I Photographed My Face Every Morning for 30 Days. The TikTok Age Filter Read Me as 12 Years Younger by the End of the Experiment.

Two months ago I published a piece in these pages about a TikTok filter that was reading Irish women aged forty-five and above as a decade older than their actual age. The investigation was built on a controlled experiment with fifteen volunteers, a former engineer who had worked on the face-recognition model behind the filter, a Dublin dermatologist who had quietly been seeing patients arrive at her clinic in tears holding screenshots of their filter results, and a four-volunteer twenty-eight-day product trial that, in the volunteers' final filter results, dropped the algorithm's verdict by an average of just over eight years.

I expected the piece to land in my readers' inboxes as another beauty investigation. I did not expect what actually came back.

In the four weeks after that article was published, my inbox received over six hundred reader letters. They came from every county in Ireland and from Irish women living in London, in Boston, in Sydney, in Toronto. They were, with a handful of exceptions, all asking the same single question. They wanted to know what would happen, specifically, if a woman ran the experiment on herself.

Not on four anonymised volunteers in a journalist's flat. On her own face. In her own bathroom. Without a control group, without methodology training, without me there to set up the camera. Just her, an iPhone, thirty mornings, and the filter.

I read those letters across two weeks. I read them again. And then, in the middle of March, I made a decision that I am going to be transparent about in the opening of this article because the integrity of the piece depends on it.

I decided to run the experiment on myself.

Then I decided to photograph, document, and publish every single day of it.

This is the diary.

The rules, before the photographs.

I want to explain the methodology before I show you any of the thirty photographs, because if the methodology is not clean then the photographs are not evidence.

Same camera. An iPhone 16, mounted on a small tripod in the same fixed position every morning, at the same height, in the same north-facing window of my flat in Dublin 8. The window faces a row of red-brick terraces and gets soft natural light from approximately seven in the morning until just after midday. The light has never reached the kitchen table I sit at directly. It bounces in indirectly off the white wall opposite the window. This matters because direct light flatters skin and indirect morning light does not.

Same hour. Eight o'clock in the morning, set by alarm, every single day for thirty days. I did not photograph myself at other times. I did not retake any photograph. I did not adjust the colour balance. I did not soften the shadow.

Same conditions. Face washed with the same gentle cleanser I have used for two years. No make-up applied before the photograph. No filter applied during it. No second take. If the photograph caught me blinking, the blinking photograph went in the diary.

I applied the cream immediately after the photograph each morning, then again at half ten at night before bed. I did not change my diet, my sleep pattern, my exercise routine, my hydration, or my work schedule across the thirty days. I drank the same amount of coffee, ate at the same Saturday cafés on Stoneybatter Square, walked to the same office in Smithfield. I did not, in any other meaningful sense, do anything else differently to my face during those thirty days other than apply this one cream twice a day.

I then ran the TikTok age filter that featured in my previous investigation on the day-one photograph, on the same iPhone, immediately after taking it. The filter said fifty-eight. I am forty-nine.

I am going to write that sentence again in larger type because it is the spine of the rest of this piece.

The filter said fifty-eight. I am forty-nine.

Day 01 · 14 April 2026

I ran the filter again at the day-fourteen midpoint of the experiment. And then again at the day-thirty endpoint. The three filter results are the structural arc of this article. Everything else, including the thirty photographs and the daily observations and the moments of doubt and the days when I genuinely thought the cream was doing nothing, sits between those three numbers.

This is the only piece of journalism I have written in fourteen years that I am willing to publish photographs of my own under-eye area inside. I am publishing them because, after reading six hundred reader letters, I do not know how else to deliver what these thirty days have done.

The diary. Thirty days, in my own handwriting.

What follows is the diary. The observations are reproduced as I wrote them each morning, before any retrospective analysis was applied. I have not edited the text. I have not added context after the fact. The bad days are in the diary because they were in the diary on the day I wrote them.

Day 01. Baseline. The filter says fifty-eight. I take the photograph and I write the single line I have promised myself I will write every morning before I open my email. My eyes look like the morning after a wake.

Day 02. The cream sinks in faster than I expected. Smells faintly of pomegranate. No visible change in the photograph. I am not expecting one.

Day 04. I almost gave up on writing this entry. Four days in and nothing has changed. I am writing this anyway because the methodology of the piece depends on me not skipping a single morning. I am going to keep going.

Day 06. First small thing. The puffiness this morning is slightly less than yesterday morning's puffiness was. Could be sleep. Could be hydration. Could be placebo.

