Issue No. 2
Maximize Your Product's Potential with AI: A Strategic Approach
New technologies like ChatGPT can be hard to ignore. But how do you know when and how to add them to your product?
Reading Time: 7 min
Source: DALLE, told to extract key phrases from this newsletter, create a prompt and using the style illustrative, colorful, futuristic create an image.
I'm glad you're here!
Unless you've been living in a hole for the past few months, you've likely heard of many new AI tools that have sprung up suddenly.
ChatGPT, especially has captured everyone's imagination.
In this issue I'm sharing:
- Why we're about to enter a new age of products more pivotal than Web 2.0, the Social Graph, and the App Store?
- How I think about how new technologies like ChatGPT fit in my product strategy.
- Bonus: Some of the fast wins I'm seeing
Is AI a more significant wave than Web 2.0, the Social Graph, and the App Store?
If you've followed the evolution of the internet, you know there are peak moments when an inflection point happens, and how we see the world is forever changed.
We're at another one of those moments.
Those who embrace it will succeed.
Those who don't will find they're obsolete.
There is a force at play far more extensive than the technology backing ChatGPT and the other remarkable AI technologies powering it.
That's consumer expectation. Tastes and expectations change rapidly; look at the fashion industry. But while tastes iterate quite frequently, expectations are cumulative.
Let me explain.
Like before, the genie is unlocked, and now the expectations of what AI promises are being set.
Users have experienced "done for them" solutions.
(Don't believe me? ChatGPT gained over a million users in 1 week, 70 days faster than Instagram and 145 days faster than Spotify. Schools are banning it, companies are battling how to detect it, and my mom is asking me about it. The genie is released, my friends).
What's "done for you" mean?
They're solutions that are natural to use and expand users' abilities without requiring them to do the work. They don't have to search through results to get what they want (ex., Google, Stackoverflow). They don't have to do a task; it's done for them (Ex. naming something, organizing notes, creating an outline, and coming up with quiz questions). They don't have to look for patterns; they're shown to me (ex., Anomaly and pattern detection).
And on and on.
But the good news is that we're early into this new cycle, and there's plenty of time for you to adapt to the new order of things.
Plus, many early adaptations are going to be superficial and hurt expectations. So you have room to think and strategically make a bet.
Curious how you can adapt? Keep reading!
Be strategic — How to think about applying new technologies like AI and ChatGPT
I highly encourage experimentation with new technologies. Give it a try, play with it, be inspired, spin up a proof of concept, or two or three.
But before adding them to your product, think strategically.
Here's how I think about new technologies and if they fit today's product strategy, if it should wait till later, or if it even fits.
1. Does it fit my current product strategy?
Notice I put a time component to this framework. New technologies must be considered in the context of your current product strategy.
You set your current product strategy for a reason.
It's okay to change your strategy if the opportunity is far larger. But don't go chasing new tech for brief momentum or fame. That's a losing strategy.
If it doesn't fit your current strategy, keep it in the future opportunities list if it makes sense. But don't just dismiss it.
Hot Tip – If you're a product leader within a company and the current "hot thing" doesn't fit your current strategy. Be sure to emphasize the "why" to your stakeholders. They may not think about the opportunity cost, misunderstand the impact, or underestimate the effort. Help them along the journey, so they have a deeper understanding.
Finally, here's an extreme example of not doing product strategically. In the late 1990s and early 2000s, companies added .com to their names even though they were not internet-based companies. Customer expectations were moving to online services, but companies said they were a .com while their software had no internet capabilities. It signaled a strategy change, but without any real change. It was superficial.
Bonus Point – Don't do gimmicks. A quick flash in the pan isn't worth the confusion to your user base.
2. Ask yourself, is this a Vitamin, Candy, or a Painkiller?
What do you accomplish if you build a technically cool feature and no one uses it? You accomplished wasting precious time, resources, energy, and market opportunity.
First, what are we talking about?
- Vitamins – Something people could use, but most can do without them.
- Candy – Something people find fun to use but don't add value.
- Painkiller – Something people need to make their lives easier.
I don't recommend spending much time on products and features that are Vitamins. The path to success is slow and full of bumps and bruises. Sometimes, it can be hard to see this because a feature seems so revolutionary.
Many AI implementations you see will be Vitamins disguised as Candy. You'll know it to be true when they get lots of attention but no traction.
Candy features can be helpful; they can attract user attention and inspire imagination. But just like the calories in Candy are shallow, so are most Candy features. But they can be an excellent place to start if you have a larger vision for a Painkiller.
Painkillers are where you want to invest. They solve a real customer's problem. They come in all sizes, massive pains and minor pains. Look at your customer job journey; you can solve many types of pains.
And if done it right, Painkillers do it without tasting like medicine. In fact, they taste like Candy.
(btw, this is why I'm excited about AI. I think they can be a Painkiller that tastes like Candy. But I'm worried most will just be adding Gummy Vitamins at best).
Good ways to tell if you have a painkiller:
- It's a 10x solution
- You already have related product feedback
- It's highly differentiating and is still aligned with your customers' Jobs to be Done
3. Can I iteratively apply it?
I snuck in iteration above from Candy -> Painkiller. Velocity matters in the product world. Some new technologies take a long road to add.
Whenever possible, I prefer to iterate to my end goal.
In fact, the biggest career mistake I ever made was not remembering this rule.
What's this mean if you can't iterate there? First, be 100% certain you can't. Ask people you trust if they see a smaller path forward. But once you do that, you've got to go with your product strategy, instinct, and feedback.
But also make sure you check the last one below.
4. What are the potential negative impacts?
Sometimes a new technology seems so wonderful we don't stop to ask ourselves, "Just because we can, should we?"
Many new technologies have hidden negative impacts.
For example, for all the wonders of social media (probably how you found this!), there are many negatives. Look at what it's done to our attention spans and the impact on children.
Not everything is so drastic though.
Here are some things to consider:
- Will it lead to misaligned customer expectations?
- Will it lead to negative or abusive user behavior?
- Will it require you to staff up in a way you're not prepared for? (like fighting spam and abuse for email providers)
Bonus: Are you adding AI to your SaaS?
If you're adding AI or ChatGPT-like features to your SaaS, I'd love to hear how! Drop me a line on Twitter @OrionSeven or reply to this email.
Some of the quick wins I'm seeing are:
- Help Support – Pour all your blog posts and help articles in and create a fantastic knowledge base (I'm sure this is being productized left and right though).
- Search Support – Why make people search their content when ChatGPT can organize it for them and help them find it.
- Creative Support – Suggest alternate or more memorable names. Start them off with something usable in their context. You see this in all sorts of marketing tools, but bring it inside your app whenever someone has to provide a name or description for something.
- Buddy Support – How can you apply AI to 10x someone? You see this with GitHub's copilot solution, reviewing your code and writes it for you.
- Done for you Solutions - But my favorite ones we'll see with AI are those that invert the work completely. Where a user starts with a complete solution and the user augments it. Imagine providing a description for a new landing page, and it builds it for you entirely.
Did you like this newsletter?
Can you do me a favor and like or retweet this thread on Twitter to help support me grow ProductFoundry?
Join the ProductFoundry Newsletter
Signup for weekly insights into making great SaaS products and companies.