I have bought seven smart-home devices in the last six years. The list: a smart speaker (Google Home), a smart bulb set (Philips Hue), a smart plug set (TP-Link), a smart thermostat (Tado), a robot vacuum (Roborock), a video doorbell (Ring), and a smart lock (Nuki). Total spent: about €1,800. Total daily-use devices: one. The smart speaker is the survivor.
The smart speaker is the device that earns its space. The other six are devices I use occasionally, devices I use rarely, or devices I have turned off entirely. The pattern across the six is interesting. The pattern explains why the smart speaker stayed.
This is the list, what each device taught me, what stayed and what did not, and the framework I now use to decide which smart-home devices are worth buying.
The survivor: the smart speaker
The smart speaker (Google Nest Audio, €100) is the device I use every day. The uses are limited but consistent: music, podcasts, timers, weather, and the occasional question.
Music. The most common use. The speaker is connected to Spotify. I say "Hey Google, play [artist]" and the music starts. The home is filled with music. The music stops when I say "Hey Google stop" or when I leave the room. The system is hands-free. The system is convenient. The system is used almost every day.
Podcasts. The second most common use. The speaker plays podcasts through Spotify. The same hands-free pattern. The podcast can be paused, resumed, and skipped with voice commands. The podcast is convenient.
Timers. The third most common use. The kitchen timer is hands-free. I say "Hey Google, set a timer for 12 minutes" and the timer counts down. The timer is especially useful when cooking. The timer is more reliable than the microwave timer.
Weather. The fourth most common use. The morning weather check is hands-free. "Hey Google, what's the weather today?" The weather is announced. The weather check is a small daily ritual.
Questions. The fifth use, less common but occasional. The speaker answers general questions. The answers are usually accurate. The answers are usually sufficient. The answers are usually hands-free.
Total uses per day: about 15-20. The uses are spread across cooking, working, and relaxing. The speaker has been used every day for four years.
The cost-per-use over four years: €100 / (4 years × 365 days × 15 uses) = €0.004 per use. The cost-per-use is essentially zero.
The speaker earned its space because it is genuinely useful, every day, for tasks that would otherwise require reaching for a phone, a tablet, or a cookbook. The speaker is the rare smart-home device that delivers on the promise.
What did not stay
The six devices that did not earn their space, in order of how often I use them now.
The smart bulbs (Philips Hue, €200 for a starter set). Used for: the bedroom and the living room. The bulbs can change color, dim, and turn on/off via voice or app. The bulbs work. The bulbs are reliable. The bulbs are not used every day.
The reason: the bulbs require thinking. To use the bulbs, I have to remember the voice command or open the app. The bulbs do not provide enough convenience to overcome the friction. The wall switch is faster. The wall switch works. The wall switch is what I use.
The bulbs have been used approximately twice a month over four years. The cost-per-use is high. The bulbs are now sitting in "always on warm white" mode, which is what they would have been as regular bulbs.
The lesson: smart bulbs are useful only when the use case is genuinely hands-free or genuinely dynamic (movie mode, party mode). The bulbs are not useful for basic on/off. The basic bulbs are an expensive solution to a non-problem.
The smart plugs (TP-Link, €80 for four). Used for: the living room lamps and the kitchen appliances. The plugs can be controlled via app or voice. The plugs work. The plugs are not used every day.
The reason: the same as the bulbs. The plugs require thinking. To use the plugs, I have to remember the voice command or open the app. The plugs are not useful for the basic use case. The plugs were useful exactly once — for the Christmas tree, which I wanted on a timer.
The plugs have been used approximately ten times in four years. The cost-per-use is high. The plugs are now sitting in "always on" mode, which is what they would have been as regular plugs.
The lesson: smart plugs are useful only when the use case is genuinely dynamic (a lamp that needs to be on a timer, a coffee maker that should start before I wake up). The plugs are not useful for the basic on/off.
The smart thermostat (Tado, €250). Used for: the central heating. The thermostat can be controlled via app or voice. The thermostat can learn schedules. The thermostat can detect when the house is empty. The thermostat works. The thermostat is not used every day.
The reason: the thermostat works automatically. The thermostat has been set up with a schedule. The thermostat does its job without intervention. The thermostat does not need me to interact with it. The thermostat is invisible to me. I have not opened the Tado app in about two years.
The thermostat is a "set and forget" device. The set-and-forget is the value proposition. The set-and-forget is what makes the thermostat worth the money. The set-and-forget is also what makes the thermostat invisible. The thermostat does not need attention to provide value.
The lesson: smart thermostats are valuable because they automate a previously-manual task (adjusting the temperature). The value is in the automation, not in the interactivity. The interactivity is a bonus.
The robot vacuum (Roborock, €400). Used for: vacuuming the apartment. The robot can be scheduled, can be controlled via app, can map the house. The robot works. The robot is used twice a week.
The reason: the robot is useful but limited. The robot cannot vacuum under all furniture. The robot gets stuck on cables. The robot needs the floor cleared before it runs. The robot takes longer than a manual vacuum. The robot is useful when I do not want to vacuum. The robot is less useful when I want the apartment vacuumed fast.
