How do I know if a SaaS idea is worth building?

Plenty of SaaS ideas solve a real problem and still don't make a business. The difference usually comes down to three things, and you can start checking all of them before you put much into building.

1. A specific group with a frequent, costly problem

"Frequent" matters because people pay every month for problems they meet every week. A problem that comes up once a year is a hard thing to charge a subscription for.

"Costly" matters because it sets what you can charge. If the problem costs a business a few pounds a month, there's not much room for a price. If it costs them hours every week or lost revenue, there is.

"Specific" matters because you can't build for, or sell to, everyone at once. Starting with the people who have the most painful version of the problem lets you build less, explain it clearly and find your next customers through the first ones.

2. You can reach them, again and again

Your first customers will probably come from people you know. The question is where customer 20 and customer 200 come from: trade shows, associations, online communities, partners who already sell to them, the software they already use, or customers telling each other. If you can't name a route that works month after month, at a cost that's less than a customer is worth to you, a good product won't save the business.

3. They'll pay more than they cost you

Each customer costs you something to serve: hosting, payment fees, setup and support time, and any AI the product uses. Your price has to cover that, with enough left over to pay for finding the next customer.

A rough check on the numbers

You don't need a spreadsheet model. A back-of-the-envelope check is enough:

1. Roughly how many businesses have this problem, that you could realistically reach?
2. What might each pay a month, based on what the problem costs them?
3. Multiply the two, then take a modest share, like 5%, that you might win in a few years.

If the answer is a few hundred pounds a month, it may be a nice side project but not a business. If it's tens of thousands a month, it's worth testing properly.

Warning signs

• Only friends and family are keen. They want you to succeed, so their enthusiasm proves little.
• Nobody does anything about the problem today. If there's no workaround and no spending, the problem may not hurt enough to pay for.
• The person with the problem can't buy. If the user loves it but the budget holder doesn't care, you'll struggle to close a sale.
• A good prompt in ChatGPT does most of it. Your promise has to be worth more than what a customer can get from a general AI tool.

Your advantage if you know the industry

If you've worked in the industry, you're well placed. You know the problem, you know the people who have it, and you know where they meet. That's the part a competitor copying your features can't copy quickly. Use it to get your first conversations, put rough versions in front of people and ask for your first commitments to pay. You'll learn whether the idea is worth building properly far sooner than by building the whole thing first.