Ask any skylight cleaning robot supplier five questions before you buy: can you show holding-force test data at my roof slope, what is the spare-parts lead time, what is the service response time, which certifications apply, and can I speak to a site running the same machine.
Most buying mistakes trace back to a question nobody asked. Price is visible. The risks are not. These five questions surface the risks while you still have room to negotiate.
1. Can you show holding-force test data at my roof slope?
Not a marketing number, a test report. The holding force of a vacuum robot depends on glass condition, temperature and moisture, so a single figure on a brochure tells you little. Ask for the test method and the derating they apply for wet glass. If the answer is a slope in degrees with no supporting data, keep asking.
2. What is the spare-parts lead time?
Suction cups, brushes, squeegee blades and filters wear out. If a cup takes four weeks to arrive from overseas, your machine is idle for four weeks and your cleaning contract is still due. A serious supplier keeps common wear parts in a regional warehouse and can quote you days, not months.
- Wear parts in stock locally: cups, brushes, blades, filters.
- Quoted lead time in calendar days, in writing.
- A named spares contact, not a generic support inbox.
3. What is the service response time?
Ask for a written response window and whether a loaner machine is available during repair. On a commercial building with a cleaning schedule, a two-day response can be fine and a three-week one is a deal-breaker. Get the SLA in the contract, not in a sales email.
4. Which certifications apply, and can I see them?
CE for the machine, at least IP54 and ideally IP65 for wet use, and a UN38.3 certificate if it runs on lithium batteries. Ask to see the documents, not a claim that they exist. This matters for shipping, insurance and any tender that vets suppliers.
5. Can I speak to a site running the same machine?
The single best test. A reference site tells you real output, real failure rates and how the supplier behaves when something breaks. If no reference is offered, or the reference is suspiciously new, that is information too.
Who should walk away here: anyone being pushed to a bigger machine than their roof needs, and anyone told a single robot will eliminate rope access entirely. Both claims signal a seller optimising for the deal, not the building. Who should buy: teams that got straight answers to all five, plus a clear picture of the hybrid robot-and-crew workflow.
What red flags should end the conversation?
A few answers tell you to walk. “The slope limit is just a guideline” is one. “We do not stock wear parts” is another. Any claim that the robot removes all need for rope access, a claim you should treat as false on any real building. And an evasive answer on who made the battery, because the battery is where cheap machines cut corners first.
Your position is stronger than it feels. Several credible suppliers compete for the same building, and a buyer who asks sharp questions gets better pricing and better terms. Run the five questions on every quote, write the answers down, and compare them side by side before you sign anything.
How should you compare competing quotes?
Put the answers in a table, not in a memory. One column per supplier, one row per question: holding-force data, spares lead time, service response, certifications, reference site. Faced with the same rows, the vague answers stand out. A supplier who wrote “IP65” but cannot produce a test report looks different beside one who attaches it to the quote.
Price goes in the last column, not the first. A machine that costs 20 percent more but ships parts in three days and answers the phone will cost less over five years than the cheaper unit that leaves you waiting a month for a suction cup. Buy the support, and the machine comes with it.
What should be in the written offer?
Insist the answers make it into the written offer, not just the sales call. The slope rating, the holding force at that slope, the spares lead time, the service response window, the certification list and the reference contact should all be in the document you sign. A verbal promise is worth nothing when a cup fails in the fifth month.
This one habit prevents most disputes. Suppliers who are confident put it in writing; suppliers who are not will resist. The resistance itself is the answer to your assessment, and it costs you nothing to ask.
How do you check a reference site properly?
Do not just ask whether the machine works. Ask the reference three specific things: how long a clean actually takes, how often it has gone down, and how fast a part arrived when it did. Those answers expose the honest picture that no brochure carries, and a site that has run the machine through a full year will tell you more than the specification ever could.
Try to see the machine on a roof, not in a meeting room. Watch a real clean on real glass and you will learn the setup time, the noise, the water handling and the operator’s own frustration points within half an hour. That single visit replaces most of the risk in the purchase decision.
Key Takeaways
- Demand holding-force test data at your actual slope, not a brochure figure.
- Wear-part lead time decides whether a small fault becomes a lost month.
- Get a written service SLA with a response window and loaner terms.
- Ask to see CE, IP and UN38.3 documents rather than trusting a claim.
- Talk to a real reference site; a supplier who cannot name one is a risk.
Want honest answers to all five against your building? Reach our team through the contact page, or read the 2026 safety and certification guide.