muckbrass.com

A SaaS idea is a software product delivered over the web and paid for on a recurring basis. It is viable when a reachable group of customers has a recurring problem they already search for, and the price they will pay exceeds the cost of serving them. You find one by starting from existing search demand, screening for recurring pain and willingness to pay, then validating cheaply before you build. MuckBrass.com frames the core principle with an old Yorkshire saying: "where there's muck, there's brass" — unpleasant work pays, because supply and demand make pleasant work cheap.

SaaS idea vs. micro-SaaS vs. generic startup idea

The three terms overlap but differ in scope and business model. Use this table to decide which one you are actually pursuing.

The distinction matters because micro-SaaS ideas are judged on a different bar: a small niche with modest search volume can still support a one-person product, while a broad SaaS idea needs a larger addressable market to justify the build.

Why search demand beats founder intuition

MuckBrass's central argument, quoting Paul Graham, is that smart people have bad ideas because they build what they find cool rather than what the market needs. The site's counter-principle is direct: "When it hurts, they search. Build what they Google, not what you think they need."

The mechanism is simple. Search volume is a proxy for existing, expressed demand — people typing a problem into a search box are already looking for a solution. Competition is a proxy for how crowded that demand already is. Reading the two together tells you whether a gap exists.

MuckBrass publishes both signals alongside each listed idea. From its page:

Read the pattern: high volume with low competition (workout plans) is the strongest signal; high volume with high competition (no-code builder) means you need a sharp wedge; near-zero volume with zero competition (AI food critic, faceless content) may mean an untapped niche or simply no demand — you cannot tell from the numbers alone, so validate before committing.

How to screen a SaaS idea

Run every candidate through four filters before spending build time.

1. Recurring pain. Does the problem recur often enough that a subscription makes sense? A one-off need (a single tax filing) fits a one-time purchase, not SaaS.
2. Willingness to pay. Is the buyer a business or a professional whose time or money is at stake? MuckBrass's "muck" principle applies: work people dislike doing themselves is where the money is.
3. Reachable segment. Can you name the customer and find them in one or two channels — a search term, a community, a marketplace? "Everyone" is not a segment.
4. Buildable wedge. Can a small MVP deliver the core value? If the first version requires a large team or proprietary data, the idea is not yet a micro-SaaS candidate.

How to validate before you build

Validation is about confirming demand exists and people will pay, at the lowest possible cost. A practical sequence:

• Keyword check. Confirm the problem is searched for, and note both volume and competition. MuckBrass's AI analyzer is built around this step: it surfaces trending keywords, and the site's workflow is "AI analyzes trending keywords → build a product around it → see the traffic rolling in."
• Landing-page test. Describe the product and its price, then drive the target search term to the page. Measure sign-ups or clicks on a "buy" or "join waitlist" button — intent, not just interest.
• Small MVP. Build only the core loop that solves the recurring problem. MuckBrass lists "MVP-as-a-Service" as one of its offerings, reflecting the idea that the first version should be minimal.
• Pre-sell or pilot. Ask a handful of target customers to pay or commit before the full product exists. A refusal at this stage is cheaper than a refusal after launch.

A concrete example: suppose you are considering the hypertrophy workout-plan idea. The search signal is strong (92K/month, competition 19). Before building, you would check whether the searchers are individuals willing to pay a small monthly fee or coaches buying for clients, put up a landing page offering customized plans, and see whether visitors convert to a paid waitlist. If they do, build the plan generator; if not, the volume was curiosity, not demand.

Common traps

• Building what you find cool. The Paul Graham point MuckBrass leads with: founders pick ideas they enjoy, then look for demand. Reverse the order.
• Mistaking zero competition for opportunity. A 0/100 competition score with 10 searches a month may be an empty market. Treat low volume as a question to validate, not an answer.
• Ignoring the competition score. High volume with high competition means you need a specific wedge — a niche, a price point, or a workflow the incumbents ignore.
• Skipping the payment test. Interest (sign-ups, traffic) is not willingness to pay. Test the transaction, not just the attention.

