Guide

Trend research on a freelance budget

Where the free signals are, how to turn them into a system, and the three situations where paying for a subscription is actually the right call.

By YakLab · Published 26 August 2026 · 8 min read

What a subscription actually buys

It is worth being precise about this, because “I can’t afford the tools” is usually the wrong diagnosis.

A forecasting subscription is not selling you access to secret information. Almost every input it uses is publicly observable: what walked, what is hanging in shops, what people are searching for and reselling, what mills are showing, what the colour authorities issued. You can see all of it.

What you are buying is three things: synthesis — somebody has already done the reading; a shared vocabulary — when a client says a season name and you know it, you sound like a peer; and cover — being able to cite a recognised source when a call is questioned.

The first is time you can spend instead of money. The second you can acquire by reading. The third only matters with certain clients, and is worth less than it looks, because a forecaster who can explain their own reasoning beats one who can only cite someone else’s.

Where the free signal is

Pull from several kinds of source, because each is biased differently and the corrections come from the overlap.

Collections and runway

Fashion press coverage of shows is extensive and free. Read it for direction and silhouette, not for product. Note who is doing something rather than only what — three houses with nothing in common arriving at the same proportion is a real signal; one house doing something striking is a look.

Retail assortment, read properly

Every major retailer publishes its entire range, with prices, sorted by newness and often by popularity. This is extraordinary competitive data given away free. Track a fixed set of retailers at the price points your client sits in, look at the same categories at the same interval, and pay attention to depth — how many colourways and sizes a style comes in tells you how much someone believes in it.

Resale and secondhand

Resale marketplaces show demand without a marketing budget behind it. Sold listings, asking prices holding above retail, and what people search for are all visible. Especially good for spotting revivals early.

Materials and colour

Mills and yarn spinners publish development stories. Trade fairs publish exhibitor lists and themes. Colour authorities publish seasonal direction. Because materials run earliest in the calendar, this is where you see things first.

Adjacent categories

Interiors, product design, packaging, film and television art direction, food photography. Colour and surface migrate between categories and usually arrive in fashion rather than leaving it. This is the cheapest genuine edge available, because most fashion people only look at fashion.

Archives and museums

Major museum collections are digitised and searchable. Revivals are constant, and being able to point at the original rather than the last revival makes a story far more convincing.

The most under-used source is a shop

Spend two hours in physical retail with a notebook and you will learn things no dataset contains. What has been marked down and how hard. What is on the front table versus buried at the back. What the fabric actually feels like at that price. Which sizes have gone — a rail stripped of the middle sizes has sold; a rail stripped of the extremes has not.

It is free, nobody does it consistently, and it is the fastest correction available for research built entirely on screens.

Build a system, not a folder

The difference between someone with a subscription and someone without is rarely access. It is that one of them has a system and the other has a folder of screenshots.

The minimum viable system:

• A fixed source list. Ten to fifteen sources you check on a set rhythm. Fixed matters more than large — you are looking for change over time, which requires looking at the same things repeatedly.
• A capture habit with provenance. Every saved image gets a source and a date. Without those two fields the image is decoration.
• A monthly cut. An hour a month reviewing what you saved, deleting most of it, and writing two or three sentences about what is moving. This is the step that turns collecting into forecasting.
• A running list of dead trends. Things you thought were rising that weren’t. Nobody keeps this and it is the fastest way to get better.

Twenty minutes a week on a fixed source list, plus an hour a month cutting, beats a subscription used sporadically.

Budgeting the time, not the money

The real constraint for a freelancer is unbilled hours. So price it honestly: for one story with a defined customer and category, budget roughly a focused day of gathering and half a day of cutting and writing. Gathering expands to fill whatever time is available, so set the cut-off before you start.

If research is for a client project, put it in the quote as a line item. Trend research presented as a deliverable gets paid for; trend research presented as preparation gets absorbed. That is a positioning choice, not a market condition.

When paying is the right call

Three situations where spending money is straightforwardly correct:

• When the client expects a named source. Some organisations will not act on research without one. Price it into the job rather than absorbing it.
• When you are working outside your knowledge. Taking a kidswear brief with a womenswear background, you are not short of images — you are short of context. Buy the context.
• When the deadline is shorter than the reading. Sometimes you need the synthesis today, and that is exactly what synthesis is for.

Outside those, the honest answer is that the constraint is usually time and structure rather than access — and structure is free.

---

The gap between a freelancer’s research and an agency’s is not the sources. It is that the agency does it on a rhythm, writes it down, and keeps a record of when it was wrong. All three are available to anyone willing to be boring about it.