10 Ways to Start A Company Using AI: Research (Part 1)
The sum of all human knowledge exists on the internet. The search engine gave us the power to browse through it. The Large Language Model (LLM) gave us the power to converse with it.
This week, we’re exploring how we can ‘converse’ with AI to research before starting our company. There are so many things to research when creating a startup, so we’ll divide this topic into two parts. Let’s see if our trusty AI-cofounder can lighten the load for us.
So far in this series we’ve looked at how startups can use AI to perform:
By the end of this article, you will be empowered with prompts that’ll take you those extra steps closer to bringing your vision to life. The scaffolding of your business plan will begin to take shape!
On the menu:
Prompts are linked so you can jump straight to the action.
There’ll likely be some overlap between the output of these prompts. In fact, some prompts will likely be useful inputs to other prompts. At a high-level, you can see how we are starting broad by looking at the market before continuously narrowing down our focus to how we might start developing our idea!
As a reminder, you can use, edit, fork any of the prompts you see here by creating an account on LowTech.ai, and adjusting to your specific needs. The goal is to inspire you with useful tools to start your AI-empowered entrepreneurial journey!
A lot of content, a lot of utility. Let’s get cracking.
1. Identifying Market Opportunities
One of the fundamental building blocks to starting a company is having some theories about the current and future states of the world. These theories normally appear in question format: what will transportation look like in 5 years time? How many people drink orange juice? What are the current problems faced by potato farming companies?
With so many questions that we wish to ask simultaneously, it can be quite a chore to sit down with a search engine and trawl through links after links to find answers we care about. That doesn’t sound too fun. Why not use AI to ask all of our questions at once?
That’s exactly what this prompt here does: Entrepreneur’s Guide to Market Research. We can make use of AI’s large 6,000 word context window and rest-assured that our many questions will be answered simultaneously. Just simply type in all the questions you have, and this prompt will answer your questions, along with insights into market trends, competitiveness, and barriers to entry. Perfect for that qualitative business research!
Example: Do UK consumers eat sandwiches for lunch? What does the market for delis look like in the UK? Do UK consumers prefer to sit-in or take-out for lunch?
2. Competitor Analysis
Now that you’re all excited about the market opportunity that you’re going after, it’s time to see who is in that market already doing business. This is a simple matter of telling our AI-cofounder that we want to do a competitive analysis of the industry we’re in and the problem we’re trying to solve. For a little extra help, I asked the AI to identify what makes the strategy of each competitor unique. All of this to see what our potential edge into the market will be.
The result is spectacular. We get a thorough response, formatted for clear interpretable insights. Here is a snippet from performing a competitive analysis for the lunchtime hoagie market in London:
Strategy: Subway is a well-known global sandwich chain that offers a variety of customizable subs and hoagies.
Approach: Their approach focuses on providing healthier fast-food options with fresh ingredients and a quick assembly process.
Seems good to me! The full list from our AI-cofounder had various high street lunchtime options available to consumers in London, including a concluding paragraph on the unique strategies of each competitor.
3. Defining the Problem
The hard work is done, so now the harder work begins! Don't worry, we'll use AI to make hard work feel simple. AI has helped us to get a feel for the market environment as well as the competitive landscape, so we have a pretty good sense of what options already exist in the market.
In order to propel our vision into existence, we really need to cement our understanding of the problem we are solving for. To find this clarity, we need to formulate a statement that meets the following criteria (1):
- Is really specific
- Identifies the target market
- Defines the market size and the specific problem
- Defines the current bad solutions
- Is written in 3rd person
I’m going to use AI to generate the problem for my hoagie startup. The only input I have to offer right now is the vision I have for what I want, and maybe a loose understanding of the problem I want to solve.
Input: I want to enjoy a scrumptious hoagie made with fresh produce for lunch while enjoying a walk by the Northern Irish shore front
Output: Problem Statement: The target market, comprising tourists and locals enjoying leisurely walks along the Northern Irish shorefront, is currently unable to find a satisfying lunch option that meets their desire for a scrumptious hoagie made from fresh produce. The market size is estimated to be around 100,000 individuals who frequent the picturesque shoreline. The current bad solution involves limited availability of food vendors along the shorefront, offering only basic snacks and generic fast food, leaving people without a satisfying lunch option that aligns with their taste preferences and dietary choices.
Even with a quite ambiguous vision, our AI-cofounder was able to discern the problem from the vision, and formulate quite a compelling problem statement with great structure. This is fantastic for us as AI-empowered entrepreneurs, as now we can focus aggressively on the specific problem we are trying to solve. We’ve saved so much time in the research process so far, that we’ll have our ideas in the real world in no time.
4. Finding the Value Proposition
As stated on Wikipedia, the value proposition is the “full mix of benefits” that a company promises to deliver to customers (2). Effectively, a company’s value proposition summarises why a consumer should buy a product or service. Having a really clear sense of what our startup’s value proposition is will help us to focus relentlessly on delivering that value.
