Usage-based insurance prices your policy partly on how and how much you actually drive, rather than only on demographic and vehicle factors. The data comes from a phone app, a plug-in device, or increasingly from the vehicle itself.
What they measure
- Mileage. The strongest single predictor, and the most objective.
- Hard braking and rapid acceleration. Used as proxies for anticipation and following distance.
- Speed, sometimes relative to posted limits.
- Time of day. Late-night driving is rated differently.
- Cornering forces.
- Phone handling, in app-based programs that detect screen interaction while moving.
The first question to ask is whether your rate can go up. Some programs only ever discount; others adjust in both directions based on what the data shows. This is disclosed, but it is not always prominent, and it is the difference between a program with no downside and one with a real one.
What the data is used for
Beyond pricing, worth reading the specifics on:
- Retention. How long the data is kept.
- Sharing. Whether it goes to third parties, and which.
- Claims. Whether it can be used in evaluating a claim you make.
- Legal process. Whether it can be produced in response to a subpoena. Driving data is detailed and time-stamped, and it exists once it is collected.
Several states regulate telematics data specifically, and some require disclosures about collection and use. Your Department of Insurance can tell you what applies where you live.
Where the discount is largest
Low-mileage drivers benefit most, because mileage is the strongest factor and is objectively measured. People who commute short distances, work from home, or are retired frequently see meaningful reductions.
Drivers with long highway commutes may see less, even with careful driving, simply because exposure is higher.
The phone problem
App-based programs measure the phone rather than the car, which creates known artefacts. Being a passenger, riding a bus, or using a rideshare can register as trips. A phone loose on a seat may record braking events that never happened.
Most programs let you reclassify trips. It requires you to notice and act, usually within a window.
Young drivers
Some insurers offer programs aimed specifically at young drivers, where the potential reduction is proportionally larger because the base rate is higher. Some also provide feedback to a parent account, which families use differently depending on their own preferences about monitoring.
Before enrolling
- Confirm whether the rate can increase, and by how much.
- Ask how long the monitoring period lasts and whether it repeats.
- Read what the data is used for, retained for, and shared with.
- Check whether you can leave the program, and what happens to the rate if you do.
- Ask whether a discount applies simply for enrolling, before any data is scored.
What we are not saying
We are not recommending or discouraging enrolment, and we receive nothing from any insurer. What we are saying is that these programs can reduce cost meaningfully for low-mileage drivers, that some can increase your rate as well as lower it, and that the data they collect is detailed enough that its retention and use is worth reading before you agree.
Where to verify this yourself
- The program terms — whether rates can increase, and the data policy.
- Your state Department of Insurance — telematics regulations where you live.
What is actually measured
Programmes differ, but the recurring inputs are consistent. Knowing which ones your programme uses matters, because several are affected by circumstances rather than by driving.
| Input | What it captures | Can it misread you? |
|---|---|---|
| Hard braking | Deceleration above a threshold | Yes — avoiding someone else's mistake registers identically |
| Rapid acceleration | Acceleration above a threshold | Sometimes — merging onto a fast road |
| Cornering forces | Lateral acceleration | Sometimes — depends on the roads you drive |
| Time of day | Late-night driving, statistically higher risk | Yes — shift work is rated the same as a night out |
| Mileage | How much you drive | No, but a long commute is not a choice for everyone |
| Phone handling | Screen interaction while moving | Yes — a passenger using your phone can register |
| Speed relative to limit | In some programmes | Sometimes — limit data can be wrong |
The hard braking metric is the one worth understanding before enrolling. Braking sharply because someone pulled out in front of you is defensive driving, and it registers exactly like braking sharply because you were not paying attention. Some programmes now attempt to distinguish context; many do not.
The three programme structures
Discount-only
- Your score can reduce the premium
- It cannot increase it
- The downside is a smaller discount, not a surcharge
- Some states require this structure
Two-way rating
- Your score can reduce or increase the premium
- Poor scores cost money
- Read the terms before enrolling
- Not permitted everywhere
There is also a third pattern: pay-per-mile, where a base rate is combined with a per-mile charge. This is a different proposition entirely — it rewards driving less rather than driving in a particular way, and for a genuinely low-mileage household it can produce a larger saving than behaviour scoring ever would.
The questions to ask before enrolling
- Can this programme increase my premium, or only reduce it?
- What is the maximum discount, and what is the maximum surcharge?
- How long is the monitoring period before a rate is set?
- Can I withdraw during the period, and what happens if I do?
- Exactly which metrics are used?
- Does it track location, and is location used in rating?
- Can I see my own data, and how often?
- Is the score portable if I switch insurers? (Usually not.)
The data questions
These programmes generate a detailed record of where a vehicle went and when. That is worth thinking about deliberately rather than accepting by default.
Worth asking the insurer directly
- What data is collected, and how precisely is location recorded?
- How long is it retained after the programme ends?
