10 Ways to Start A Company Using AI: Research (Part 2)

Introduction

How do we take the thought of an opportunity into the actions to create a startup? What are the stages of research that we should go through in order to test and refine our ideas?

You might initially ask yourself questions such as: what is the market opportunity for this idea? Who might my competitors be? What exactly is the problem I am solving, and what exactly is the value I am bringing? What are my core beliefs for this idea?

Fortunately, there’s no need to spend ages rummaging websites, databases, or libraries for the answers to your questions. We can use Artificial Intelligence (AI) to help uncover answers. We’ve spoken before about how Generative AI (specifically Large Language Models) gives us the chance to get fantastic answers from the wealth of human knowledge by tactical prompting with simple English questions and commands.

For a recap on how we used AI to answer the initial research questions for our idea, check out part one of this two-part article.

In part two, we’re going to build upon part one’s explanation of simple idea generations, problem statements, and value propositions. Indeed, the goal of this article is to find the meaningful steps we should take to start making our idea, taking it to market, and putting it in the hands of our desired customers.

AI will be used extensively throughout the way, with some intuition as to how each prompt was created with a certain purpose in mind. But you don’t need to worry about the specifics. You’re busy enough trying to bring your idea into the world. So, all of these prompts are readily available on LowTech.ai for you to use. A simple fill-in-the-box approach makes it easy for you to focus on your specific idea without the hassle of knowing how to prompt AI.

Without further ado, let’s crack on.

On the menu:

Part 1 (last article: here)

1. Identifying Market Opportunities: Prompt

2. Competitor Analysis: Prompt

3. Defining the Problem Statement: Prompt

4. Finding the Value Proposition: Prompt

5. Creating a Mission Statement: Prompt

Part 2 <-- This article

1. Generating User Personas: Prompt

2. Finding a Minimum Viable Product: Prompt

3. Forming a Pricing Strategy: Prompt

4. Iterating to Find Product-Market Fit: Prompt

5. Planning the Iterative Development Process: Prompt

Generating User Personas

First up: how can we figure out who the target demographic of our startup idea is? Knowing who we are targeting and selling to is essential in our quest to find product-market fit. Amazingly, we can use AI to conceive of sample user personas that embody the type of customer that the AI thinks would use our product or service.

At this point, we currently have the problem statement and value proposition for our startup. In other words, we have written down in simple sentences both exactly what the problem is that we’re trying to solve, and also what the value is that we are bringing to our customers. If we feed this information into our AI-cofounder, we should be able to come up with imaginary but very plausible user profiles that personify the traits of our desired demographic.

This prompt is inspired by Sam Cantor’s AI Prompts for Product Managers here. Within it, the prompt asks the AI to generate quite a structured overview for their imaginary user persona. Specifically, we are able to conceive of the following information for our user:

Alright, let’s give it go. As per the example used through Part 1, my AI-cofounder and I are still working on bringing a hoagie deli called Hoagie Heaven to the idyllic shorelines of my Northern Irish town. Check out Part 1 to see how I found the value proposition using this prompt.

AI response:

Persona Name: The Stroller Gourmet

Who are they? The Stroller Gourmet is an individual with an appetite for fresh, locally sourced food that can be enjoyed on the go. They may be a local resident or a visiting tourist, fascinated by the beautiful Northern Irish seaside, and enjoy consuming high-quality hoagies sourced from fresh produce whilst strolling around.

Industries: The Stroller Gourmet could be primarily involved in the tourism and hospitality industry, either providing or seeking unique eats, or they could be in other occupations, coming for a leisure lunch. They are willing to spend around $100-$200 monthly on this.

Why is it worth that much? For The Stroller Gourmet, the value is not just in the delicious hoagies but also in the unique seaside experience that reinforces their love for fresh, natural ingredients. Plus, it connects them with the local food culture, making it a much more enriching experience.