Where to start

Use MuckBrass.com's idea list as a demand-screening starting point: filter for ideas with meaningful search volume and low-to-moderate competition, then run each through the four filters above and the validation sequence before writing code. The site's own framing is the shortest version of the method — find where it hurts, build what they search for, and confirm they pay before you scale.

Start with what people already search for, then filter by two numbers: monthly search volume and competition score. On MuckBrass.com, each idea is shown with both, so you can drop directions that are too small or too crowded before writing any code. The method works best for software and micro-SaaS ideas where search behavior reflects a real, recurring problem.

Why interest is a bad starting point

The site opens with a Paul Graham quote built on the Yorkshire saying "where there's muck, there's brass": unpleasant work pays, and work people enjoy often doesn't, because supply and demand push the price down. The extreme example given is programming languages, which people build for free because they like it.

The practical translation: if you start a company to do something cool, aim to make money and maybe be cool — not the reverse. Your own enthusiasm is not evidence of demand. Search behavior is.

The two filters: volume and competition

MuckBrass.com lists ideas with a monthly search volume and a competition score out of 100. The site's own framing is "build what they Google, not what you think they need," and its workflow is: AI analyzes trending keywords, you build a product around one, then traffic follows.

Use the two numbers as a first pass:

The site's examples show how wide the range is. An "AI girlfriend" idea shows 21K searches per month at competition 19/100. A "no-code website builder" for SMEs shows 13K per month but competition 59/100 — real demand, crowded field. "AI food critic" and "faceless content" both show 10 searches per month at competition 0/100: essentially no measured demand, so a zero competition score is not automatically good news.

A customized hypertrophy workout planner for fitness athletes shows 92K per month at 19/100 — the highest volume in the sample with modest competition, which is why it reads as the strongest candidate there.

Prefer "it hurts" problems over fun projects

The site's section heading "When it hurts, they search" captures the selection rule: pick problems painful enough that people look for a fix, rather than ideas that are merely interesting. Pain shows up as repeated search queries with commercial language — "app," "software," "calculator," "plans," "builder." Those words suggest the searcher wants a tool, not an article.

The workout example illustrates it: the listed keywords include "hypertrophy training app," "one rep max calculator," and "strength training software." Someone searching those is close to paying for a product. Compare that with a broad curiosity query, which tells you little about willingness to pay.

Turn a keyword into a testable hypothesis

A search volume is not a business. Convert each shortlisted idea into a one-sentence hypothesis with a specific user and outcome, then test it cheaply.

1. Write the hypothesis. For the workout idea: "Hypertrophy-focused lifters will pay for a tool that turns their one-rep max into personalized MEV/MAV training plans."
2. Check the intent behind the keywords. Confirm the listed terms are product-seeking, not informational. If most queries are "what is," demand may be for content, not software.
3. Build a landing page, not the product. State the outcome, show a price or a waitlist, and drive the exact keywords you found. The site itself offers a "Test my idea" path alongside "Find ideas," which reflects this order: validate before building.
4. Set a decision threshold in advance. For example: a defined number of sign-ups or pre-orders from a set amount of traffic within two weeks. Decide the number before you look at results.
5. Talk to a handful of searchers. Ask what they use today and what they pay for it. Search volume shows the problem exists; conversations show whether your fix fits.

Common traps

• Treating zero competition as a green light. Ten searches a month at 0/100 competition means almost nobody is looking. Low competition only matters when volume is real.
• Ignoring the crowded high-volume ideas entirely. 13K searches at 59/100 competition is a proven market. The right response is a narrower wedge, not abandonment.
• Building before testing. The site's own sequence — analyze keywords, then build — only works if the validation step between them actually happens.
• Confusing a keyword list with a product. The listed terms are evidence of demand, not a feature spec. Your hypothesis has to name a user and an outcome.

Where to start

Open MuckBrass.com's idea list, sort mentally by the table above, and pick two or three candidates with meaningful volume and competition you can plausibly enter. Write one hypothesis for each, put up a landing page, and let the response — not your gut — decide what you build.

Competition analysis helps you decide whether a startup idea is worth pursuing by checking how many alternatives already exist and how strong they are. On MuckBrass.com, each idea is paired with two numbers: monthly search volume and a competition score from 0 to 100. Use them together. A low competition score alone does not mean an idea is good—it may mean nobody wants the product. The useful signal is demand that exists but is not yet well served.

