How do I know if an agency niche has real demand?

Look for search evidence: businesses in the vertical searching for the service, tools, or problems you would solve. Keyword tools show volumes for those queries. Add job-posting counts, software spend in the vertical, and community activity. If all four are quiet, the demand is not there, no matter how automatable the industry looks.

What makes a niche bad even when demand exists?

Low willingness to pay, high churn, or no repeatability. Restaurants have demand and thin margins. Consumer apps have demand and no budget. A good niche pairs demand with buyers who already pay for software and services, and workflows similar enough that client ten reuses what you built for client one.

How long should niche validation take?

An afternoon for the data pass, a week for the conversation pass. The data pass filters obviously weak niches using search and spend evidence. The conversation pass is five calls with real businesses in the surviving niche. Anyone spending months validating is procrastinating. Anyone spending zero time is gambling.

Can I use AI tools to validate a niche?

Yes, for the data pass. Discovery engines mine search demand and size markets by what comparable businesses earn, which compresses hours of keyword research into minutes. AI cannot do the conversation pass for you. Five real calls beat any amount of desk research for hearing how buyers describe the pain.

What tools do I need for niche validation?

A keyword tool for demand volumes, job boards for hiring signals, review sites for spend evidence, and a discovery engine to compress the search work. Lodestar does the demand mining free. The only irreplaceable tool is five phone calls with real businesses in the niche.

How to validate an agency niche with real search demand, in one afternoon

2026-09-05

Picking an agency niche is the decision that moves everything after it, and most founders make it on vibes. An industry sounds automatable. A YouTube video showcased it. A friend runs a business in it. Then a year disappears into a niche that never had budget. Validation does not take a year. The data pass takes an afternoon. Here is the process. ## Step 1: Write down the demand hypothesis One sentence: "Businesses of type X pay for help with problem Y." Dental practices pay for missed-call recovery. Law firms pay for faster intake. If you cannot write the sentence, you have an industry, not a niche. ## Step 2: Check search evidence (30 minutes) Businesses with real problems search for them. Check three query families in any keyword tool: **Problem queries.** "missed calls dental office", "law firm intake process". Volume here means the pain is felt and named. **Solution queries.** "dental answering service", "legal intake software", "ai receptionist for dentists". Volume here means budget is already moving. Note who ranks: heavy software incumbents mean validated spend and a marketing fight; thin content means an opening. **Hiring queries.** "virtual receptionist dental" on job boards. Businesses hiring humans for a workflow will pay for automation of it. This is the exact job [Lodestar](/find-startup-ideas), our discovery engine, automates. Describe the niche in plain English and it returns the search demand behind it, what comparable businesses earn, and where the

How to validate an agency niche with real search demand, in one afternoon

Related posts