Review Methodology
Most tech publications don’t explain how their reviews work. You read a score, you read a verdict, and you’re expected to trust it. That’s a reasonable thing to push back on, especially when a review is helping you decide whether to spend $100 or $1600 on a phone you’ll use every day for the next two years.
This page explains exactly how AndroidBeep reviews work: how we decide what to cover, how long we test things, what we actually test for, how scores are assigned, and what happens when a manufacturer sends us a device for free. None of this is a promise of perfection. It’s a description of the process, explained specifically enough that you can decide for yourself how much weight to give our conclusions.
If something in a published review doesn’t match what we describe here, we want to know about it. The contact details are at the bottom of this page.
What We Decide to Review and Why
The editorial team makes coverage decisions based on one question: Is this product relevant and useful to AndroidBeep readers? That filters out a significant amount of the Android market. A phone that will only be sold in two markets where we have minimal readership is not a priority. A phone announced at a price that puts it in the same bracket as three other devices we’ve already reviewed recently may get pushed down the list.
The categories we actively review are Android smartphones, tablets, smartwatches and fitness trackers, wireless earbuds and headphones with Android-focused features, and Android-specific accessories where there’s a meaningful editorial case to make. Apps and services are reviewed selectively when they’re significant enough that readers genuinely need help evaluating them.
No manufacturer can pay to have a product reviewed. No advertiser can guarantee coverage by spending money with us. Review decisions are made editorially and stay editorial.
Where Review Units Come From
Some products we review are bought at retail, with our own money, from standard sales channels. Others are provided by manufacturers or PR agencies as loaner units or press samples specifically for review. When a device was provided rather than purchased, the review says so near the top.
That disclosure exists because readers deserve to know. What it doesn’t mean is that a provided device gets treated differently. The testing process is identical regardless of where the unit came from. A phone sent to us by a manufacturer goes through the same protocol as one we pulled off a store shelf.
Receiving a device creates no obligation to publish a review, to publish a positive review, or to share review content with the manufacturer before it goes live. We don’t offer pre-publication review access. We don’t give manufacturers the opportunity to request changes. Loaner units are returned after testing when that’s requested; otherwise, we note in the review that the device was kept.
Minimum Testing Periods
Rushed reviews produce unreliable conclusions. A phone that performs well during a two-day press event might have a battery that degrades noticeably under sustained real-world use, a heat issue that only appears after extended gaming or navigation, or a camera that struggles in conditions the review never tested. We’ve seen all of those things, which is why we have minimum testing periods before a final review is published.
Smartphones: ten days as a primary daily driver before a final verdict is written. This is a floor, not a ceiling. Complex devices, or devices that require more time to evaluate fairly, take longer.
Tablets: seven days of active use across multiple contexts, including productivity tasks, media consumption, and, where relevant, stylus input and keyboard accessories.
Smartwatches and fitness trackers: fourteen days minimum, because health and fitness tracking accuracy requires data across multiple activity types and sleep cycles to evaluate meaningfully.
Wireless earbuds and headphones: five days across varied listening environments, including commuting, exercise, and quiet indoor use.
Accessories: testing period varies by product type but is always long enough to assess whether the core function holds up under normal use.
First impressions and hands-on previews, described in more detail below, are published before these minimums are met. They’re clearly labeled as such.
How We Test Products
Battery Life Testing
Every smartphone and tablet we review goes through a standardized battery test: screen brightness set to 200 nits, volume at 50 percent, and a looping video file played until the battery dies. We record the result and compare it against other devices tested under the same conditions.
That number appears in the review, but it sits alongside real-world battery observations, because a controlled test doesn’t capture how a phone performs when you’re using GPS navigation for two hours in direct sunlight or running background sync across a dozen apps. Both matter and both appear in the review.
Camera Testing
Camera testing follows a fixed set of conditions for every device: outdoor daylight, indoor artificial light, low light without flash, low light with night mode where available, and video at the maximum resolution the device supports. Portrait mode is tested separately where applicable.
Cameras are compared against other phones in the same price range, not against phones that cost twice as much. A $350 phone’s camera is evaluated as a $350 phone camera. We publish sample images in reviews so readers can make their own judgment rather than relying solely on ours.