Why will they use this instead of the alternatives? The Stroller Gourmet opts for Hoagie Deli because of its focus on fresh, regional ingredients which other fast food or food stalls might not offer. Paired with the memorable experience of enjoying their meal while walking along the gorgeous seaside, it becomes an unbeatable offer.

Quote: "Thanks to Hoagie Deli, I no longer need to give up on my love for fresh produce while indulging on-the-go. It's my seaside culinary adventure."

Goals:

1. Enjoy fresh and healthy meals while strolling or sight-seeing

2. Be part of a unique local food experience

3. Contribute to local businesses

4. Explore a variety of ingredients and hoagie styles

5. Have a reliable go-to lunch option

Pain Points:

1. Difficulty finding fresh, healthy fast food

2. Lack of authentic local food experiences

3. Inconvenience in managing meals while touring or walking

4. Limited variety in hoagie options

5. Finding a consistent source of quality lunch meals

Important Features:

1. A wide variety of hoagies made from fresh, locally sourced ingredients

2. Convenient packaging for on-the-go eating

3. Regularly updated menu to keep the options exciting

4. Affiliation with local produce suppliers for authenticity

5. Easy availability for locals and tourists.

Epic. It’s great to have personified The Stroller Gourmet and be able to visualise the type of customer we’re solving problems for. This will help significantly to focus on features that are important to this type of customer.

Try it out yourself:

Prompt: https://lowtech.ai/matthew/startup-user-persona-generator

Finding a Minimum Viable Product

This is arguably one of the most important stages in creating our startup. We want to get as much feedback as possible to test if customer’s like our idea. However, we want to do it with as few resources as possible – after all, we’re still a very early stage concept with probably not much financial backing just yet.

The goal of finding a MVP is to create the most basic version of our product that contains just enough features to satisfy the needs of early customers. We want our MVP to be implementable quickly, cheaply, and impactfully. We want to validate any assumptions we’ve made, test the market for our idea, and learn from real user interactions. Combining all these factors together, we can deliver core functionality with minimal resources, reducing risk and time-wasted building a product that doesn’t have true product-market fit.

In recent years, there have been excellent examples of great MVP’s from some of the contemporary world’s biggest companies. Amazon initially focussed solely on ordering books through the internet, with Jeff Bezos himself purchasing and delivering the books to the intended recipients. Dropbox only created a video to showcase their file-syncing idea, allowing them to gauge interest without wasting time on product development. Airbnb started with only a simple web page with an ad to rent an air bed in Brian Chesky’s apartment.

Source: https://siamcomputing.com/mvp/what-can-mvp-builders-learn-from-amazon-dropbox-and-airbnb/

It doesn’t need to take much to get your idea in the open and start gaining feedback. What’s great is that now our AI-cofounder can help with some of the creativity needed to break down a problem / idea / value proposition into a simpler MVP.

To get the AI thinking along the correct lines, we can create a prompt that ‘seeds’ the Amazon, Dropbox, and Airbnb examples above. Seeding will effectively sprinkle in additional context, so the AI knows what types of answers we’re expecting to see. In this instance, the MVP examples help to steer the AI towards generating very straight forward product concepts that fully encompass the core functionality of our startup’s value proposition.

All that’s left to complete the prompt is simply adding in any vision, idea, value proposition that you have, and leaving the heavy-lifting to our AI-cofounder. Check out the response for Hoagie Heaven’s value proposition:

The minimum viable product (MVP) for Hoagie Heaven could be a small food stand or pop-up booth at a frequented beach or waterfront location. The menu would be limited, maybe one to three types of hoagies initially. This simple setup would allow you to test out your concept, gauge customer interest, and receive feedback on the taste and quality of your sandwiches.

You can either rent a space or use a mobile food stand to keep the overhead costs down. Additionally, rather than employing staff initially, you may opt to manage the operations yourself. You might also consider using convenient payment options such as mobile payment apps for transactions. Your initial investment would include the cost of ingredients, basic cooking equipment such as a grill or oven, and marketing materials like banners or flyers.

