Hey everyone,
​
If you're new here, welcome to Tech Flavor—where I share what’s been on my mind professionally, along with breakthroughs in IT, AI, and Health.  

Now onto this week’s issue…

The truth is, if I paid for a real human assistant for most people, they would not know what to do with them for a while. Their established routines would overshadow this new reality every time.

This scenario mirrors the challenges faced by AI assistants in corporate apps. Just in the past five years, we saw this common scenario unfold before our eyes a couple of times:

A big corporation decides to create an internal product. End users identified. Decision makers identified. Everyone recognizes the problem. We are asked to build the solution.

The MVP is out in five months. User presentations go well. The app is well-received. The downloads are good. Weekly active users? 5%.

What’s wrong here?

Well, it’s one thing most companies always overlook when they seemingly do everything right: adoption.

It turns out it’s not enough to build a solid solution, even for internal users who should be motivated to use it.

Just like any corporate app launched with limited marketing budgets, the adoption of AI-driven apps is significantly underestimated. While these solutions may impress during board meetings, the reality is different.

For any technology or routine to stick, the target group must be willing to change their habits or adopt new technology. We have yet to see real-world testimonials of immediate adoption for commercial applications pushed by corporations.

This raises the question: what's the solution?

The core issue is that people are creatures of habit. Introducing an AI assistant, no matter how sophisticated, requires users to alter their established workflows.

Remember what I started this newsletter with? If you suddenly had a real human assistant today, it would still take time to figure out what its capabilities are, its strengths and weaknesses, speed of performance, etc. Meanwhile, you’d be better off doing everything yourself.

Here are three common mistakes in AI assistant development:

1. Overlooking User Experience: One major pitfall is neglecting the user experience. AI assistants must be intuitive and seamlessly integrate into existing workflows. Complex or cumbersome interfaces deter adoption.

2. Lack of Personalization: AI assistants that fail to personalize interactions based on user behavior and preferences can feel generic and less useful. Personalized AI provides more relevant and valuable assistance.

3. Ignoring Continuous Improvement: Another common mistake is treating AI development as a one-time effort. Successful AI assistants require ongoing updates and improvements based on user feedback and evolving needs.

Corporate leaders often view AI assistants through a lens of efficiency and potential cost savings. They imagine streamlined operations, quick problem resolution, and enhanced customer experiences.

However, employees and customers might see these tools as an additional layer to navigate, complicating rather than simplifying their tasks.

After several unsuccessful launches, we managed to figure out how to boost adoption for our clients’ apps.

Use these strategies to overcome adoption barriers for your apps:

1. Clear Communication and Training: The first step in overcoming adoption barriers is clear communication. Companies (read: direct managers) must articulate the benefits of AI assistants, addressing common concerns and misconceptions. If your team believes in the mission, adoption becomes less bumpy.

2. Incremental Implementation: Instead of a full-scale rollout, consider an incremental approach. Start with pilot programs in specific departments, gather feedback, and make necessary adjustments before a wider launch. This phased approach helps in ironing out initial kinks and builds confidence among users.

3. User-Centric Design: AI assistants must be designed with the end-user in mind (obviously). This means intuitive interfaces, seamless integration with existing workflows, and providing real value in everyday tasks. You’d be surprised how many times we encountered vast differences in problem understanding between business and field. Sometimes involving users in the design process can lead to more effective solutions that are readily adopted.

4. Feedback Loops and Iteration: Track EVERYTHING. Establish continuous feedback loops to understand how the AI assistant is being used and where improvements are needed. Iterative updates based on user feedback can significantly enhance the tool's effectiveness and adoption rates.

5. Highlighting Quick Wins: Showcase quick wins and success stories within the organization. When users see tangible benefits from their colleagues, they are more likely to embrace the new technology. Celebrate these successes to build momentum.

Understand this: when you decide to build something as unconventional for your organization as an AI solution, you’re stepping into startup territory.

So you need to behave like one too. Adoption is as important as the solution itself. Make sure you think about it before you commit to spending millions on development.

Good luck!

That’s all for this week…but one more thing. If you’re enjoying this, can you do me a favor and forward it to a friend? Thanks.

-Alex

Want to partner with my needle-moving software consultancy? → Techery

Follow me on Twitter/X: @pshenianykov

Connect with me on LinkedIn: Alex Pshenianykov

What else was on my mind last week 👇

NEXT IN AI
Another assistant

Yes, I know. But the times dictate what we talk about. And this time, it’s Amazon. They just launched their Q - you guessed it, their AI assistant.

I don’t think it’s groundbreaking, but it does these 3 things well (probably):

• Q Developer offers industry-leading code generation, testing, debugging, reasoning, and agents for step-by-step planning.

• Q Business connects to company data repositories, enabling users to easily get answers, summarize info, analyze trends, and interact with enterprise data.

• A new Q Apps feature allows non-technical users to create custom AI applications using natural language prompts from company data.

FUTURE OF HEALTH
Covid expertise wasn’t wasted

Pandemic, just like war, forces all civil institutions to boost their effectiveness. It looks like the Vaccine department really took off.

There’s a new mRNA cancer vaccine making waves by sparking a fierce immune response against tumors. It utilizes the same principle as the COVID-19 vaccines, teaching body to hunt down and destroy cancer cells.

I personally think (as I follow this particular topic for a while) that cancer vaccines are coming in the next 5-10 years. It feels like we’re making a lot of great progress in this direction lately.

🫠 THIS MADE ME SMILE