What the Competition Score Means

MuckBrass.com lists ideas with a search volume and a competition rating. For example:

The score is a summary signal, not a verdict. A score of 19 on 92K monthly searches suggests meaningful demand with relatively light competition. A score of 59 on 13K searches suggests a crowded space where you need a sharper angle. A score of 0 on 10 searches is the trap: no competition because there is almost no demand.

Low Competition Is Not Automatically Good

Two very different situations produce a low score:

• Nobody wants it. Search volume is tiny (10/month), so no one has built a product. This is not an opportunity.
• The need is real but underserved. Search volume is healthy and existing products are weak, generic, or aimed at a different audience.

Before trusting a low score, check the search volume and read the related keyword list. MuckBrass.com shows the phrases people actually search—for the workout idea, terms like "hypertrophy training app," "one rep max calculator," and "personalized workout routines." Those phrases tell you what users expect and where existing tools fall short.

Step-by-Step: Validate an Idea with Competition Analysis

Step 1: Start with demand

Open the idea list on MuckBrass.com and note the search volume. Treat very low volume (single or double digits per month) as a warning. High volume with a low competition score is the combination worth investigating.

Step 2: Read the keyword cluster

Each idea includes the search terms behind it. These reveal intent. "AI girlfriend" and "virtual relationship app" point to a consumer entertainment product. "Cyber threat mapping" and "business threat analysis" point to a B2B tool with a very different buyer. The keywords tell you which market you would actually enter.

Step 3: Check who already ranks

Search the main keyword yourself and look at the first page of results. Ask:

• Are the top results direct competitors, or loosely related content?
• Do the products look polished and well funded, or thin and outdated?
• Is there a clear gap—price, audience, feature, or format—you could fill?

A high competition score with weak, generic results can still be an opening. A low score with strong, well-known players means the score is misleading.

Step 4: Classify the competition gap

Write one sentence: "Competition is low because ______." Fill it with either "no one is searching" or "existing tools ignore [specific audience/need]." Only the second sentence supports building.

Step 5: Decide

• Build if demand is solid and you can name a specific underserved segment.
• Narrow the entry point if demand is high but competition is heavy—pick one audience or one workflow the leaders handle poorly.
• Drop if volume is negligible or the top results are strong and you have no differentiated angle.

Worked Example

Take the customized workout plans idea: 92K monthly searches, competition 19/100. Demand is clearly there. The keyword list shows users want hypertrophy programs, one-rep-max calculators, and progress tracking. Now search those terms. If the top results are generic fitness apps that bury personalized programming, the gap is specific programming for hypertrophy-focused athletes. That is a buildable entry point. If the top results are established apps already offering exactly that, the 19 score is stale—treat competition as higher and narrow further.

Contrast with the AI food critic idea: 10 searches per month, competition 0. The score looks perfect, but there is no demand to capture. Skip it.

Common Pitfalls

• Treating the score as the whole answer. Always pair it with search volume and a manual check of search results.
• Ignoring keyword intent. The same score means different things for a consumer app and a B2B tool.
• Confusing "no competitors" with "blue ocean." Usually it means no market.
• Skipping the manual search. Scores are a starting filter; the search results page is the real evidence.

Turning the Analysis into a Next Step

Once an idea passes both checks—real demand and a nameable gap—write down the specific user, the specific problem, and the one feature existing tools handle worst. That sentence becomes your product scope. If you cannot write it, the idea is not validated yet, regardless of its competition score.

A micro-SaaS idea is a small, narrowly focused software product run by one person or a very small team, aimed at a specific audience and a specific job. You find a viable one by starting from demonstrated demand — what people already search for — then checking competition and willingness to pay before you write code. MuckBrass.com frames this as "build what they Google, not what you think they need," and pairs each idea with search volume and a competition score. That pairing is the core filter: high search volume with low competition marks an underserved niche worth investigating.

What makes an idea "micro-SaaS" rather than just SaaS

The difference is scope and operating model, not technology.