By focusing on a small-scale operation like this, you can validate your concept and test the market's appetite for the product, while also collecting data and customer feedback that will help you improve and evolve the business as needed. This MVP will give you an idea of the earning potential of your idea without a hefty initial investment

Prompt: https://lowtech.ai/matthew/find-your-startups-mvp

Forming a Pricing Strategy

Let’s continue exploring the creative potential of AI. This time, we’ll use AI to consider different pricing models and potential ancillary monetisation approaches that our startup could pursue. We’ll also get specific contextual examples, the pros and cons of each approach for our startup, and then an overall recommendation on which strategy to implement.

At this point, we’ve learnt a lot of new tools and prompts that are effective for starting our company. Now, we can really start to piece the outputs together to help create even better AI responses.

Here are the inputs we’ll use to help the AI give us the most contextual answers:

Let’s see what ideas we come up with, using The Stroller Gourmet persona from Step 2, and the value proposition from Part 1. Here’s the recommended strategy that our AI-cofounder suggests to us based on these parameters:

Recommended Pricing Strategy:

Given the startup's value proposition and the target persona's willingness to spend $100-$200 monthly, the best pricing strategy is a combination of Premium Pricing and Subscription/Loyalty Program. This approach emphasizes the quality ingredients and unique experience while also encouraging repeat business and brand loyalty. An ancillary monetization strategy of Hosting Food Tours/Events can best leverage 'The Stroller Gourmet’s' goal of being part of unique local food experiences

I love that the pricing strategy and ancillary revenue suggestions are specific to our startup’s target demographic. That intuition will help to guide our approach as we go about bringing our company to life.

Prompt: https://lowtech.ai/matthew/startup-monetisation-strategies-tailored-for-you

Iterating to Find Product-Market Fit

The holy grail of creating a company is finding product-market fit – the point at which your startup’s offering perfectly aligns with the needs and demands of a target market. It’s an iterative process to get there. Pivoting initial ideas and hypotheses is common. The important aspect is getting constant feedback and adapting your startup accordingly.

It would be cool if AI could suggest a potential pivot or feature that we could implement that might take us closer to identifying product-market fit. Maybe we can anticipate the consequence of such an iteration. Maybe we can have more clarity on what specific change we are testing.

As we have seen so far, AI’s creativity is able to take our current startup idea / MVP and return actionable recommendations. So, what recommendations might we get if we feed in our value proposition and our MVP? The AI will know what the intended value is, and then can suggest an iteration based on the MVP fed in.

Again, this is an example of us using previous AI outputs as inputs to more AI prompts! We’re really chaining together a lot of prompts to create a comprehensive business proposal.

Check out a snippet of the response:

Feature / Recommendation / Suggestion:

Implement a "Hoagie of the Day" concept where you introduce a unique, special hoagie each day that is inspired by the Northern Irish culture, local produce, or even customer suggestions.

Consequences of Implementing Your Suggestion:

1. Greater Customer Engagement: A "Hoagie of the Day" concept not only adds variety to your food offerings but also builds curiosity and anticipation among your customers. They will look forward to seeing what's new each day and this will give them motivation to return more frequently.

2. More Feedback: By rotating a range of hoagies, instead of just sticking to a few staples, you’ll gain insight on a broader array of your offerings. You will get a sense of which types of hoagies are most popular and which are not, aiding your future menu strategy.

Try out the full prompt yourself to see the full output!

Prompt: https://lowtech.ai/matthew/boost-your-startup-with-innovative-suggestions

Planning the Iterative Development Process

Part of iterating to find product-market fit is getting your offering in front of customers. It is through interacting with your idea that users and customers will give meaningful feedback. That feedback serves to adjust your approach and get you closer to product-market fit.

If we’re trying to create a hoagie deli, how do we get feedback quickly? We’ve seen how we can create an MVP – an offering that meets our value proposition with minimum resources. How do we iterate after that? We don’t simply hide away to create a big reveal after one round of feedback.

