Eng: I saved $300k on a project... How I did this:

This article is a reference to my previous material “I lost $300k on a project... 3 lessons I learned”:
https://www.indiehackers.com/post/i-lost-300k-on-a-project-3-lessons-i-learned-c3a779b5ba

I want to share my experience of how I learned from the last article.

What are we doing?
M1-project.com is a service that allows you to research and describe Ideal Customers, as well as create a Marketing Strategy for business promotion using AI.

Team.
2 co-founders (one of them is me).
1 marketer at an hourly rate
1 programmer at an hourly rate
1 designer at an hourly rate
2 assistants at an hourly rate (engaged in operational and routine work)

Market research.
Initially, we had the idea of making an advertisement generator using AI. Instead of development, we paid for a consultation with 10 targetologists (people who run Facebook ads for clients).

It quickly became clear to us that there was no need for the product.

They know how to write advertising texts themselves and often use GPT. At the same time, no one bothers with the quality of the advertisements; usually they simply describe the product as “reliable, steel doors! 10% discount”
“Get a manicure and be beautiful! 20% discount!”
and run advertising while losing clients’ money.

Pivot.
After that, we went to marketers who told us that in order to write profitable ads, you need to know everything not only about the product, but also about the target audience (its goals, problems, pain points, barriers and triggers).

Our first pivot happened here. We decided to make a product based on defining the target audience. This is how we created the “Ideal Customer profile generator” tool (https://m1-project.com/ideal-customer-profile-generator).

Idea validation.
I used the Riskiest Assumption Test (https://medium.com/mindsea-development-inc/why-your-rat-riskiest-assumption-test-is-the-real-mvp-177d66cde3e1) in my edition :D

I have formulated 4 main risks:

1. Nobody uses the Ideal Customer Profile in their work.
2. Nobody wants to pay for an Ideal Customer Profile created by AI.
3. We cannot create a high-quality Ideal Customer Profile using AI.

How did we check the risks?
Risk #1: Nobody uses the Ideal Customer Profile in their work.

Verification: We checked 20+ interviews with experienced marketers (5-10 years of experience) and found that ICP is actively used in companies. At the same time, the ICP can either be drawn up in the form of a document or simply be at the level of intuition among the founders and the company’s marketer. In any case, they know their target audience (goals, problems, pain points, barriers, triggers, JTBD, etc.) and apply this knowledge.

Risk #2: Nobody wants to pay for an AI-created Ideal Customer Profile.

Verification: We made a landing page with a registration form and ICP examples (we made them manually). We prepared a launch on Product Hunt and ordered advertising on Twitter. The whole thing cost $1,500. Here is our launch https://www.producthunt.com/products/m1-project#m1-project
We received (I don’t remember exactly) about 500 registrations and as many as 3 payments!

People actually paid for the product, despite the fact that we had no functionality; we asked for data about their business and sold it via email (by the way, this is super ineffective).

Risk #3: We cannot create a high-quality Ideal Customer Profile using AI.

Verification: By “quality” I mean “corresponding to reality.”

We learned how to do ICP using AI, in short, we did it like this:

1. Spent a month communicating with marketers to understand what ICP is and how it is created.
2. Another month was spent writing prom.
3. At the same time, we learned to enrich AI with current market data from the network.

But the best test is a test in practice.
We are currently advertising and accepting orders.
We provide a 7-day money back guarantee if the customer is not satisfied. We currently have a 5% return rate under warranty.

After the sale, 20 days later we contact the client and ask what results he received from applying the information from the document.

Based on feedback, we see that clients are getting positive results.
Case studies from our clients:
74% Higher than Average CTR with Ads (https://m1-project.com/case-studies/envidual)
2X LinkedIn Visibility with Marketing Strategy (https://m1-project.com/case-studies/cynthia-barnes)
30% Conversion to Paid Clients with the Right Targeting (https://m1-project.com/case-studies/all-around-express)
https://www.capterra.com/p/10016869/M1-project/ other reviews

Bottom line.
Active work on the project: 5 months.
Expenses: $40k.
Of these, advertising costs: $10k.
I will not tell the revenue, but it’s fine ;) Just don’t want to help competitors.

As a result, we checked the main risks and are now at the stage of searching for product market fit and developing new products.

If you liked this article, please support the launch of our new tool on Product Hunt: AI-Driven Copy and Targeting for Your Facebook Ads ❤️❤️❤️ https://www.producthunt.com/posts/m1-project-2
❤️❤️❤️
Our tool writes truly converting ads (based on the problems, pain points, needs of the target audience, and not on the product description).
It also provides high-quality settings for FB advertising campaigns.

+seven-day money back guarantee :)

I will also be pleased to hear your thoughts and ideas about the product :)