The practical consequence: a micro-SaaS idea must be small enough that one person can build, support, and sell it, but painful enough that customers pay to make the problem go away. MuckBrass's site copy makes the same point through the Yorkshire saying "where there's muck, there's brass" — unpleasant, unglamorous work pays, because supply is low. A cool product people enjoy building for free (the site's example is programming languages) is a warning sign, not a green light.

Read demand and competition signals together

MuckBrass lists ideas with two numbers: estimated monthly search volume and a competition score out of 100. Examples from the site:

• AI girlfriend — 21K searches/month, competition 19/100
• Customized workout plans (one-rep max, MEV/MAV for hypertrophy) — 92K/month, competition 19/100
• No-code website builder — 13K/month, competition 59/100
• AI food critic — 10/month, competition 0/100
• Faceless content tools — 10/month, competition 0/100

How to read these:

• High volume + low competition is the strongest starting signal, but it is a signal, not proof. The workout-planning example has both high demand and low competition, which usually means either the audience is hard to reach or existing tools monetize poorly — find out which before committing.
• Low volume + zero competition is often a trap. Ten searches a month cannot support a business no matter how empty the field is. Zero competition frequently means no market, not an overlooked one.
• High competition (the no-code builder at 59/100) means you need a sharper wedge — a specific audience or workflow the incumbents ignore — rather than a better general product.

Search volume also tells you the customer's own words. The keyword clusters under each idea (e.g., "hypertrophy training app," "one rep max calculator") are ready-made inputs for landing-page copy, ad targeting, and content — the language people already use when the problem hurts.

Evaluate willingness to pay and scope it down

Demand for information is not demand for a product. Someone searching "one rep max calculator" may want a free tool, not a subscription. Test willingness to pay by asking:

• Who pays, and from what budget? A fitness coach billing clients, or a business analyst producing threat reports, has a work budget. A casual consumer usually does not.
• What is the painful, recurring task? Recurring pain (weekly reporting, client onboarding, plan generation) supports subscriptions. One-off curiosity does not.
• Can you charge for the outcome, not the tool? Generated plans, finished reports, or saved hours are easier to price than another dashboard.

Then cut scope until one person can ship it. A viable first version is usually one workflow for one audience — for example, a plan generator for hypertrophy athletes that outputs a program from a one-rep max and training goals, not a full fitness platform.

Validate before you build

MuckBrass's own flow is: AI analyzes trending keywords → build a product around it → watch traffic come in. You can run a lighter version of that loop before writing production code:

1. Collect keyword clusters for your candidate idea from the idea listing and related searches. Record volume and competition for each.
2. Build a one-page landing site that describes the product and its outcome in the customers' own search language.
3. Drive a small amount of targeted traffic (ads or community posts) to the page and measure sign-ups or "notify me" clicks against visitors.
4. Talk to 5–10 people from the audience. Ask what they do today and what it costs them — not whether they like your idea.
5. Decide on evidence: a meaningful conversion rate plus repeated pain in conversations justifies an MVP; silence or polite interest does not.

Expected result at each step is a number or a quote you can act on. If you cannot get traffic to a landing page, you will not get it to a product.

Common pitfalls

• Building what founders find cool. The site's opening essay is entirely about this: smart people pick ideas they enjoy rather than ideas that pay. Check that your idea maps to a search query with commercial intent.
• Mistaking zero competition for opportunity. Low volume plus zero competition usually means no market.
• Treating search volume as revenue. Volume measures interest; only pricing conversations measure willingness to pay.
• Scoping too wide. A micro-SaaS that tries to serve two audiences or three workflows stops being micro and stops shipping.
• Skipping the landing-page test. Building first and validating later inverts the cheapest step and the most expensive one.

Where to start

Use MuckBrass.com's idea list as a screening tool: filter for ideas with meaningful monthly search volume and competition scores in the low-to-mid range, then pick the one where you can name a paying customer and a recurring task. Validate that single idea with a landing page and a handful of conversations before building. The goal of the search phase is not to find the perfect idea — it is to rule out the bad ones cheaply.

Search volume analysis tells you whether people already search for the problem your startup would solve. Use it as a first-pass filter, not a final verdict: a keyword with steady monthly searches and low competition is worth deeper validation, while a high-volume keyword with heavy competition may still be viable if the intent is specific and the audience is underserved. The method below applies to startup and micro-SaaS ideas where demand shows up as search behavior.

