You have an AI product, how to get users?
You've built this great thing, but there are no users 😱 Don't worry, we've got you covered!
DevStats (sponsored)
With DevStats, you can Improve your process, identify bottlenecks, and ship better products, faster.
Deliver more business impact and make data-driven decisions.
Improve your developer experience and avoid unrealistic workloads.
Better visibility to identify bottlenecks and risks to your codebase.
Check out DevStats and sign up for a 14 day trial (no credit card required).
Thanks to our great long-term partner DevStats for sponsoring this newsletter article!
Let’s get back to this week’s thought.
Intro
Building real-world side projects is one of the best ways to level up as an engineer. And in the last year or so, a particular niche for building products became very popular. AI products.
Building side projects is not just great for leveling up, but it can be a great additional income.
We, as engineers are great in technical details and building things, but when it comes to sales and marketing (getting the users), it’s usually not our cup of tea.
But don’t worry, we’ve got you covered!
I have teamed up with
, who is a Best-selling Author, a GTM strategist, Advisor & Mentor and also a fellow Slovene.She is sharing her insights on how to get users when building AI products.
After reading this article, you will never go through this again: “Spending a long time building something, while nobody is interested in the product”.
Let’s get straight into it.
Whenever you are building a product, you need to have a go-to-market plan
In most simple terms, your go-to-market strategy is your plan for attracting customers for your product.
It starts pre-launch (when you envision your product and select your target audience) and continues post-launch when you try to find predictable and repeatable ways to scale your business.
A go-to-market strategy is the difference between a hope and a plan.
— J.P. Eggers, Associate Professor at NYU Sternm School of Business
The first obvious question is how a go-to-market strategy for AI products differs from “normal products”
Not that big of a difference. You still have to follow core business-building principles:
Go where the audience is.
Create 10X value added to differentiate.
Test the willingness to pay as soon as possible so you are not building products that users will not pay for.
Based on my experience working with over 50 AI applications in 2024, 4 principles make the go-to-market strategy for AI products a bit different than for other products:
Target audience might not be that easy to guess
Since the technology is relatively new, it is tough to guess the target audience.
For most teams, it works better to “do something” and then reverse-engineer what works well and recalibrate our future product development, marketing, and sales efforts on users who are well-retained with the product.
Don’t rely on people buying the product because it’s an AI product
Most people will not magically buy it just because it is an AI product.
As a new product on the market, we often have to deliver 10x value and develop UVPs (unique value proposition = your promise to the audience) and USPs (unique selling proposition = how are you different/the best at serving this customer).
Often, we have to communicate that without even mentioning AI.
Test willingness to pay as soon as possible
If you are building on top of an AI model, your application will have actual usage costs that you must incorporate in your pricing so you are not losing money when users happily use your AI product. Remember User ≄ Customer.
Users can say they “love it” - but will they pay for it? This is important if you are bootstrapping or building on a shoestring budget.
First, prioritize Free or Inexpensive ways to attract users to your product
Save yourself some $$$ by focusing on low-hanging fruits before building fancy marketing and sales products.
People are eager to explore AI solutions; by now, most people feel safe doing so.
Use this to your advantage and start building to harvest the interest of curious potential customers in communities, events, social media, and other free and inexpensive platforms where your audience is.
Let’s get next into the case study of the go-to-market plan for an AI product.
Case study: 5-stage go-to-market plan for MedAsk AI
MedAsk AI is an LMS for symptom checking built upon an OpenAI model developed by a doctor, pharmacist, and developer as a part-time project on a zero budget.
Here is the demo of the product: DEMO
All “marketing” came down to:
Investing $150 in Google ads for the acquisition of beta testers
Posting to 10 relevant social media groups
WOM (word of mouth) and warm intros from personal networks to get clients
This is how we visualized their go-to-market plan:
Stage 1: Get 1000 users into the product
Mission Critical: See and measure how users interact with your product.
Tactics Used: $150 invested in Google ads, asking audience for feedback.
With this investment, they got approximately 1200 users for the product.
You must be getting your target audience to the product. Always go where the audience is actively searching for solutions. If you do not know, ask them.
Watch out: Invest in analytics setup so you will know how many users you get on the website and what they do. You will first want to measure website hits vs. registrations or user activations.
Stage 2: Get the first 100 Weekly active users (WAU)
Mission Critical: Get signals of the product market fit by seeing the product retention (min 10-20%) users come back in a week.
Tactics Used: The team listed 10ish Facebook and Reddit groups and posted their solution. It resulted in 1000+ registrations when posted in 5 local Facebook groups. If you do not know what to write, follow this example:
Watch out: At this stage, it is very important to note that you need to measure activation (how many users experience the value of the product for the very first time) and retention.
