keyword-research·6 min read·
Read keyword data like a builder: free UK tools to size demand (2026)
Most keyword-tool roundups are written for marketers chasing rankings. This one is for builders deciding whether to write code at all. Search demand is the cheapest, fastest demand signal there is - free, readable in an hour, before you commit a weekend. Here is the free UK stack, how to read volume and intent like a build/no-build decision, the winnable-battle heuristic, and a small Claude Code script that aggregates it all into one CSV.
Most articles about keyword tools are written for marketers. Their question is "how do I rank for traffic I can monetise?" Yours is different. Your question is "should I write this code at all?"
Search demand is the cheapest, fastest, most honest demand signal you can read. It is people typing their problem into Google with their own fingers - no survey bias, no founder optimism, no friend being nice to you. And you can read it in an hour, for free, before you commit a single weekend to building. This post teaches you to read it like a builder: not chasing traffic, but making a build / no-build / niche-down decision.
This pairs with the weekend validation harness - keyword data is the demand leg, the smoke-test landing page is the willingness-to-pay leg. Read demand first; it is free and it filters out the ideas not worth a Saturday.
The free UK stack
Four tools, layered. None is complete on its own. Together they give you a picture good enough to decide on.
1. Google Keyword Planner - first-party volume, straight from Google. The single most reliable free source for absolute monthly search volume. You need a Google Ads account to access it, but you do not need to run a single ad. This is your "how many people actually search this" number.
2. Google Trends - relative popularity over time. Keyword Planner tells you how many; Trends tells you whether the line is going up, down, or sideways, and when it spikes. Crucial for catching seasonality (some UK ideas are January-only) and for spotting a rising topic before it is obvious. Filter to United Kingdom and you get regional breakdowns too.
3. AnswerThePublic - the questions people actually ask. Enter a topic and it returns the who/what/why/how queries clustered around it. Browse from the UK and it auto-tailors to UK search patterns. This shows you the shape of the demand - the specific sub-problems people are stuck on, which is where product wedges hide.
4. Google autocomplete - the fastest long-tail miner there is, and completely free. Start typing your problem into Google and read the suggestions. Each is a real query people type. Add a letter and the suggestions branch. Ten minutes of this surfaces long-tail terms no tool charges you to find.
Reading volume like a builder
Marketers want big volume. Builders want the right volume - big enough to matter, small enough that incumbents ignore it.
| Monthly UK searches (core term) | Builder read |
|---|---|
| Under 100 | Thin - viable only with strong commercial intent and a niche you can own whole |
| 100 - 1,000 | The indie sweet spot - matters to you, usually beneath the radar of funded players |
| 1,000 - 5,000 | Real demand, expect moderate competition |
| Over 5,000 | Strong demand, expect serious competition - you will need a sharp wedge |
The sweet spot for UK micro-SaaS is that 100 to 1,000 band. It is the range where a focused indie tool can rank, convert, and grow without waking a competitor with a marketing budget. As the first 100 organic visitors playbook covers, those visitors are easier to win on a low-competition long-tail than on a high-volume head term.
Volume is not the whole story - read intent and competition too
A common mistake: chasing the biggest number. A 5,000-search informational query ("what is X") with ten funded competitors is a worse target than a 200-search commercial query ("X tool for UK Y") with two thin indie results. Read three things together:
- Volume - is anyone searching?
- Intent - are they trying to do or buy something (commercial), or just learn (informational)? Commercial intent converts; informational intent feeds a blog.
- Competition - who already ranks, and how hard would they be to displace?
The winnable-battle heuristic
Search your core keyword and look at the top three results with a builder's eye. The signal you want: indie tools, forum threads, or thin content with few backlinks - not established players with deep domain authority. If the first page is funded incumbents and big publishers, ranking is a long, expensive grind. If it is two small tools and a Reddit thread, that is a gap you can take.
The heuristic that has held for years: if the top three results are indie tools with fewer than about 1,000 backlinks each, it is a winnable SEO battle. Pair a winnable SERP with a real volume and you have a target worth a weekend.
