TL;DR
Routinely is a free iOS app that builds a skincare routine from a quiz, scans product labels, and grades them against your goals. The quiz is thoughtful. The recommendations skew toward more steps than most readers need. Useful as a starting framework for beginners; over-prescribes for everyone else.
Routine builders have a reputation problem. Most of them are gamified product placement: take a quiz, get a routine that conveniently uses the partner brand’s products, repeat. Routinely by MWM is one of the better ones I have used, and it is still a routine builder, which means it is still prone to the genre’s biggest flaw: assuming more steps means more results.
I tested it against a four-step minimalist routine for six weeks across three readers. The results are more interesting than the genre usually allows.
What Routinely is and isn’t
Routinely is a free iOS app from MWM. The onboarding is a quiz, around fifteen to twenty minutes long, covering skin type, current concerns, climate, age, lifestyle stressors, and history with actives. Based on the responses, the AI generates a morning and evening routine with specific step categories and example products. You can then scan labels with the barcode and OCR scanner, and the app grades each product against the routine slot it is supposed to fill and your stated goals.
There is a skin journal layer, themed flows (travel, stress relief, self care), and usage logging. The AI updates the routine over time based on what you log.
It is not a diagnostic tool, not a dermatologist substitute, and not a substitute for understanding what your actives are actually doing. The quiz can place a retinoid in your routine without ever explaining what it is. That is the kind of gap a beginner can fall into.
Who it’s for
Genuine beginners. People who have never built a routine and need a starting framework with categories rather than a vague “cleanser, moisturizer, sunscreen” instruction. Travelers using the themed flows for jet-lagged skin. People who like usage logging and routine streaks (the gamification works for some readers, against others). Anyone who has bought twelve products on TikTok and needs help triaging them.
It is not for someone with an established routine who is happy with it. The app will offer to add steps. The honest answer is usually no.
It is also not for someone with a real condition (active rosacea, perioral dermatitis, severe eczema). The app does not know the difference between irritation-prone skin and a clinical inflammatory condition. For those, see a derm before you see an app.
The features that matter
The quiz is the keystone. It is the most thoughtful onboarding quiz I have seen in the category. It asks about humidity, hard water, sleep, and the kind of cleansers your skin has historically liked. Most builders skip this and over-fit on “oily versus dry,” which is the least useful dimension. Routinely’s quiz produces a more textured profile.
The barcode and OCR scanner is the second feature worth flagging. The grading logic against your routine slot is decent. Drop a moisturizer that is mostly an exfoliating toner formulation into the moisturizer slot, and the app flags it. The grading is less reliable for nuanced cases (a moisturizer with active acids in a low concentration meant to be supportive, for instance), but for obvious mismatches the flag is useful.
The themed flows are the most surprising. Travel flow drops actives, adds gentler hydrators, suggests SPF reapplication strategy on flights. Stress flow drops aggressive treatments and emphasizes calming actives. These are well-built and align with what most editorial advice already says.
Skin journal logging is the fourth feature. Useful for some readers, friction for others. The app’s pattern detection (which logs correlate with which skin events) needs a lot of data to be useful, more than most casual users will produce.
The over-prescription problem
This is the section the brand will want me to soften. I will not. The default routine the AI generates is usually seven to nine steps. The slow-skincare position, which Elelaf has held since launch, is that most readers do better with three to five. A cleanser, a moisturizer, a sunscreen, and one well-chosen active, sometimes a second. That is enough for most people, most of the time.
The app’s seven-step morning routine includes a toner most readers do not need, an essence most readers cannot tell apart from their serum, and a separate eye cream that is almost always a duplicate of a good face cream in slightly different packaging. The app will hand you all three. The slow-skincare answer is: pick one.
The AI does this not because the team is greedy. It does it because the underlying logic of the algorithm rewards specificity (more steps means more recommendations means more apparent value), and most quiz-based builders use the same logic. The fix is editorial discipline, not algorithmic. Use Routinely as a categories map. Then cut the routine down by 40 percent.
Real-world test
I tested Routinely for six weeks with three readers. The most useful result was from a 23-year-old college student with no established routine, sensitive normal skin, and zero idea where to start. Routinely’s quiz produced an eight-step routine. I cut it to four (gentle cleanser, ceramide moisturizer, mineral SPF, azelaic acid 10 percent in the evening twice a week, scaling up to nightly).
Six weeks later, her redness was visibly down (from constant low-level flushing across the cheeks to flushing only after heat or stress), her texture was smoother, and she had spent $73 on her entire routine instead of the $214 the app’s recommended product list would have cost. The app’s quiz gave her the starting framework. The slow-skincare cull is what made the framework usable.
The second tester, a 36-year-old with established hormonal acne, did not gain much from the app. Her existing routine (the one she had built with a dermatologist) was simpler and better than what Routinely proposed. The app tried to add a vitamin C she did not need and a separate eye treatment she did not need. She closed it after two weeks.
How it stacks against the Glass app
The Glass app is the closest competitor and is more popular. Glass is better at community features and product discovery. Routinely is better at structured onboarding and routine generation. For a beginner who wants a routine handed to them, Routinely wins. For an enthusiast browsing other people’s routines and tracking favorites, Glass wins.
Against a one-time dermatologist consultation, Routinely is cheaper and faster but gives you generic advice. A dermatologist takes one look at your skin and tells you what you specifically need. The app is a useful precursor to that visit, not a replacement.
Against writing your own routine after reading three good editorial guides, the app is a faster starting point. The editorial route gives you the why, which the app does not. The trade-off is roughly four to six hours of reading versus twenty minutes of quiz answering.
Pairing it with a slow skincare philosophy
The framework Elelaf recommends as a sanity check on any AI-generated routine is the three-step minimalist routine. Start there. Anything the AI adds on top has to earn its place by addressing a real, specific concern. If you cannot name the concern, the step does not belong.
If you are a beginner, the skincare 101 beginner’s guide is the longer read. If you are not sure of your skin type, the identify your skin type piece is the prerequisite to any quiz, in any app. The app’s accuracy depends on the accuracy of your answers, and most readers misidentify their skin type on the first try.
The hero product Elelaf pairs with a thoughtfully minimal routine is the Mindful Masks line, used once a week, as the ritual rather than the workhorse. The workhorse stays the simple core. The mask is the punctuation. For more, our skinimalism tag collects the philosophy we keep returning to.
FAQ
Is Routinely actually free? The base app is free. Some advanced features and theme flows sit behind a paid tier. The free tier is enough for the routine builder.
Will it recommend products I have to buy? Yes, it suggests products. You do not have to buy them. The structure of the routine (the category slots) is the useful part.
Can I trust the safety scores on scanned products? Treat them as one input. The grading is good for obvious mismatches and less reliable for nuanced cases. Cross-check with a second source for anything you are about to buy.
What if my skin is sensitive or has rosacea? Use the app’s gentle profile or its stress flow, and be ready to cut steps. The default routine for sensitive skin is still too aggressive for many rosacea readers.
Does it handle pregnancy? The pregnancy profile exists and screens for retinoids and a few other actives. Cross-check with the conservative pregnancy list and your OB. Our pregnancy-safe piece has the full version.
How long do I need to use it before the AI gets useful? The routine is generated immediately. The pattern detection layer (which correlates your logs to skin events) needs four to six weeks of daily logging to produce anything meaningful.
Sources
AAD.org/” rel=”noopener” target=”_blank”>American Academy of Dermatology, position statement on consumer skincare routines and product layering, 2024. Journal of Cosmetic Dermatology, review on quiz-based personalization in consumer skincare apps, 2023.