For that, you will need user-tracking analytics or some server proxy. You cannot improve what you cannot manage.
There is no need to collect emails if you do not want to do that (even though it is recommended so you can invite users back to the product), but user-centric analytics is a must.
Stage 3: First 1-2 Paid customers
Mission Critical: Get 1-2 companies who will pay for the integration of this product to their website, apps (API)
Tactics Planned: Based on Initial traction and product quality assessment, a team will prepare a sales deck and get meetings with their network's complaints with direct outreach and warm introductions.
The first 5-10 customers are probably already in your phone book or one call away.
If this is not the case, you can consider joining enterprise acceleration programs, Pitch competitions, or using PR and events to get the interest of partners and investors.
Here is a quick guide on how you can get to #1 on ProductHunt if your audience is there: Win on Product Hunt: A Step-by-Step Action Plan.
Watch out: First, a founder must be included in the sales and business development process. It is tough to guide someone to see a solution if you have not done it before.
Stage 4: First 5-10 Customers
Mission Critical: OK - we solved one case - now let’s find 5-10 people who need a similar solution.
Tactics Planned: Creating a case study and presenting it at events, media in professional magazines and communities.
Watch out: Focus is the name of the game at this stage - you will see so many new opportunities, but what you want at this gentle stage is more repeatability and a clearer picture of who your target audience is.
Most companies must select a niche segment and do “more of the same” before securing sufficient traction and resources to move up the market.
This approach is called the beachhead strategy - win a smaller segment before you move on. Meta, Figma, Google, Amazon, and many other companies used it in their early go-to-market journey.
Stage 5: Scaling to 25 paid customers
Mission Critical: Find repeatable and predictable Go-to-Market Motions
Tactics Planned: Partnerships, Account Based Marketing and Sales and Enterprise communities
Watch out: In most cases, only 1-3 GTM motions work well. Always go where the audience is first, and be realistic. If you have less than 5 people in marketing, how many things can you do simultaneously? Do less, but do it better so you will achieve a critical mass.
My key takeaways from Maja’s insights
Gregor here again. Thanks to Maja for sharing all these great insights! Here are my key takeaways:
Build your MVP as soon as possible and invest a small amount of $ in acquiring your first beta testers and get feedback.
After that, check for user retention (min 10-20%) users should come back in a week.
Get paying customers as soon as possible as that would be the ultimate signal of the product-market fit.
Following this strategy will prevent you from spending a huge amount of time and effort on development and of course $ on marketing, just to figure out the fit!
Last words
This has been more product / business oriented article, but VERY important to know and understand as an engineer or a manager.
Thanks to Maja for sharing her insights on this very important topic. Make sure to follow her on LinkedIn for more go-to-market tips and subscribe to her newsletter
where she shares weekly long-form tips on go-to-market topics!You can also use this simple template called GTM Power Hour to list your initial go-to-market assumption and build a plan like we did for this case study.
Maja has been kind enough to also give a 20% discount (use code “GREGOR20”) on her best-selling book called Go to Market Strategist or a bundle with the 100+ Go-to-Market Checklist that will guide you to your launch. You can check it out here: https://gtmstrategist.com/
Liked this article? Make sure to 💙 click the like button.
Feedback or addition? Make sure to 💬 comment.
Know someone that would find this helpful? Make sure to 🔁 share this post.
Whenever you are ready, here is how I can help you further
Join the Cohort course Senior Engineer to Lead: Grow and thrive in the role here.
Interested in sponsoring this newsletter? Check the sponsorship options here.
Take a look at the cool swag in the Engineering Leadership Store here.
Want to work with me? You can see all the options here.
Get in touch
You can find me on LinkedIn or Twitter.
If you wish to make a request on particular topic you would like to read, you can send me an email to info@gregorojstersek.com.
This newsletter is funded by paid subscriptions from readers like yourself.
If you aren’t already, consider becoming a paid subscriber to receive the full experience!
You are more than welcome to find whatever interests you here and try it out in your particular case. Let me know how it went! Topics are normally about all things engineering related, leadership, management, developing scalable products, building teams etc.
Thanks so much for the opportunity to present my take on go-to-market to your awesome community, Gregor. It is my first time collaborating with a more technical community, so I hope these concepts will resonate as well as they do in growth and product management circles. I am happy to answer any questions in the comments if someone wants to dive deeper into a topic. Thanks so much for reading, and good luck on your go-to-market endeavors, Engineering Leaders!