The development process we choose must have close interaction with the end user at every step of the way. To do so, let’s use an Agile methodology. Agile works on the concept of Sprints – fixed-length time chunks in which, by the end of a Sprint, you can ship a functioning version of your product to users to get feedback. Their feedback informs you of changing priorities and influences your work for the following sprint.

Source: https://www.krasamo.com/agile-development-process/

Those are the benefits of working in an Agile way. This might be a new way of thinking or working for you, so how do we actually go about doing it?

We’ll use AI to plan out Sprints for our startup, using our value proposition as the overall goal and the MVP as the initial iteration we want to accomplish. Let’s leave it to the AI to figure out how long each Sprint should be and how many Sprints there should be.

Naturally, this is just a template for how you might structure your project. At the end of each Sprint, you’ll have user feedback, and as such that might influence your plans subsequent Sprints. That’s ok.

Let’s see it in action:

Sprint 1: Concept Validation and MVP Development (Duration: 2 Weeks)

This sprint involves research and development for the hoagie concept and preliminary discussions with local suppliers for quality ingredients. Feedback needed at the end of the sprint should focus on the concept's viability and gathering ideas for potential hoagie recipes.

Sprint 2: Cart Designing and Set-Up (Duration: 3 Weeks)

In this sprint, you will be purchasing and setting up your pop-up cart or stand. Simultaneously, you would want to nail down the logistics of operating in multiple locations throughout the Northern Irish shoreline. Feedback at the end of this sprint should focus on the mobility of the stand, its appeal to passersby, and any operational inefficiencies that need to be addressed.

Sprint 3: MVP Launch (Duration: 2 Weeks)

This sprint includes the launch of the hoagie stand with the limited menu. Feedback at the end of this period should entail customer reviews on the taste of the hoagies, their receptiveness to the concept, and feedback on pricing.

Sprint 4: Refine and Update (Duration: 4 Weeks)

Taking into account the feedback received, this sprint focuses on refining the menu and updating operational processes. This could also include promotional activities to attract more customers. At this point, feedback should concentrate on improvements in operations, the effectiveness of advertising, and overall acceptance of the product line.

Sprint 5: Expansion (Duration: 6 Weeks)

This sprint focuses on potential expansion- either through location, menu, or diversifying offerings like drinks or sides. After assessing the business viability, steps can be taken to expand in a strategic location or to include additional menu offerings based on the customer feedback and demand.

Sprint 6: Tech Implementation-Digital Menu and Online Ordering (Duration: 4 Weeks)

If all things go well, you could explore using technology in your operations. Rolling out a digital menu board or implementing online and mobile ordering could be the focus here. Feedback should center on how these technological implementations are affecting business efficiency and customer experience.

Each sprint's end-goal should be evaluated with a retrospective, thus identifying what works, what doesn’t, and what can be improved. The agile process doesn’t end after the MVP creation; it’s a continual process to ensure the product is evolving and improving

Great! Six sprints, each with a time duration, what to focus on, and the specific feedback to obtain. In this example, over the course of 21 weeks you’ll have a fully functioning startup. It makes the process seem manageable and reassuring since you’re collecting feedback at every stage along the way.

Prompt: https://lowtech.ai/matthew/agile-startup-sprint-planner

Conclusion

That’s a wrap on the two-part series on how to start a company using AI. This was quite a comprehensive dive into many aspects of creating a startup. By no means is this an exhaustive list, but it should certainly point you in the right direction to hit the ground running. With 10 simple AI prompts, you’ll have a clearer sense of your startup idea in only one hour! That’s very exciting.

Check out the prompts on LowTech.ai. You can fork my prompts if you want to customise them slightly to your tastes. I’m using the LowTech.ai Pro version – this gives me access to GPT-4, an amazing step up from the free GPT-3.5 model. All answers generated are extremely concise and specific. It really is a delight to work with.

For more help on starting a company with AI, here are my other articles: https://lowtech.ai/blog/author/matthew

mc