Updated:
How to research fashion trends: a step-by-step guide to trend analysis for designers
Fashion trend analysis means counting the signals that repeat across publications, runways and stores to show what is rising. Here is how to research fashion trends in six steps, from the first question to a finished trend board.
What is fashion trend analysis, and how is it different from forecasting?
In practice, you pick a period and log what repeats in it (a product, fabric, print, silhouette or color) across publications, runways, stores and the street, looking for evidence that a signal is spreading.
Forecasting goes further and anticipates the seasons ahead, often well in advance, using expert teams, archives and runway, retail, search and social data. Analysis is its foundation.
| Approach | Question | Relies on | Output |
|---|---|---|---|
| Tracking | What's talked about now? | Daily news and images | A reading list |
| Analysis | What's growing or fading? | Tagged, counted items | Evidence-backed trends |
| Forecasting | What will matter in seasons ahead? | Archives, experts, retail and search data | Season directions |
For an independent designer or a small brand, analysis is the realistic job. The six steps below show how.
How fashion trends are identified: where the signals come from
Signals arrive through several channels, each at its own speed:
- Runways: designers' proposals for the season. They set direction, but not every look reaches stores.
- Fashion publications: what editors keep covering, showing how far a signal has spread.
- Trade and textile news: fabric and yarn coverage, showing how close a material is to production.
- Retail: what brands actually make and price.
- Street style: what people wear; fashion-week street style is often styled, so count everyday looks separately.
- Search interest: what people are curious about, visible in free tools like Google Trends.
- Culture: TV, music, exhibitions and sports often explain why a color takes off.
A signal becomes a trend when it repeats across several channels in the same period; one show isn't enough. To spot a trend early, watch for a runway signal reaching trade news or retailers' new-arrivals pages.
Step 1: Narrow the question
“What's in this season?” can't be analyzed; the answer is everything. Narrow it with four fields:
- Product category: knit tops, denim pants, eveningwear.
- Gender and age: women, men, teens, kids.
- Region or market: the US, Europe, the Gulf.
- Season: the one you're designing for.
For example: “Which prints and colors are gaining ground in women's knit tops for spring/summer in Europe?” Put it at the top of your research file: it says what to collect and what to skip.
Step 2: Build a source list you can cite
Before adding a source, check that it publishes regularly, covers your market and dates every item with a link to the original.
- International and local fashion magazines
- Runway and collection archives
- Apparel and textile trade publications
- Fabric and apparel trade fairs
- Retailers' new-arrivals pages
- Search trend tools
Keep it balanced: international magazines alone miss your local market, and social media alone measures an algorithm.
Step 3: Tag every item the same way
Fill in the same fields for every item. If tags drift, periods stop comparing: “cream” one week and “ecru” the next count as two colors.
| Field | Examples | Read it from |
|---|---|---|
| Product category | Dress, jacket, knit top | Image and headline |
| Fabric and finish | Linen, satin, suede | Fiber from the text; visible finishes from the image, marked “by appearance” |
| Stripes, check, floral | Image and text | |
| Silhouette | Oversized, A-line, wide leg | Image and text |
| Design detail | Ruching, asymmetric collar | Image and text |
| Color | Terracotta, ice blue, olive | The garment only |
| Season | Spring/summer, fall/winter | Text |
| Source and date | Publication, link, date | The item's page |
Don't guess fiber from a photo: silk and polyester satin, or wool and acrylic knits, look alike on screen. Read color from the garment only; background, skin and hair muddy a palette.
Step 4: Compare shares, not counts
Raw counts mislead: in a busy week, everything goes up. Use share instead, the proportion of a period's items that carry a tag.
Pick two equal periods
Say, the last four weeks against the four before. Single weeks are noisy: a fashion week tilts the mix toward runway coverage.
Work out each tag's share
Divide the items carrying the tag by all items in the period.
Count the sources
Coverage across many publications says more than repeated coverage in one.
Set thin periods aside
With very few items, one story swings the share; leave such periods out or flag them.
You end up with three lists: gained share, lost share and most frequent. Most frequent isn't the same as rising: a black blazer can appear every period while its share barely moves.
In Stiloir
Insights (Analizler) computes each feature's share of analyzed items in seven dimensions, from product and fabric to color family. It compares the latest weeks with as many weeks before: at most four against four, as few as one against one while data is thin, so read early results with care. It skips weeks under 200 items and shows each feature's publication count. Fabric, print and silhouette tags come from article text; images give only garment colors.
Step 5: Micro trend, seasonal trend or lasting shift?
Numbers show what changed, not why. For each rising signal, ask how fast it rose, how many channels show it and what drives it.
| Type | Typical signs | Place in a collection |
|---|---|---|
| Micro trend | Rises fast, tied to one channel or event, often fades within weeks or months | An accessory or a capsule |
| Seasonal trend | On runways, in publications and in retail the same season | The season's color and product calls |
| Lasting shift | Repeats for several seasons, driven by changes in production or lifestyle | The core of your range |
Calling something lasting from one period's data is forecasting, not analysis. Without a long history, say so: “Spreading in recent weeks; too early to tell if it will last.”
