Every review on NinjaCreamer.com follows the same written protocol so that a number on one page can be compared with the same number on another. This policy sets out how machines are bought, how they are tested, how scores are calculated, how mistakes are corrected and what we will and will not do with photographs and writing tools. Where a review departs from this protocol, the review says so.
Purchase policy
We buy every machine we review on Amazon, through the same listings we link to, with the site’s own money. We do not request review samples from SharkNinja or any retailer, and we decline them when offered. We do not return a machine after testing. We own all six current Ninja Creami models (NC301, NC299AMZ, NC351, NC501, NC601 and NC701) plus accessory pint packs for each of the three pint systems, and we keep using them so that the 30 day notes in each review come from continued use rather than memory. Each review states the date testing began.
Test protocol
The four bases
Every machine is tested on the same four bases, chosen to cover the four kinds of result a Creami owner is likely to want.
- Full-fat vanilla: heavy cream, whole milk, sugar, vanilla. Spun on Ice Cream.
- Protein: Fairlife skim milk, whey isolate, sugar-free pudding mix. Spun on Lite Ice Cream, or on CreamiFit where the machine has it.
- Mango sorbet: canned mango in syrup, blended. Spun on Sorbet.
- Dairy-free oat: full-fat oat milk, coconut cream, sugar, guar gum. Spun on Lite Ice Cream, or on Dairy-Free where the machine has it.
The recipes for these bases are published in full in our base ratios guide so that anyone can repeat the test.
Freezer, freeze time and the flat-top rule
Every base is frozen in the pint supplied with the machine for 24 hours in an upright freezer set to 0 F, checked with a fridge thermometer before each batch. The pint sits on a level shelf so that the top of the base freezes flat. A base that freezes tilted or domed stresses the paddle and produces a crumbly first spin, so a pint that is not level is discarded and refrozen rather than tested. We call this the flat-top rule and we apply it to every pint on every machine.
The re-spin rule
After the first spin we score the texture. If it is crumbly or powdery we re-spin once on the same program. If it is still crumbly after one re-spin, we add 1 tbsp of liquid matched to the base (milk, oat milk or syrup) and re-spin a second time. We do not re-spin more than twice. The number of re-spins a base needs is recorded and reported in the test results table, because it is the clearest single measure of how well a program matches a base.
Texture score
Texture is scored from 1 to 5 after the re-spin rule has been applied: 1 is icy or powdery and will not hold a scoop, 3 is scoopable with visible ice or crumb, 5 is dense, smooth and holds a clean scoop straight from the machine. Every test is photographed after the first spin and after any re-spin.
Cycle time
Cycle time is recorded with a stopwatch from the moment the program button is pressed to the moment the motor stops, in seconds. Re-spin time is recorded the same way.
Noise
Noise is measured with a phone dB meter held 1 m from the front of the machine at counter height, recording the peak reading during the Ice Cream program on the full-fat vanilla base. We report the peak, not the average, because the peak is what you hear in the next room.
Footprint and weight
Width, depth and height are measured with a tape measure to the nearest tenth of an inch, with the lid closed and again with the lid fully open, because the open height is what determines whether the machine fits under a cabinet. Weight is the motor base only, on a kitchen scale.
Dishwasher test
Pints, lids, the Creamerizer paddle, the outer bowl and the outer bowl lid go through 20 cycles on the top rack of a domestic dishwasher on a normal program. After the twentieth cycle we record cloudiness, warping, lid seal and any wear to the paddle. On the NC701 the dispenser nozzle and handle assembly are included.
The 30 day update
Thirty days after testing begins, each review is updated with what has changed: paddle and lid lock wear, pint and lid condition, any change in noise, what broke, and which programs we have stopped using. The score is reviewed at the same time and changed if the evidence warrants it, with a note explaining why.
Scoring
Each machine receives a score out of 10 built from six weighted criteria: texture 30 percent, programs 20 percent, capacity 15 percent, noise 10 percent, ease and cleaning 15 percent, and value 10 percent. Texture is the average of the four base scores after the re-spin rule. Programs rewards useful programs that matched their base, not the raw count. Capacity is servings per pint and pints included. Noise is the measured dB, with readings under 80 scoring highest. Ease and cleaning covers the lid lock, button clarity, paddle removal and the dishwasher results. Value is the model’s position relative to its siblings for what it adds, judged without reference to price, because we do not print prices and they change weekly. The rubric and the current scores are listed on our about page.
Independence
NinjaCreamer.com is an Amazon Associate and earns a commission on qualifying purchases made through its links. That commission does not vary by model and has no input to the scoring. We do not accept payment for placement, sponsored posts, or advertising of any kind. We have no relationship with SharkNinja. Our disclosure page has the full statement.
Corrections
If a measurement, a compatibility claim or a program description on this site is wrong, we want to know. Email hello@ninjacreamer.com or use the contact form. A confirmed error is corrected on the page where it appears, and a dated correction note is added at the bottom of that page stating what changed. Errors in pint compatibility are treated as urgent and corrected the same day, because a wrong answer there costs a reader a pint pack that does not fit. We do not silently edit measurements.
AI and writing tools
Every test on this site was carried out by hand in our kitchen, and every measurement was taken by a person with a stopwatch, a dB meter, a tape measure or a scale. We use software to check spelling, to catch forbidden characters and to validate structured data. We do not publish test results, measurements or product experiences that were generated rather than observed, and no number on the site comes from anywhere other than our own testing or the manufacturer’s printed specifications.
Photo policy
All photographs on NinjaCreamer.com are original and were taken by us of machines we own, on a white or grey surface, in the same lighting, so that product cards and comparison pages are visually consistent. We do not use SharkNinja product renders, lifestyle photography, marketing imagery or logos anywhere on the site, and we do not use stock photography of ice cream. Where an Amazon product image is shown it is loaded through Amazon’s own tools. Every review carries at least ten original photos covering the box contents, the machine from the front and side, the control panel, the paddle, the lid lock, the pint before and after processing, a scoop close-up and the dishwasher rack. Every recipe carries at least four: the frozen base, the result after the first spin, after the re-spin and the final scoop. Images are not retouched beyond cropping and exposure correction.
Who is responsible
Mike Lala tests every machine, writes every page and is responsible for every number on the site. Questions about this policy go to hello@ninjacreamer.com.