, device fingerprint and behavioural pattern.
- 03
Block and document
Suspicious traffic is flagged immediately and a timestamped evidence record is written for each event.
- 04
Dispute
Export a dispute-ready report and submit it to the ad platform for refund review.
What the engine catches
Every kind of fake click leaves a shape.
Detection is pattern and velocity analysis, not guesswork. Each category below has a signature the engine is trained to recognise, and a plain-language reason it writes into the evidence record.
Explore detection categoriesAutomated bots
Scripted browsers and crawlers clicking at a pace and precision no person can hold.
Tell Abnormal click velocity, headless markers, zero pointer movement
Click farms
Real devices operated in bulk, rotating identities to look like separate people.
Tell Dense IP ranges, fingerprint collisions, synchronised timing
Competitor click abuse
Someone repeatedly clicking your ads to drain the daily budget early.
Tell One device, repeat clicks, same keywords, working hours
Low-quality placements
Display and network sites that inflate clicks through layout tricks or bought traffic.
Tell Placement history, accidental-tap rates, instant bounces
Proxy and geo masking
Traffic routed through proxies to appear inside a targeted region.
Tell Proxy exits, timezone and language mismatch
Data center traffic
Clicks originating from cloud and hosting networks rather than homes or phones.
Tell ASN type, uniform device traits, off-hours bursts
Evidence over accusation
A block without a reason is just an opinion.
Every click Maddet flags ships with a record of exactly what was seen. It is the difference between telling a platform you think something is wrong and showing it.
- Event ID A unique reference you can quote in a dispute
- Timestamp UTC to the millisecond, click and detection time
- Campaign and ad group Exactly where the charge was incurred
- Click ID The platform's own identifier, GCLID or FBCLID
- IP and /24 range The address and the subnet it belongs to
- ASN and network type Consumer ISP, mobile carrier, hosting or proxy
- Device fingerprint A hashed identifier, never raw personal data
- Pattern matched The plain-language reason plus the signals that fired
- Score and cost The risk score and what the click was billed at
#A1029
Bot pattern detected
40 clicks per minute from a single IP across one ad group, with no pointer movement before any click.
- Timestamp (UTC)
- 04:54:07.312
- Campaign
- Search - Brand - IN
- Click ID
- gclid ...Kx9Qe
- IP range
- 203.0.113.0/24
- Network
- Hosting ASN
- Fingerprint
- fp_7c1e...a90
Plainly stated
What the AI does. And what it doesn't.
Fraud tools tend to overclaim. We would rather you know precisely where the engine's job ends and yours begins.
Maddet does
- Detects bot traffic, click farms and competitor click abuse through pattern and velocity analysis
- Scores publisher and placement quality across your display and network campaigns
- Builds an evidence trail of timestamps, IPs and click patterns for every flagged event
- Exports dispute-ready reports in the structure platforms review
Maddet does not
- Issue refunds. It flags and documents; the final refund decision rests with the ad platform
- Change your bids, budgets or creatives without you
- Claim certainty it doesn't have. Borderline clicks are marked for review, not blocked
- Request write access to your ad accounts to do its scoring
Built for two kinds of teams
Whoever answers for the spend.
Click scoring as traffic arrives, not an end-of-day batch review.
Signals evaluated on every click across network, device, behaviour and context.
Evidence fields written for each blocked click, so every block ships with a reason.
Platform families connected: Google Ads, Meta Ads and programmatic display.
Exposure estimate
Somewhere in last month's spend, there is a number nobody measured.
Enter what you spend on paid media each month. The slip shows the range typically lost to invalid traffic at that level. The audit replaces the range with your real figure.
- Lower band (7.000000000000001%)
- ₹35,000
- Typical (14.000000000000002%)
- ₹70,000
- Upper band (22%)
- ₹1,10,000
- Typical, over 12 months
- ₹8,40,000
This is an illustrative estimate based on typical invalid-traffic rates, not a live scan of your account. Connect your account for your actual numbers.
Swap the estimate for the real figureGet a free traffic audit