The robot is used twice a week, on a schedule. The robot is genuinely useful. The robot is not the daily-use device the smart speaker is.
The lesson: robot vacuums are useful for people who do not want to vacuum manually. The robot is useful for medium-traffic apartments. The robot is less useful for high-traffic areas or for people who want fast cleaning. The robot is worth it for the time savings, not for the cleaning quality.
The video doorbell (Ring, €200). Used for: the front door. The doorbell can show video, can be controlled via app, can detect motion. The doorbell works. The doorbell is used occasionally.
The reason: the doorbell requires the Ring app to be useful. The app has notifications. The notifications are mostly false positives (a car passing, a leaf blowing). The notifications train me to ignore them. The actual useful notifications (someone at the door) are rare.
The doorbell has been used approximately once a week for four years. The cost-per-use is high. The doorbell is now used as a regular doorbell, with the smart features turned off.
The lesson: video doorbells are useful only when the user wants to see who is at the door without opening it. The doorbells are less useful for people who are usually home, or for people who do not get many visitors. The doorbell is a niche product.
The smart lock (Nuki, €200). Used for: the front door. The lock can be controlled via app, can detect who is home, can auto-unlock. The lock works. The lock is used occasionally.
The reason: the lock requires the Nuki app or a phone to unlock. The lock is slower than a key. The lock has failed twice (battery died) and required a key. The lock is useful when my hands are full. The lock is less useful when my hands are empty and a key is faster.
The lock has been used approximately three times a week for two years. The cost-per-use is acceptable. The lock is worth it for the convenience of not carrying keys.
The lesson: smart locks are useful when the user does not want to carry keys. The locks are less useful when a key is faster. The locks are useful for families where multiple people need access. The locks are niche.
The framework
The framework I now use to decide whether a smart-home device is worth buying.
The daily-use test. Will I use this device every day? If yes, it is probably worth it. If no, the cost-per-use will be high. The smart speaker passed the daily-use test. The other six did not.
The friction test. Does the device remove friction from a task I do often? Or does the device add friction? The smart speaker removed friction (no reaching for the phone, no unlocking, just voice). The smart bulbs added friction (voice command vs wall switch). The friction test is the most reliable predictor of whether a device will be used.
The "set and forget" test. Can the device be set up once and then operate automatically? If yes, the device is worth it even without daily interaction. The smart thermostat passed this test. The smart bulbs did not.
The privacy test. Does the device require sending my data to a third party? Does the device have a camera or microphone? Does the trust in the manufacturer justify the data being sent? The smart speaker (microphone, but limited to "Hey Google" wake word) was acceptable. The video doorbell (video of my front door) was acceptable because of the door location. The smart bulbs (no data) were ideal from a privacy perspective.
The cost-per-use test. What is the cost of the device divided by the expected number of uses? The smart speaker: €100 / (15 uses × 365 days × 4 years) = €0.004 per use. The smart bulbs: €200 / (4 uses × 12 months × 4 years) = €1 per use. The cost-per-use is a useful sanity check.
What I would buy again
If I were starting from scratch, the devices I would buy, in order.
Step one: a smart speaker. About €100. The speaker is the daily-use device. The speaker delivers on the promise. The speaker is worth the money.
Step two: a smart thermostat. About €250. The thermostat is the set-and-forget device. The thermostat saves money on bills. The thermostat is worth the money.
Step three: a robot vacuum. About €400. The vacuum is the time-saver. The vacuum is useful twice a week. The vacuum is worth the money for the time saved.
After those three, the collection is complete. The other devices (smart bulbs, smart plugs, video doorbell, smart lock) are not worth the money for most people. The other devices are niche products.
What I would skip
The devices I would not buy again.
Smart bulbs. The bulbs add friction. The bulbs do not remove friction. The bulbs are not worth the money for most people. The regular market is faster.
Smart plugs. The plugs are useful only for specific cases (a Christmas tree on a timer, a coffee maker that starts before I wake up). The plugs are not worth the money as a general category.
Video doorbell. The doorbells are useful only for people who are often away from home, who get many visitors, or who have specific security concerns. The doorbells are not worth the money as a general category.
Smart lock. The locks are useful only for people who do not want to carry keys, or for families with multiple users. The locks are not worth the money for people who do not have those specific needs.
The honest summary
The smart speaker is the only smart-home device that has earned its space in my home. The other six devices were useful in the moment of purchase but did not survive the daily-use test. The pattern across the failures is consistent: the devices added friction, did not remove friction, or were set-and-forget but did not need to be smart to be set-and-forget.
The framework that has worked: the daily-use test, the friction test, the set-and-forget test, the privacy test, the cost-per-use test. The framework is what I use to decide whether a new device is worth buying. The framework is what most smart-home manufacturers hope I do not use.
The smart-home category is full of devices that look useful in marketing and become expensive dust collectors in real life. The smart-home category has a few devices that genuinely improve daily life. The smart speaker is in the second category. Most of the rest are in the first.