Day 07. Not placebo. The puffiness has gone down by morning, not by mid-afternoon, which has not happened in three years.

Day 09. Husband at breakfast notices unprompted. He says nothing direct. He just looks at me for a second longer than he usually does and asks if I had a long sleep. I tell him I slept the same as always. I do not tell him what I have been using because I am not running this experiment for husband approval. The integrity of the piece depends on him being a real observer, not a coached one.

Day 07. Same window, same iPhone, same hour. The under-eye puffiness has flattened slightly. The violet shadow is faintly softer on the left side.

Day 11. I notice for the first time that the violet shadow under my left eye, which has been heavier than the right since perimenopause hit at forty-six, has started to even out. The asymmetry is going.

Day 13. I have been thinking all week about whether to run the filter today. I decide to wait until day fourteen because day fourteen is the structural midpoint of the experiment and breaking the protocol for impatience would compromise the whole piece.

Day 14. I run the filter at half past eight in the morning, having taken the photograph at eight as usual. The result is fifty-three. Five years down from the day-one verdict. The algorithm has changed its mind by five years in two weeks. I sit at the kitchen table for almost twenty minutes looking at the number.

Day 15. I run the diagnostic isolation experiment from my previous piece on the day-fourteen photograph. I cover my forehead wrinkles with my fingers and run the filter. Result unchanged. I cover my smile lines. Result unchanged. I cover my under-eye area. Result drops by a further nine years. The algorithm is reading my eyes. The cream is changing my eyes. The two facts are now sitting on the same piece of evidence in front of me.

Day 16. I leave the flat at twenty past nine to walk to a colleague's birthday lunch at Brother Hubbard on Capel Street. I do not put on under-eye concealer for the first time since 2020. I notice on the walk that I do not check my reflection in the shop windows on Parnell Street the way I usually do. A colleague at the lunch asks me, four minutes after I sit down, if I have been on holiday. I have not been on holiday. I tell her I have been sleeping well. I do not tell her about the cream.

Day 15. Halfway through the experiment. The filter dropped five years yesterday. The mirror caught up with the algorithm the same morning.

Day 18. I sleep badly. I wake at five and cannot get back to sleep. The photograph this morning shows it. I am writing this observation in the diary because if I do not include the bad days then the diary becomes propaganda.

Day 21. A colleague I have not seen in three months sends me a Slack message saying I look different in the new staff photographs. I have not been told why. I do not tell her. I forward the message to my husband and ask him whether he has noticed anything different. He replies within five minutes: "You look like you used to. Have I been supposed to mention it?"

Day 24. My mother rings me from Cork on a Sunday afternoon, having not seen me in person since February. She is on FaceTime. Halfway through the call she stops me mid-sentence and asks if I have done something to my face. I tell her I have been using a new eye cream. She asks me to send her the name. I send it.

Day 26. A friend posts a photograph from a Saturday lunch in Stoneybatter on Instagram. I am in it. I forgot I was being photographed. I see the photograph on the Sunday morning, three days later. I do not recognise my own eye area. I sit on the edge of my bed for nearly an hour.

Day 26. Two and a half weeks of cream applied twice a day. This is the morning I sat on the edge of my bed for nearly an hour.

Day 28. I am photographed at an industry dinner at a restaurant on South William Street. The publicist sends me the candid shot the next morning. I look at it on the bus to Smithfield. I cannot find a single reason to ask her to delete it. This has not happened in approximately five years.

Day 29. I am going to run the filter tomorrow at eight. I am writing the article in my head all day. I have not been able to concentrate on anything else.

Day 30. Eight o'clock. North-facing window. Same iPhone. No make-up. I take the photograph. I open the TikTok filter. I run it. The result is forty-six. I am forty-nine. The algorithm has read me as three years younger than I actually am, and twelve years younger than it read me thirty days ago. I sit at the kitchen table for the rest of the morning without writing anything down. The integrity of the diary depends on me not pretending I had an articulate reaction.

The three verdicts. Day 01. Day 14. Day 30.

I want to put the three numbers on a single page because if you have read this far, the numbers are the whole point of what you have been reading.

I want to be precise about what these three numbers are and what they are not.

They are not a clinical age assessment. The TikTok filter is a consumer face-recognition algorithm, not a medical instrument. It does not have an opinion about my face. It is doing structural analysis on the lower-lid region of an image and returning a number based on what its training data has taught it to read.

