Why the Cheap Ones Are Cheap

The five cuts, at a glance

Each of those deserves proper attention, because each fails in a different way.

1. The camera: 8–10 MP versus 36–48 MP

This is the one that sounds like a marketing number and is not. These machines work by identifying and measuring features — counting pores, measuring the area of a pigment patch, counting hairs per follicle mouth. Everything downstream is built on what the sensor can actually resolve.

• Fewer pixels over the same face means less detail per feature. A fine line, a small pore or an individual follicle occupies a certain fraction of the frame; below a certain resolution it stops being distinguishable from its neighbours.
• The indicator still produces a number. That is the trap. The report looks the same — a score, a grade, a percentage. There is simply less information behind it.
• It hits tracking hardest. If a real improvement is smaller than what the sensor can resolve, your follow-up scan shows no change. The client concludes the treatment did not work. It may have; the camera could not see it.
• Magnification multiplies the problem. 200× micro-inspection on a low-resolution sensor is a large blurry image. Magnification without resolution shows you nothing new — and this is exactly where scalp analysis lives.

What to ask: the sensor's megapixel count and whether it is industrial grade. If the answer to the second is vague, treat the first as the whole answer.

2. Report and analysis depth

The report is the product. A client never sees the sensor; they read the report, and you hand it over several times a day. This is where a cheaper machine is most visibly thinner — but only once you try to use it properly.

What a full system produces, concretely:

• 8 light spectra rather than two or three — each one revealing a different layer. What each one shows.
• 17 to 27+ indicators, with score, grade, count, area and percentage coverage on each.
• Heatmaps — sensitivity, pigment, red pigment — which show where, not just how much. Usually the most persuasive artefact in the whole report.
• A–E grading, colour-coded, so a client reads a level instead of interpreting a number.
• Three report modes, including a fully editable one where you annotate and rewrite — so the document is yours rather than generic software output.
• Three or four comparison modes plus contrastive analysis against previous sessions, and a radar chart of everything grading below C.
• H5 mobile report with a QR code, so it leaves the clinic on the client's phone.
• 19 languages on the B3, and customisable watermarks on the X6.

3. No system updates

An analyser is a software product in a metal case. Roughly half of what you paid for is code: the AI models, the indicator algorithms, the report templates, the interface, the language files, and the operating system underneath.

A machine that never receives an update is frozen at the moment you bought it.

• Bugs are permanent. A mis-scoring indicator, a report layout that cuts off text, a translation error in a client-facing document — all yours to live with.
• Improvements never arrive. The algorithms in this category are actively developed. A frozen machine falls progressively further behind one that is maintained.
• Security patching stops too. These run on Android in several cases. An unpatched device holding your clients' facial photographs is a genuine exposure, not a theoretical one.
• It can simply stop working. If the software talks to a cloud service and that service's API changes, an un-updated client eventually breaks. The machine is fine; the software can no longer reach the server.

Ours are covered by backend cloud support that keeps the device connected, updated and operational with remote assistance. What to ask any supplier: how long are updates provided, how are they delivered, what do they cost after the first period, and what happens to the machine if updates stop.

4. Unstable server — and your client data is on it

This is the sharpest of the five and the least discussed, because it only becomes visible at the worst possible moment.

These systems store client images, reports and scan history in the cloud — that is what makes client files, history and cross-visit comparison work. Which means the vendor's server is not a technical detail. It is where your client records live.

• Server down, mid-consultation. The client is seated, you have taken the scan, and the report will not generate. There is no graceful way to handle that in front of a paying customer.
• Server down on a follow-up appointment. The comparison you booked them in for cannot be retrieved. The one appointment type that depends entirely on history is the one that fails.
• Vendor disappears, history goes with it. Your scan archive — months or years of client progress, the single most valuable thing the machine produced — is on infrastructure you do not control and cannot audit.
• You cannot check any of this in a demo. A server is fine until it is not. Uptime is invisible on the day you buy.

What to ask, and get in writing:

• Where is the server, whose is it, and in which country?
• Can the machine capture and produce a report fully offline, with the cloud only for sync?
• Can you export all scans and reports in bulk, in a standard format, at any time?
• What happens to stored data if you stop paying, or if the vendor stops trading?
• Is there any uptime commitment at all, or is it best-effort?
• Can the supplier or manufacturer view your clients' images?

Ours run backend cloud server support with real-time sync, client file management and multi-device management. Bulk export is the question to press on with any supplier, ours included — because an export you have tested is the only real insurance against every scenario above.

5. No backend support

“Support” in this category should mean more than a phone number. On a connected machine it can mean genuine remote capability — and where that exists, it changes the economics of a fault.

What a real backend provides, from our own T3's system:

• Serial-number tracking with remote troubleshooting — a technical issue identified and resolved through the backstage system, without shipping anything.
• A backstage management platform — analytics dashboard with footfall, client age distribution, gender ratio, and problem proportions across five severity levels.
• Multi-device unified management — several units administered from one place, which is the difference between one clinic and a chain.
• Client file management with search and full test history.
• Language and system settings managed centrally.

With no backend, every fault becomes a courier job and a machine standing idle. There is no remote diagnosis, no central view across devices, and no business analytics — and a chain running four units has no way to see them together.

What the price difference actually buys

Put the five together and the pattern is consistent: every cut is invisible on the day of purchase and expensive from month three onward.

Which is why comparing these machines on price alone compares the wrong thing. The honest comparison is: what is this machine worth in month twelve, when you are running follow-up appointments off its history?

The five questions, to send every supplier

Identical questions, sent to everyone including us. The replies are comparable in a way that spec sheets are not.

• What is the sensor's megapixel count, and is it industrial grade?
• How many indicators, how many spectra, how many report modes — and does it support comparison across visits and a mobile report?
• How long are software updates provided, how are they delivered, and what do they cost afterwards?
• Where is the server, can the machine work fully offline, and can I export all my data in bulk?
• Is there a backend management system with remote diagnostics — and can I see it?

A supplier who answers all five in writing is being straight with you regardless of their price. One who cannot answer question four has told you where their savings came from.