Always build what people actually want.
Sounds obvious but so so so (so) many entrepreneurs ignore this.
And we instead build what we want. What we think would be cool.
Don’t do this!
We’re going to make our lives a lot easier by finding existing problems in our market. Then we’re going to build our micro tools to solve these existing problems.
No guess work, no reliance on our (probably wrong) ideas.
Let’s get started:
Here's the thing about building micro tools - the hard part is no longer the building. With modern AI and no-code tools, that's become surprisingly straightforward.
The hard part is knowing what to build. Knowing what people will actually want.
Here’s the shortcut. Our future users (and customers) are already telling you exactly what they want. They're typing it into Google 20,000+ times a month like our business name example from the last Part. Your job is simply to listen.
We are going to use Google search volume as a proxy for demand.
Basically, how many people each month are looking for the solution to certain problems. We find their problems. And provide solutions by building micro tools.
But Kyle, haven’t you heard, Google is dead? People won’t use Google to answer questions anymore now that we have AI!!
Don’t worry - plenty of time. Google (w/ Youtube) have 50% of web traffic locked down still:
First up, we need real numbers. ChatGPT might be clever, but it can't tell you actual search volumes. It doesn’t have access to Google’s keyword volume database. It will just make stuff up. That won’t work!
So, we have a few options:
The free route? Google Ads Keyword Planner. You'll need to set up an account, but you don't need to spend anything. It'll give you search volumes and competition levels for any term you plug in. This gives you direct access to keyword volumes straight from the horses’ mouth.
If you're serious about this, paid tools like Ahrefs or SEMrush will make your life easier. They draw on the direct Google data but add layers of information to tell which keywords are particularly attractive. You can work this out yourself using the free Google tool but it’s a bit more work. These premium tools will show you everything from search volumes to keyword difficulty to what's currently ranking in a couple of clicks.
Finally, if you want the premium data but don't want to pay for tools? Here's a hack: hire someone on Fiverr or Upwork who has access to these tools. For about $20-30, they can pull all the keyword data you need. They’ll just run a report for you. Much cheaper than monthly subscriptions when you're just starting out.
Let's walk through an example in the mortgage niche. Here's what our initial keyword research might show:
Each entry is:
Gather up your keywords using whatever tool you prefer. And get them all into a CSV or similar document. Generally the more the merrier at this point!
Not all of these keywords are right for micro tools.
Broadly there are 4 types of keywords:
Different phrases are used at different points in our customer’s journey.
Initially they might start with a broad informational keyphrase like “how do mortgages work?” or “do I need a mortgage?”. These are informational.
Then they might start actively looking for a mortgage with a navigational search like “UK mortgage providers”.
Moving onto commercial key phrases we’ll have something like “best 40 year mortgage” and transactional searches like “natwest 40 year mortgage sign up”.
Lots of people focus on Commercial and Transactional keywords. Because this is where the money is! These people are ready to buy.
These keywords also have the most competition and low volumes.
We don’t want to compete here. Instead we want to compete at the informational end - providing a micro tool that answers an early informational request.
At this end we have high volume, low purchase intent and low competition. Perfect for a micro tool to gain attention and capture emails. We then internally move the lead towards purchase using our sales funnel.
We need to focus on informational queries - people looking to understand, calculate, research or get an overview about something. Let's run our keyword list through our intent classifier prompt:
You are a search intent analyst. Categorise these keywords into:
- Informational: Learning/understanding
- Navigational: Finding specific sites
- Commercial: Researching products/services
- Transactional: Ready to buy
For each keyword, provide:
1. Intent category
2. Reasoning
3. Tool potential (High/Medium/Low)
We're looking for informational queries that could be solved with a calculator or generator tool.
Keywords:
[Previous keyword list]After filtering our examples, we're left with six prime candidates:
We’ve gone from 15 to 6. Good progress.
Now let's feed these into our tool ideation prompt:
You are an AI tool strategist. Based on these keyword groups, suggest potential tools and variations. Each tool should solve a specific problem and have potential variations to capture related searches.
Keywords with volumes:
[Give informational keyword shortlist]
For each potential tool:
1. Core tool concept
2. Primary search term it targets
3. Potential variations and their target keywords
4. Total potential monthly traffic
5. Key features neededHere's what we get back - three clear opportunities:
This second prompt has taken our short list and grouped up similar key phrases into a smaller set of tools. In this case 3 tools.
These are 3 micro tools we can take forward and build. Each connected to a set of known key phrases that real people (customers!) search for online.
You can probably already see the possibilities here! Next up we're diving into creating your AI solution. We'll take one of these validated opportunities and turn it into a reliable tool that delivers consistent value.