Keep observation, interpretation and decision apart. For example, “striped shirts gained share across many publications” is an observation; “this may reflect a return to the office”, an interpretation; “let's add a striped shirt”, a decision.
Step 6: Turn findings into a trend board and a palette
For each trend you pick, make a small board with:
- Its name and a one-line observation.
- Evidence images from different sources, with links.
- A palette taken from the garments only.
- Fabric and print notes a supplier would understand.
- A note on whether it suits your customer.
See the fashion mood board guide for the board and the collection color palette guide for the palette. To turn a photo of your own sample into a flat sketch, see the photo to flat sketch guide.
Common mistakes in fashion trend research
- One source. A single magazine or feed shows one source's taste, not a trend.
- Coverage taken as sales. Attention in publications doesn't mean a piece will sell.
- The wrong season. Fall/winter coverage skews a spring/summer collection, and publish dates won't help: spring/summer shows are covered in the fall. Note the season each item covers.
A research calendar for independent designers
Trend research works best as a small, regular habit, not a big push before the season.
| When | What to do |
|---|---|
| Weekly | Scan new items; tag the ones that fit your research question. |
| Monthly | Compare the last four weeks with the four before; interpret the changes. |
| When planning a collection | Update your boards; move seasonal signals into your range, micro trends into capsules. |
| Graduation project | Open with your research question and sources; credit every evidence image. |
How fashion trend analysis works in Stiloir
You can run this method with a spreadsheet. Stiloir, a fashion trend analysis app, does the collecting, tagging and counting; interpretation and design decisions stay with you.
| Step | On your own | In Stiloir |
|---|---|---|
| Sources | Visit publications regularly | Discover collects new items hourly from 340+ feeds of 280+ publications, each linked to its original; most feeds refresh every 1–3 hours |
| Filtering | Sort items by hand | On the web: 22 product groups, gender and age, region, season, designer, color |
| Tagging | Fill in a spreadsheet | Fabric, print and silhouette from article text; colors from photos that show a garment |
| Comparing | Work out shares yourself | Insights compares the latest weeks with as many before, at most four against four |
| Following a topic | Search again regularly | Reports (Raporlar) searches registered fashion publications daily, weekly or on demand; 15 credits per run |
| Interpreting | Separate observation and decision | The assistant keeps observation, interpretation, any outlook and suggestions apart |
| Board | Collect images and colors | Add an image or its palette to a new or existing board |
Limits
Stiloir doesn't forecast sales or demand: it counts appearances in publications, with no sales, stock or order data. Any short outlook from the assistant rests on Stiloir's data and carries a confidence level. “Rising” means a larger share than the period before, not that it will sell. Collection began in September 2026, so there's no multi-season history yet. Colors come from on-screen images and aren't fabric-accurate. The assistant cites no sources.
See what uses credits on the Plans page, or browse all guides.
Frequently asked
What is the difference between trend analysis and trend forecasting?
Trend analysis looks at the present: it collects, counts and compares what repeats across publications, runways, stores and the street. Forecasting builds on it to anticipate the seasons ahead, usually with expert teams, archives and retail, search and social data. Stiloir analyzes publication coverage; it doesn't forecast sales or demand.
Where do fashion designers find trends?
Across several channels: runway archives, fashion magazines, trade publications and fairs, retailers' new-arrivals pages, street style and search tools. When a signal repeats in several of them in the same period, log it, compare its share across two equal periods and keep tracking it. One source isn't enough.
Are search trends enough to spot a fashion trend?
No. Search interest shows curiosity, not whether brands produce something or editors cover it, and it spikes around single events. Treat search data as one channel next to publications, runways and retail, and look for the same signal in several.
How do I know a trend isn't a one-off?
Look at speed, spread and cause. A signal that spikes and fades within weeks or months, tied to one channel or event, is usually a micro trend. One seen on runways, in publications and in retail the same season is seasonal. Lasting takes several seasons.
Can I do fashion trend analysis for free?
Yes. A spreadsheet, free publications and runway archives are enough; the cost is your time. In Stiloir, browsing Discover and Insights and reading news use no credits, and new accounts get 40 trial credits for 7 days. The Plans page lists what uses credits.
Does a rising trend mean it will sell?
No. A rising trend gets more coverage in publications, not necessarily more buyers; sales also depend on price, quality, channel and your customer. Stiloir uses no sales, stock or order data, so treat a rising signal as an input, not a decision.
How do I cite trend sources in a student project?
Under every evidence image or quote, give the publication, headline, date and link. Open with your research question and source list, say which periods you compared, and follow your school's citation style if it has one.