Searching for information about a robot product returns mostly material written by or for the seller. It is not necessarily inaccurate; it is simply one-sided. Finding sources that discuss the weaknesses takes deliberate effort and it is worth making.
Recognising promotional material
Signals that a piece is selling rather than assessing.
Only advantages. No product is without drawbacks. A piece naming none is an introduction rather than an evaluation.
Drawbacks that are advantages in disguise. The stated weakness being that it has so many features they take time to learn.
Published at launch. The writer has not used it long enough to know anything.
Affiliate links or discount codes. Not disqualifying, and worth knowing the writer has an interest.
Images matching the manufacturer's. Real users do not photograph things that way.
No mention of running costs. Actual users discuss consumables, batteries and long-term expense constantly.
No specific failure situation. Genuine assessments describe a case where the product struggled.
Where useful information is
User communities. Groups where owners ask each other for help when something goes wrong. The best source, because nobody there has an incentive to praise.
Reviews written after six months. Search with a time qualifier. Writers at six months discuss what first-week reviews cannot know.
Uncut video. Continuous footage rather than edited sequences shows considerably more, including hesitations and retries.
Someone you know who bought one. The best source if available. Ask directly whether they would buy it again.
Repair businesses. People who service equipment know which models fail and how. Rarely consulted and highly informative.
Reference sites. Ask the supplier for a working installation and visit it. Ask the site owner rather than the supplier what they would change.
Questions that produce useful answers
For owners in a community or at a reference site.
- How long have you had it and what has failed?
- What does it cost you monthly in consumables?
- How much preparation does it need before each run?
- What is your space like? To judge whether their conditions resemble yours.
- Would you buy the same model again? The best single question, and the answers are usually blunt.
- What did you not know when you bought it? Consistently produces the most valuable information.
What not to ask. Which model is best. That question returns opinions. Questions about someone's specific experience return facts.
The last two questions in the list are worth more than everything else combined, because they surface the things nobody thinks to mention until asked directly.
Common sales techniques to recognise
Not necessarily dishonest, and worth identifying.
An inflated original price with a large discount. The original figure is frequently one nobody ever paid.
Limited availability or a deadline. Creating pressure to decide quickly. Good products rarely need this.
Comparison against an unnamed alternative. Twice as capable as conventional models — which models are unspecified.
Impressive-sounding certification. Worth checking who issues it and what it covers.
Content formatted as news. Paid placement presented as reporting.
Endorsement by a well-known figure. Says nothing about fit for your situation.
Features promised in a future update. Evaluate what it does now. Roadmaps are not commitments.
Volume claims without a denominator. Thousands of units deployed worldwide tells you nothing about how many are running in conditions like yours, or how many are still running. Ask for the local figure and for a site you can visit.
An unusually long warranty. Read the exclusions, which are usually where the length is paid for.
A research sequence before buying
Assembled into a workable order.
One: define what you need. Three specific things, written down. Do this before reading anything, because reading first leads toward features that sound appealing rather than toward requirements.
Two: identify three to five candidate products. Within budget.
Three: find the user community for each. Read the problem threads, not the showcase posts.
Four: ask the questions above. In the community or at reference sites.
Five: verify spares, support and warranty locally. Before shortlisting further.
Six: see it operating live, or arrange a trial.
Seven: decide, and keep the documentation.
The sequence takes a few days spread over a couple of weeks, and it prevents the category of mistake that costs far more than the time invested.
Frequently asked questions
How do you recognise promotional content?
It lists only advantages, or drawbacks that are advantages in disguise, was published at launch before anyone could know much, and says nothing about running costs — which actual owners discuss constantly.
What is the best information source?
User communities where owners ask each other for help when something goes wrong, because nobody there has an incentive to praise. Followed by reviews written six months after purchase and by repair businesses.
Which questions produce the most useful answers?
Whether they would buy the same model again, and what they did not know when they bought it. Both surface information nobody thinks to mention until asked directly.
Why define requirements before researching?
Because reading first leads toward features that sound appealing rather than toward what you actually need, and those features inflate the price while contributing nothing to the job.
More in Humanoids at work and Myths and questions.
