I’ve been a WHOOP user for some time and used WHOOP 4.0 for approximately a year alongside the same Samsung watch without experiencing this persistent level of discrepancy.
Since moving to WHOOP 5.0, that has changed significantly.
I’ve been raising the issue with WHOOP Support since May and, rather than relying on a few isolated examples, I spent several months recording WHOOP 5.0 against Samsung Health.
Here is the data I’ve now provided to WHOOP:
| Date | Samsung | WHOOP 5.0 | Difference |
|---|---|---|---|
| 4 Jun | 14,439 | 13,397 | 7.2% lower |
| 10 Jun | 14,269 | 11,275 | 21.0% lower |
| 12 Jun | 12,210 | 10,156 | 16.8% lower |
| 18 Jun | 11,366 | 8,449 | 25.7% lower |
| 21 Jun | 10,444 | 8,928 | 14.5% lower |
| 4 Jul | 9,670 | 9,368 | 3.1% lower |
| 9 Jul | 16,547 | 15,997 | 3.3% lower |
| 10 Jul | 11,970 | 12,778 | 6.7% higher |
| 11 Jul | 14,841 | 10,286 | 30.7% lower |
| 12 Jul | 9,700 | 11,705 | 20.7% higher |
| 15 Jul | 11,852 | 11,070 | 6.6% lower |
| 18 Jul | 10,939 | 12,143 | 11.0% higher |
| 19 Jul | 11,201 | 11,993 | 7.1% higher |
| 20 Jul | 14,506 | 15,237 | 5.0% higher |
| 21 Jul | 17,081 | 17,696 | 3.6% higher |
| 22 Jul | 17,257 | 11,729 | 32.0% lower |
| 25 Jul | 4,038 | 1,448 | 64.1% lower |
| 28 Jul | 9,773 | 7,020 | 28.2% lower |
| 29 Jul | 11,613 | 9,213 | 20.7% lower |
| 30 Jul | 7,282 | 7,056 | 3.1% lower |
| 31 Jul | 7,731 | 12,640 | 63.5% higher |
| 2 Aug | 13,126 | 10,862 | 17.2% lower |
| 9 Aug | 11,969 | 11,789 | 1.5% lower |
| 10 Aug | 10,981 | 9,353 | 14.8% lower |
| 11 Aug | 11,581 | 9,162 | 20.9% lower |
| 12 Aug | 4,811 | 4,334 | 9.9% lower |
| 13 Aug | 12,557 | 13,135 | 4.6% higher |
| 14 Aug | 9,968 | 7,649 | 23.3% lower |
| 20 Aug | 9,551 | 7,276 | 23.8% lower |
| 21 Aug | 12,336 | 10,760 | 12.8% lower |
| 22 Aug | 14,011 | 11,407 | 18.6% lower |
| 23 Aug | 11,124 | 10,111 | 9.1% lower |
| 24 Aug | 13,165 | 10,459 | 20.6% lower |
| 25 Aug | 6,405 | 5,394 | 15.8% lower |
| 26 Aug | 8,495 | 6,743 | 20.6% lower |
| 27 Aug | 12,259 | 12,635 | 3.1% higher |
| 28 Aug | 9,023 | 8,227 | 8.8% lower |
| 29 Aug | 6,392 | 4,766 | 25.4% lower |
| 30 Aug | 9,576 | 7,014 | 26.8% lower |
| 31 Aug | 10,410 | 7,331 | 29.6% lower |
| 1 Sep | 9,245 | 7,063 | 23.6% lower |
| 2 Sep | 11,958 | 9,579 | 19.9% lower |
| 3 Sep | 11,723 | 9,121 | 22.2% lower |
There were also two days, 24 July and 1 August, where WHOOP provided no step record at all despite the device being worn.
And I did controlled walking comparisons after WHOOP Support specifically requested them:
31 May: Samsung 1,201 vs WHOOP 798
1 June: Samsung 3,812 vs WHOOP 2,705
2 June: Samsung 2,473 vs WHOOP 1,903
What concerns me isn’t simply that two trackers disagree
I don’t expect Samsung and WHOOP to produce identical numbers.
What concerns me is the inconsistency.
On some days WHOOP and Samsung are remarkably close: 1.5%, 3.1%, 3.3%.
On other days, using the same devices and wearing them in the same way, WHOOP suddenly records 20%, 25%, 30% or substantially fewer steps.
My most recent five consecutive days averaged:
Samsung: 10,582 steps/day
WHOOP: 8,022 steps/day
That’s an average WHOOP undercount of 2,561 steps per day, or 24.2%.
WHOOP Support keeps returning to how I wear it
This is where I’m struggling with WHOOP’s explanation.
I wore WHOOP 4.0 for approximately a year, on the same wrist, in essentially the same position, against the same Samsung watch.
It performed fine.
Since upgrading to 5.0, I’ve repeatedly been asked to reconsider how I’m wearing the device. I’ve now even supplied photographs of the WHOOP on my wrist from multiple angles.
Support has most recently told me that my account contains periods of “non optimal sensor contact” and “off body” detection.
I’ve therefore asked them a simple question:
Do those events actually correspond to the dates where WHOOP undercounted by 20–30%+, and what do the sensor logs show on the dates where WHOOP and Samsung were within 1–3%?
If sensor contact is causing the discrepancy, there should be a relationship in the data.
And apparently I’m not alone
After looking through this Community, I’ve now found other WHOOP 5.0 users reporting very similar experiences.
One user reports WHOOP 5.0 being 3,000–6,000 steps below Apple Watch after upgrading and says Support also told them to change the position on their wrist.
Another Galaxy Watch 8 user reports an average discrepancy of approximately 5,000 steps, while another Samsung user reports WHOOP consistently being at least 2,000 steps lower.
There is also an entire Community discussion titled “REGRETTING UPGRADE TO Whoop 5.0 – less accurate device than 4.0,” containing reports from long term users who believe their 5.0 performs worse than their previous 4.0.
Interestingly, not everyone has the problem. Another 5.0 MG user reports their WHOOP and Garmin being within 20–30 steps during a running workout.
That makes me even more interested in whether some individual WHOOP 5.0 devices are faulty, rather than accepting that large discrepancies are simply how the new algorithm works.
I’ve now asked WHOOP to replace my 5.0 so we can test exactly that.
I upgraded expecting an upgrade
This is probably the most disappointing part.
I was happy with WHOOP 4.0.
I moved to WHOOP 5.0 expecting the newer generation to improve on the previous product. Instead, I’ve spent months collecting data and troubleshooting a basic metric that my previous WHOOP did not consistently struggle with.
WHOOP is also an ongoing paid subscription, not simply a one off hardware purchase.
There are several highly competitive fitness and health trackers available now. I don’t think it’s unreasonable for customers upgrading from WHOOP 4.0 to expect WHOOP 5.0 to perform at least as reliably as the product it replaced.
At this stage, my experience feels less like an upgrade and more like I’ve been pushed onto a newer generation that has downgraded one of the product’s basic measurements.
I’m interested to hear from other 4.0 → 5.0 users specifically:
Did your step accuracy change after upgrading? What are you comparing it against, and roughly how large is your daily discrepancy?