We’re going to use AI to help draft up some value proposition ideas. From Step 3, we have defined a problem statement that our startup is trying to solve for - let’s use that as one of the inputs to help figure out our value proposition. The other input we’ll need is what our proposed solution is. If we give our AI-cofounder 1) the problem we’re trying to solve, and 2) how we intend on solving it, our AI should be able to come up with some strong value propositions.
Let’s give it a go.
Problem: [ Using the same problem statement as generated above in Step 3 ]
Proposed Solution: Hoagie Heaven - a local hoagie deli positioned close to the seaside that sells scrumptious hoagies with fresh produce and flavours from around the world. Easy to hold hoagies while walking along the shore front and soaking in the beautiful Northern Irish views.
Value Proposition: "Hoagie Heaven: Elevating Shorefront Dining. Our fresh, flavorful hoagies cater to the discerning tastes of locals and tourists seeking a satisfying lunch option as they wander along the picturesque Northern Irish shoreline, filling a gap in the market and delighting hungry beach-goers with a taste of heaven."
That’s pretty amazing. I’m getting quite excited to start my idea now!
5. Creating a Mission Statement
So far we have assessed an entire market, identified competitors and barriers to entry, defined specifically the problem we’re trying to solve, and finally clarified what is the exact value proposition of our startup. With all this knowledge, we’re on our way to creating!
Creating a mission statement gives your startup an explicit statement about it’s goals, values, objectives - effectively it’s purpose for being. This sense of purpose will be quite important for you as a founder because, not only will you be working towards delivering on your value proposition, you also want to ensure that you are upholding the values of the entity you are creating. Philosophically, it is as if you are detaching yourself - your ego - from this other entity that will be living and breathing in its own way, and you want to make sure that you align with it.
A mission statement is also important for the wider world looking in on your startup! Who is this company? What does it stand for? What is its outlook on the world? By crafting a neat mission statement, your startup will have its own personality and beliefs that people can engage with.
With all that said, how can we go about crafting an impactful mission statement? I’m going to use AI to draft up some options. I’m hoping for brevity in the message, as well as full communication of the startup’s goals, values, mission, and vision.
The inputs we’ll need:
- The value proposition (e.g. from Step 4) ← this is great for our AI to have context
- The problem statement (e.g. from Step 3) ← again, great for context
- Goals: e.g. “have a deli location close to the seaside and hire friendly staff to create great atmosphere”
- Values: e.g. “fresh produce, fresh bread, scrumptious flavours, friendly staff, fast service, exciting atmosphere”
- Mission: e.g. “serve tasty hoagies from a deli for locals and tourists”
- Vision: e.g. “become the goto lunch option for locals and tourists visiting the lovely seaside, and a popular sit-in location with great atmosphere, fast service, and friendly staff. Great hoagies for all weather”
How’s this for a mission statement?
Delighting Beach-Goers, One Hoagie at a Time: Bringing joy to the picturesque shores of Northern Ireland, our deli offers tantalising hoagies made with love, delivering a taste of heaven while creating a friendly, vibrant atmosphere for all to enjoy.
Fantastic! We got quite a detailed explanation of what our startup’s purpose of being is. I quite like the notion of “bringing joy” to the shores of Northern Ireland - that’s an idea that I would be sure to hone in on when building out my startup!
It may feel like populating many of the fields in the prompt are redundant (e.g. defining what your values are or what your mission is when we already provided the value proposition). I think it is ok to feel like you are repeating yourself in these prompts - in fact, providing the AI with more context over the entire situation, it is able to absorb all thoughts and combine into one effective mission statement. If you check the actual prompt (click “Show Prompt” on the prompt shown here), you can see that the prompt actually provides some example mission statements from large established companies. In doing so, the AI is ‘steered’ towards knowing what type of output it should be generating. It’s quite a neat trick, and gives us much better results.
Recap and What’s Next…
This isn’t a formal conclusion on the topic of researching your startup with AI, as we still have a lot more material to cover. What we have done today sets the stage for all that is to come in Part 2.
We began by solely having questions about a specific industry, demographic, country, or preferences. That was all we had to work with. We learnt how to use AI to ask all of these questions at once, and get actionable insights on potential ideas to tackle the challenges that arise from the questions we asked.
From that, over the course of four additional steps, we were able to go from having questions to then having a formal value proposition and mission statement for a startup that we want to create. That’s quite powerful.
All of this was made possible by our AI-cofounder. We saw that it can actually be beneficial to feed the output of one prompt as the input to another prompt. This chaining together of thoughts from the AI gives us the chance to explore a whole host of creative possibilities.
What’s left to cover? Well, now that we have the precise problem we want to solve, the value that we are proposing to our customers, and a clear sense of what our startup’s mission statement is, we can turn our attention to researching the steps to bringing the startup to life. First up on the menu in part 2: generating our user personas.
Part 2 is here!