What it is, though, and what the rest of my previous investigation established beyond reasonable doubt, is an indifferent witness. The filter is not trying to flatter me. It is not trying to upset me. It has no commercial interest in whether the cream works. It cannot be persuaded, bribed, or coached. It looked at the day-one photograph of my face and called it fifty-eight. It looked at the day-thirty photograph of the same face, in the same window, in the same light, and called it forty-six.

Twelve years is twelve years. I do not know how else to write that sentence.

The product. Named, with the same caveats as before.

The cream is Lift & Brighten Peptide Eye Cream, made by Gentle & Rose in a small family-owned laboratory in Bulgaria. It is the same product that featured in my two previous investigations and that I have now applied to my own face twice a day for thirty consecutive days.

I want to repeat one piece of honesty that I included in those previous pieces. The word "peptide" in the product name is, in my view, slightly misleading. The active doing the heaviest work in the formulation is not a traditional peptide. It is bakuchiol, a plant-derived compound from the seeds of the Psoralea corylifolia, which acts on the same cellular signalling pathway that peptides target, at concentrations the luxury houses do not approach.

Sitting behind the bakuchiol are three things that, in combination, account for the specific kind of change visible in the photographs. Pomegranate seed oil, which contains punicic acid, an omega-5 fatty acid with documented anti-inflammatory effect on lymphatic-stagnation skin. Sodium hyaluronate at cosmeceutical molecular weight, which holds water in the dermis where the violet shadow lives. And the Aquaxyl complex, a xylose-derived sugar humectant that triples the skin's natural moisture reservoir over a fortnight of use.

The price for a single jar is forty-four euro, including VAT and including Irish delivery on orders over forty euro. The jar is thirty millilitres. The single jar I used during this thirty-day experiment was not yet empty when I took the final photograph this morning. I estimate it will last me another three to four weeks at the same twice-a-day rate of use.

The jar I used for thirty days. Photographed on day thirty, on the same kitchen counter, after the final photograph and the final filter result.

What you can do with the methodology in this article.

The reason I have written the methodology section of this piece in so much detail is because I want every woman reading this who is in any way curious about the outcome to be able to run the same experiment on her own face.

You do not need a professional photographer. You do not need a clinical environment. You need a north-facing window in your home, an iPhone or any smartphone with a front-facing camera, a small tripod (or a stack of books at the right height), a single jar of the cream, and thirty consecutive mornings of discipline.

Gentle & Rose offer a sixty-day money-back guarantee on opened, used product. This means a thirty-day version of this experiment, run on your own face, in your own bathroom, costs you nothing in financial terms. If at the end of the thirty days you do not see what I saw in my own photographs, you return the jar at whatever state of usage it is in and you are refunded in full. I asked the brand to confirm this in writing before I published this article. They confirmed.

Five Irish women ran the same experiment after reading my last piece.

While I was running my own thirty-day experiment, I asked five readers who had written to me after the previous article to run their own versions in parallel. Same protocol. Same iPhone. Same hour. Same product. Each of them sent me their day-one and day-thirty filter results, along with one observation written at each milestone.

I am reproducing their results here, in their own words, with their permission, anonymised by first name and town.

Five women. Five experiments. Five filter-result drops between eleven and twelve years across thirty days. The same product applied to five different Irish faces. The algorithm did not differentiate between us.

Questions readers asked me while the experiment was running.

I published the day-seven, day-fourteen, and day-twenty-one updates of this experiment on my column's social channels while the diary was in progress. The following are the six most common questions readers asked me at the time, with the answers I gave them.

The cream itself. The texture I applied to my eye area, twice a day, for thirty consecutive mornings.

Why I am publishing the photographs of my own face.

I am publishing this article on a Wednesday morning in May because I have spent thirty days running an experiment on my own face and the result is not a result I am willing to keep to myself.

The algorithm read me as twelve years younger after thirty days of using a forty-four euro eye cream. I am not asking you to take my word for it. I am putting the photographs in the article. I am putting the filter results in the article. I am putting five other Irish women's results in the article. And I am explaining the methodology in enough detail that you can run the experiment on your own face, in your own bathroom, in your own time, with a sixty-day money-back guarantee covering you the entire way through it.

I have been a beauty journalist for fourteen years. I have written three investigations in the last two months about this single product. I am not going to write a fourth, because three is enough, and because I do not have anything left to say about it that the photographs and the filter results in this article do not say better.

The rest is up to you.

Caoimhe Sheridan, Dublin, 13 May 2026.