- Is it shared with third parties, and which?
- Would it be used to investigate a claim I make?
- Can I request deletion, and does any state law give me that right?
- What happens to the data if I cancel the policy?
The fourth question is the one people do not think to ask. Data collected to price a policy exists and can potentially be relevant to a claim — establishing speed, time, or whether a phone was in use. Insurers' practices differ, and the answer is worth having in writing before rather than after.
Several states have adopted rules governing what may be collected and how it may be used in rating. Your Department of Insurance can tell you what applies where you live.
Who tends to do well and who does not
| Profile | Typical outcome |
|---|---|
| Low annual mileage, mostly daytime | Frequently the largest savings, especially on pay-per-mile |
| Retired, short local trips | Generally favourable |
| Remote worker with occasional driving | Generally favourable |
| Night shift worker | Penalised by the time-of-day metric regardless of driving quality |
| Long motorway commute | High mileage, though motorway driving is otherwise low-risk |
| Dense urban driving | More braking and cornering events by necessity |
| Shared vehicle in a household | Someone else's driving affects your score |
The last row deserves a specific mention. Most programmes score the vehicle, not the driver. In a household where a teenager drives the same car, their driving is your score — which cuts both ways and is worth discussing before enrolling.
A sensible approach to trying one
-
Before enrolling
Get a conventional quote first
So you know what you are comparing against. A telematics discount off a higher base rate is not necessarily a saving.
-
Before enrolling
Confirm whether it can surcharge you
In writing. This is the single most important answer.
-
During
Check your data regularly
Most programmes show a running score. If events are being logged that do not match your driving, raise it early rather than at the end.
-
During
Note anything contextual
A week of unusual driving, another driver using the vehicle, a period of heavy roadworks. Some insurers will consider context if raised.
-
At the end
Compare the result against the market
A telematics discount is only a saving if the final premium beats what someone else would charge you outright.
What we are not saying
We are not telling you to enrol or to avoid these programmes, and we do not sell insurance. What we are saying is that the metrics used are not always a fair proxy for how well you drive, that whether a programme can surcharge you is the first thing to establish, that most programmes score the vehicle rather than the person, and that the data questions deserve an answer before you agree rather than after.
Device, app, or built in
| Method | How it works | Considerations |
|---|---|---|
| Plug-in device | A dongle in the diagnostic port | Scores the vehicle regardless of who drives; must be returned |
| Phone app | Uses the phone's sensors | Can log you as a passenger in someone else's car; battery drain; needs permissions kept on |
| Built-in connected services | Data from the manufacturer's own system | Nothing to install; check what the manufacturer shares and with whom |
The phone app row causes the most avoidable problems. Riding as a passenger, or travelling by bus or train, can be recorded as a trip unless the app detects otherwise. Most apps let you reclassify a trip — find that feature at the start rather than after a month of misattributed journeys.
If your score looks wrong
What to do
- Review the individual trip records rather than only the summary score
- Reclassify any trip where you were a passenger, if the app allows it
- Note the date and circumstances of any event you dispute
- Raise it with the insurer during the monitoring period, not after
- Ask whether events can be reviewed and adjusted, and on what basis
- If the programme can surcharge and the data is wrong, ask about withdrawing
Raising it early matters. Once a monitoring period closes and a rate is set, the conversation is much harder than it would have been in week two.
Common misunderstandings
“A telematics discount means I am paying less than anyone else” Not necessarily. A discount is applied to that insurer's base rate. Another insurer's ordinary rate may still be lower. Compare final premiums, not discounts.
“My good score follows me if I switch” Generally not. Scores are proprietary to each programme and do not transfer, so switching means starting a new monitoring period.
“It only tracks how I drive, not where” Many programmes record location. Whether it is used in rating is a separate question from whether it is collected.
“I can turn it off when I want” Interrupting data collection is frequently treated as non-participation, which can forfeit the discount or end the programme.
The alternative worth pricing alongside it
Before enrolling, ask what discounts you already qualify for and are not receiving — a defensive driving course, low annual mileage declared conventionally, multi-policy, paid-in-full, or a vehicle safety feature discount.
Those require no monitoring and no data sharing, and for some households they add up to more than a telematics programme would deliver. It is worth knowing the size of both before choosing.
One question to end on
Ask the insurer what your premium would be with the best possible score, and what it would be with the worst. Those two numbers bound the entire decision, and an insurer that will not state them is telling you something about how the programme is structured.
Get both figures in writing before enrolling, and keep them. At the end of the monitoring period they are what you measure the result against, rather than relying on how the outcome is presented to you.
A programme that delivers less than it promised is worth leaving, and knowing the two boundary figures is what tells you whether that happened.
This is general education, not advice. Insurance law and claim rules vary by state and change over time. Nothing here is legal, financial, or insurance advice for your situation, and reading it does not create any professional relationship. For your specific case, consult a licensed professional in your state or contact your state Department of Insurance.