Automate the demand snapshot with Claude Code
Doing this by hand for one idea is an hour. Doing it for a shortlist of ten is a slog - exactly the kind of repetitive fetch-and-tabulate job to hand to a small script. Two of the four signals have lightweight unofficial endpoints: Google autocomplete and Google Trends interest-over-time. Have Claude Code write a fetcher:
# demand_snapshot.py - aggregate free demand signals into one CSV
import csv, requests
from urllib.parse import quote
def autocomplete(seed):
# Google's public suggest endpoint
r = requests.get(
"https://suggestqueries.google.com/complete/search",
params={"client": "firefox", "q": seed, "gl": "uk", "hl": "en-GB"},
timeout=10,
)
return r.json()[1] if r.ok else []
def expand(seeds):
rows = []
for seed in seeds:
for term in autocomplete(seed):
rows.append({"seed": seed, "suggestion": term})
return rows
if __name__ == "__main__":
seeds = ["uk freelancer invoice", "uk landlord compliance", "uk vat return"]
rows = expand(seeds)
with open("demand_snapshot.csv", "w", newline="", encoding="utf-8") as f:
w = csv.DictWriter(f, fieldnames=["seed", "suggestion"])
w.writeheader()
w.writerows(rows)
print(f"wrote {len(rows)} suggestions")
Ask Claude Code to extend it with a Google Trends pull (the pytrends library wraps the interest-over-time endpoint) so each surviving term gets a 12-month trend direction alongside it. Keyword Planner volume still wants a manual export - Google guards that API - so the pragmatic split is: automate autocomplete and Trends for breadth and direction, then hand-enter Planner volume for the handful of terms that survive the first cut. The whole script is under 100 lines and costs pennies in Claude tokens to write. Run it weekly against your idea shortlist and you have a living demand board.
Putting it together: a build/no-build read in an hour
For any idea:
- Autocomplete + AnswerThePublic (10 min) - what do people actually search, and what sub-problems cluster around it?
- Keyword Planner (10 min) - absolute volume on the three best terms. Is it in the 100-1,000 sweet spot or above?
- Trends (5 min) - is the line rising, flat, or dying? Seasonal?
- SERP eyeball (10 min) - top three results: indie tools and forums (winnable) or funded incumbents (grind)?
- Decide: real volume + commercial intent + winnable SERP = build it. Thin volume or a wall of incumbents = niche down or move on.
That is an hour against a weekend. Search demand will not validate willingness to pay - that is what the smoke-test landing page and pre-sale are for. But it is the cheapest filter you have, and it runs before you write a line of code.
New here? IdeaStack publishes one deeply researched UK business opportunity every Thursday - real keyword data, competitor analysis, builder prompts. See the latest free report.
Frequently asked
Why should a builder care about keyword data - is that not a marketing thing?
It is the cheapest demand signal you can read, and you can read it in an hour for free before committing a single weekend to code. Most keyword roundups are written for marketers who want to rank for traffic they then monetise with ads. A builder uses the same data for a different question: does anyone actually search for the problem I want to solve, and is the gap winnable? Search volume is people typing their problem into Google with their own fingers - that is more honest than a survey and more reliable than your gut. It will not tell you everything (some great products serve demand that has no search term yet) but for most UK micro-SaaS the search signal is the fastest way to separate a real itch from a founder fantasy.
What are the best free keyword tools for UK builders in 2026?
Four, layered. Google Keyword Planner gives first-party volume data straight from Google - the most reliable free source for absolute search volume, and free with any Google Ads account (you do not need to run ads). Google Trends shows relative popularity over time, so you catch seasonality and whether a topic is rising or dying, with UK-specific and regional breakdowns. AnswerThePublic surfaces the actual questions people ask around a topic and auto-tailors to UK search patterns when you browse from the UK - gold for understanding the shape of the demand. And Google autocomplete is the fastest long-tail miner there is: start typing your problem and read what Google suggests. No single tool is complete; the power is in layering them.
What search volume is big enough to justify building?
There is no universal floor, but useful bands for a UK micro-SaaS: under 100 monthly UK searches for the core term is thin - viable only if intent is very commercial and you can win the whole niche; 100 to 1,000 is the sweet spot for indie SaaS, big enough to matter and usually small enough that incumbents ignore it; over 5,000 means real demand but expect serious competition. Volume alone is not the decision though - a 200-search term with strong buying intent and weak competition beats a 5,000-search informational term every time. Read volume, intent, and competition together, not volume in isolation.
How do I judge whether the SEO competition is winnable?
Search your core keyword and look at the top three Google results with a builder's eye. The winnable signal: the top results are indie tools, forum threads, or thin content with few backlinks, rather than established players with deep domain authority. If the first page is dominated by funded incumbents and big publishers, ranking will be a long, expensive grind. If it is a couple of small tools and a Reddit thread, that is a gap you can take with good content and a focused product. The heuristic that has held: if the top three results are indie tools with fewer than about 1,000 backlinks each, it is a winnable SEO battle. Pair that with a real search volume and you have a target.
Can I automate keyword demand-sizing with Claude Code?
Yes, and it is a small script. The free signals - Google autocomplete suggestions and Google Trends interest-over-time - both have lightweight unofficial endpoints, and Claude Code can write a fetcher that pulls autocomplete expansions for a seed term, grabs the Trends interest curve, and writes the lot to a single CSV demand snapshot you can re-run weekly. Keyword Planner volume still wants a manual export (Google guards that API), so the pragmatic build is: automate autocomplete plus Trends for breadth and trend direction, then hand-enter Planner volume for the handful of terms that survive. The whole script is well under 100 lines and costs pennies in Claude tokens to write.