Performance Testing
We run benchmark tests using AnTuTu and Geekbench on every smartphone we review. Those numbers appear in the review as reference data. We don’t let benchmark scores drive the verdict, because benchmark performance and real-world performance don’t always track together.
What we’re looking for in practice: how the phone handles switching between demanding apps, whether it maintains performance under sustained load, how it handles heat during extended use, and how responsive it feels during everyday tasks like scrolling, typing, and opening apps.
Software Evaluation
We test the software as it ships and note the Android version and security patch level in every review. For Samsung devices, we document the One UI version. For Pixel devices, we note which version of stock Android the device is running at the time of review.
Software evaluation covers the pre-installed app situation, any significant manufacturer additions to the interface, update history for the manufacturer where it’s relevant, and any software-specific issues we encountered during testing.
How We Score Products
How the Rating Scale Works
AndroidBeep uses a ten-point scale for reviews. Scores are assigned based on how a product performs relative to its price category and what a reasonable buyer would expect at that price. A score reflects real-world utility, not a checklist of features.
What Each Score Means
- 9 to 10: Among the best available at this price. Recommended without significant reservations.
- 7 to 8: Good product with real strengths. Worth buying for most people in the target audience, with clearly noted limitations.
- 5 to 6: Average. Does its job without distinction. Specific buyers may find it worthwhile; most should look at alternatives.
- 3 to 4: Below average. Meaningful problems that affect everyday use. Hard to recommend at the listed price.
- 1 to 2: Significant failures. We rarely assign these, and when we do, the review explains specifically why.
Scores are arrived at after testing is complete, not assigned before writing begins and then justified. The written review and the score should be consistent with each other. If a review reads positively but the score is a 6, or vice versa, that’s an editorial failure. Readers who spot that kind of inconsistency are encouraged to point it out.
What Happens After a Review Is Published
When a significant software update changes how a device performs, we update the review. Update notes appear at the top of the article with a date and a plain description of what changed. We don’t rewrite original reviews to hide what the device was like at launch; we add to them.
If a manufacturer issues a fix for a problem we documented in a review, we note that in the review. If a problem we didn’t document turns out to be widespread after publication, we add a note covering it.
First Impressions, Previews, and Buying Guides
A first impressions article is published when we’ve spent time with a device but haven’t completed our full testing period. It’s clearly labeled as first impressions. Readers should treat it as an early, incomplete picture.
A hands-on preview is typically published after a press event or manufacturer briefing where we had access to a device but didn’t take it away for extended testing. These are observations, not verdicts.
A buying guide is an editorial recommendation based on accumulated review experience across multiple devices. Buying guides are updated when the competitive landscape changes. Affiliate links appear in buying guides with disclosure at the top of the article.
Conflicts of Interest and Affiliate Links
Writers are required to disclose any relationship with a manufacturer whose product they’re reviewing before the review is assigned. That includes past paid work, personal relationships, and financial stakes. Where a conflict exists, the review is reassigned.
Affiliate links in reviews and buying guides are disclosed at the top of the article. The links are added to products we’ve already decided to recommend. A product that earns a poor score doesn’t receive an affiliate link regardless of commission rate. A product we recommend strongly gets an affiliate link regardless of whether one is available.
The Editorial Process
Every review draft is edited by a senior editor before publication. Editing covers factual accuracy, consistency between the written assessment and the score, and whether the review gives readers enough information to make a real decision. The editor has the final say on whether a review is ready to publish.
Fact-checking covers specifications against primary manufacturer sources, benchmark results against our own recorded data, and any comparative claims against other reviews we’ve published. Claims that can’t be verified don’t go in.
Responding to a Review or Flagging an Issue
If a published review contains a factual error, email editor@androidbeep.com with the article URL and a description of the problem. We investigate every report and respond within two business days.
If you think a product we haven’t reviewed deserves coverage, use the same address. We read every suggestion. We can’t promise coverage for every request, but reader recommendations do influence the coverage queue.
If you disagree with a verdict, we’re happy to hear why. Disagreement isn’t a factual error and doesn’t trigger the corrections process, but informed pushback from readers who’ve used a device we reviewed is worth reading, and occasionally it changes how we think about a product.