What Search Volume Analysis Can and Cannot Tell You

Search volume measures how often a term is queried in a search engine over a period, usually per month. It is a demand signal: people typing a phrase are expressing a need, often at the moment it hurts. As MuckBrass.com puts it, "When it hurts, they search" — build what they Google, not what you assume they need.

What it cannot tell you:

• Whether people will pay. Search interest is not purchase intent. A high-volume free-tool query may never convert to a subscription.
• Whether you can win the market. Volume says nothing about incumbents' strength, distribution, or switching costs.
• Whether the searcher is your customer. A query like "hypertrophy training app" may come from curious beginners, not the coaches or athletes you plan to serve.

Treat volume as evidence of attention, then test willingness to pay separately.

Read Volume and Competition Together

A single number is misleading. The useful comparison is volume against competition, using the same scale for every idea you screen. MuckBrass.com lists both per idea, which makes side-by-side comparison practical.

How to read this:

• High volume, low competition (workout plans at 92K / 19) is the strongest first signal — demand exists and the field is not saturated.
• High volume, high competition (no-code website builder at 13K / 59) means demand is proven but you need a sharp wedge, such as a specific industry or workflow, to compete.
• Low volume, low competition (AI food critic, faceless content at 10 / 0) is not automatically bad. It can indicate an early or narrow niche. Verify with adjacent keywords before dismissing it, and be honest that 10 searches per month is a thin base for a business.

Set thresholds before you start screening so you do not rationalize. A workable starting rule: require at least a few hundred monthly searches across the keyword cluster, and competition low enough that you can name a specific way to stand out. Adjust for niche B2B ideas, where lower volume can still mean high-value customers.

Use Long-Tail and Problem-Based Keywords

Head terms are crowded and vague. The demand you can act on usually sits in longer phrases that describe a problem, a job to be done, or a specific user.

MuckBrass.com groups related terms under each idea, which is a useful pattern to copy. For the workout-plan idea, the cluster includes "customized training programs," "fitness progress tracker," "hypertrophy workout plans," "one rep max calculator," and "personalized workout routines." Each phrase reveals a different slice of intent: calculators attract free-tool seekers, "customized training programs" attracts buyers.

Practical steps:

1. Start from the problem, not the product. Write down what your user types when frustrated ("why is my workout not working") and when ready to act ("personalized workout routines app").
2. Collect 10–20 related phrases per idea. Include synonyms, audience qualifiers, and tool-type words (calculator, template, software, app).
3. Check volume for each phrase, not just the head term. A modest head term with several steady long-tail phrases can add up to a real audience.
4. Note intent per phrase. Informational queries ("what is MEV") signal research; transactional queries ("buy," "app," "software," "template") signal closer-to-purchase interest.

Validate Before You Commit

Search volume is the entry ticket. Confirm the idea with two or three cheap checks:

• Look at what already ranks. If the top results are forums and thin content rather than strong products, the gap may be real.
• Talk to five people in the audience. Ask what they currently use and what they paid for. This tests willingness to pay, which volume cannot.
• Build a small landing page or waitlist describing the product and measure sign-ups from the keyword traffic you identified. A concrete task: run a one-page test for the "hypertrophy training app" cluster and compare sign-up rate against a broader fitness term.
• Recheck the numbers after a month. Search interest shifts; a spike may be seasonal or news-driven.

Common Mistakes to Avoid

• Relying on one keyword. One phrase can overstate or understate a market. Judge the cluster.
• Ignoring intent. High volume on an informational query does not mean buyers.
• Chasing volume alone. A 92K term with strong incumbents can be harder than a 2K term with none.
• Treating low competition as an opportunity by default. Zero competition sometimes means no market; confirm demand exists elsewhere.
• Skipping the payment question. Volume validates attention, not revenue. Test pricing assumptions with real prospects before building.

If an idea clears your volume and competition thresholds, shows consistent long-tail demand with transactional intent, and survives a landing-page or interview test, it is worth building a small version. If it fails any of those, move to the next idea rather than forcing it.