What Are Keyboard Typosquats?
Published 2026-11-14 · Reviewed 2026-11-14 · By the Ingenious AI team
The short answer
A keyboard typosquat substitutes one letter of your brand with a neighbouring key — the fat-finger error. TakedownCoop's generator uses a keyboard-neighbour map, taking the first neighbour for each letter position to keep the candidate set bounded, then checks every variant against DNS and Certificate Transparency like any other candidate.
What is keyboard typosquats?
Keyboard typosquatting exploits the physical layout of typing: each key sits beside others, and hurried fingers land on them. A domain built from a neighbour substitution looks near-identical to the brand in a results list, and on a phone screen the difference can be a single pixel row.
What is keyboard typosquats?
Keyboard typosquatting exploits the physical layout of typing: each key sits beside others, and hurried fingers land on them. A domain built from a neighbour substitution looks near-identical to the brand in a results list, and on a phone screen the difference can be a single pixel row.
TakedownCoop's free brand scan is where keyboard typosquats stops being a suspicion and becomes a record. You enter one domain — no account, no card — and the scanner generates dozens of lookalike candidates from your brand name: deleted-letter, transposed-letter, duplicated-letter and keyboard-neighbour typosquats, homoglyph swaps, affix squats built from prefixes and suffixes, and the same name on different top-level domains. Every candidate is then checked against real DNS resolution, so a name that merely exists on paper is never reported as a live threat.
How the scan ranks and records it
Certificate Transparency is the scan's second witness. When a lookalike holds a publicly logged certificate, the scan records it, and names that appear in the logs but were never generated as candidates are added as CT discoveries. Hits are then prioritised the way a small team needs them: a live, resolving lookalike with a certificate scores 90, a live one without a certificate 80, a registered but non-resolving name 50 with a certificate or 40 without, and a certificate-only discovery 35 or 30. Live threats are filed first; dormant registrations are watched, not panicked over.
Bounded generation: why the scanner takes the first neighbour only
Every letter has several keyboard neighbours, so an exhaustive neighbour set would explode the candidate list and bury real hits in noise. The generator makes a deliberate trade-off, visible in its own code comment: duplication plus keyboard substitution uses the first neighbour only, which keeps the set bounded. The most probable fat-finger variant of each letter is always checked; exhaustive coverage is traded for a scan that finishes and a hit list a human can act on.
That trade-off is worth understanding because it defines what the scan promises. It does not claim every conceivable misspelling — it claims a deterministic, repeatable candidate set across seven families, checked against real DNS and CT evidence, with anything it could not check stated honestly. For takedown purposes that is the stronger position: the packet describes exactly how the lookalike was found, and the method can be re-run to prove it.
A worked example, end to end
A fictional accountancy app, Ledgerlane at ledgerlane.com, scans and finds ledgerlsne.com — the a replaced by its neighbour s — resolving to a credential-harvesting login clone with a certificate: priority 90. The firm files registrar and host packets the same day. In the case notes it records the kind, keyboard substitution, because a portfolio of neighbour-swap registrations is part of the intent picture if the registrar asks why this is abuse rather than coincidence.
Turning evidence into a filing
Finding the problem is only the first third of the work. On Pro ($49/mo) and Agency ($99/mo), each confirmed hit can generate a takedown packet: a complete, channel-correct evidence packet with a filing checklist written for the specific channel — registrar abuse desk, hosting provider, Meta, Google Ads, marketplaces or app stores. The packet carries the genuine brand and domain, the lookalike, how it was detected, whether it resolved at scan time, the DNS and Certificate Transparency evidence, and the priority score, so the person receiving it does not have to take your word for anything.
- Run the scan so the first-neighbour variant of every letter is checked
- File on resolving keyboard variants at the same priority as any live hit
- Screenshot credential or payment forms the clone presents as priority evidence
- Watch dormant neighbour variants under monitoring
- Record the candidate kind in the packet's case notes
After you file: tracking the case
After filing, the case — not your inbox — is the system of record. TakedownCoop's case tracking follows each target through evidence collected, packet built, submitted, provider response recorded, resolved, and, if the impersonator returns, reappeared and re-opened. Two of those states are labelled as holder attestations, because the product records what you report about a submission or a provider's reply; a resolution only counts as verified when a re-scan confirms the item is actually gone. A submitted case is scheduled for recheck seven days after submission.
Where efforts like this usually go wrong
A second failure is overreach — padding a filing with suspicions the evidence does not carry. TakedownCoop's scan is careful to separate what it confirmed from what it could not check, and a filing should inherit that discipline exactly: claim the resolving domain, the certificate, the screenshots you actually hold, and stop there. Reviewers act fastest on filings whose every sentence they can verify, and a single inflated claim taxes all the true ones beside it.
Finally, do not let a win delete the record. Brands that close a takedown by celebrating and clearing the folder meet the reappearance defenceless; brands that keep the case — evidence, tickets, verified resolution and all — meet it with precedent. In this product the record is not paperwork after the work. Kept properly, it is most of the work, because the second filing is always the one that ends the pattern.
Keeping it from coming back
The last step is keeping the street watched. Re-appearance monitoring on Pro and Agency re-scans your domain and regenerates packets when an impostor returns, instead of waiting for a customer to find the fake first. The public verified-genuine page gives customers a door they can check — one page that confirms they are at your real business — and the evidence timeline exports as JSON or CSV for payment processors and insurers. Dealt with once, keyboard typosquats is an incident; watched continuously, it becomes a routine.
Frequently asked questions
Why only the first keyboard neighbour per letter?
To keep the candidate set bounded, as the generator's own comment states. The most likely substitution per position is always covered.
Are keyboard typosquats mostly a mobile problem?
They occur on any keyboard, but small touch targets make neighbour errors more frequent on phones — and phone screens make the wrong domain hardest to spot.
How is a keyboard variant scored?
Like every hit: 90 resolving with a certificate, 80 resolving without, 50/40 registered-only, 35/30 for CT-only discoveries.
Could the first-neighbour rule miss a squatted variant?
The generated set is one source; Certificate Transparency is the second. Names in the CT logs that generation missed are added as CT discoveries.
Keep exploring
TakedownCoop home — run the free brand scan · All TakedownCoop guides · Pricing — Free, Pro and Agency · Get a license key · Takedown cases · My street (dashboard) · Terms · Privacy · Ingenious AI — all platforms