Documento elaborado por:

Lucho Delgado
 AI for Product Management

📍
1. Intro & An overview of AI and
product management
¿Qué es AI?
La habilidad de una computadora de desarrollar funciones cognitivas que usualmente
asociamos con inteligencia humana. Es una disciplina bien amplia y hay unos sub campos,
algunos de ellos:

Machine learning: A branch of AI that focuses on the use of data and algorithms to imitate
the way humans learn and gradually improve its accuracy. Deep learning: A type of AI that
can process a wider range of data types (e.g. images and sound) and often produces more
accurate results than traditional machine learning. Natural language processing (NLP): A
technology that helps machines process and understand human language so that they can
automatically perform repetitive tasks. Generative AI: An artificial intelligence model that
generates content in response to a prompt. A common type of generative AI are large
language models, or LLMs, which are trained on textual data and then used to output new
text.

Two ways product managers should think about AI:

1. As a tool to help you build your product and improve internal processes
2. As a capability you can build into your product in the form of AI-powered features

Machine Learning: Algoritmos que detectan patrones Product Recommendation on a Retail

Site Deep Learning: More accurate result / Driverless Cars detect signs & people to move
Natural Language Processing: Autocorrectores y Asistentes Inteligentes Gen AI: LLMs,
ChatGPT o Google Bard

2 formas que un PM debe pensar la AI:

1. As a tool to help you build your product

2. As a capability you can build into your product

AI’s place in building products:

1. Data analysis: recognize patterns & trends in a Dataset

2. Experimentation: Learn what's working and implement changes quickly
3. Communication: Automate the creation of key product documentation
Every PM would be an AI PM
2 areas AI will amplify (Not replace) PM's work:

● Being customer centric (collect & analyze customer data)

● Having good business sense (better Business Cases & Decisions)

How product managers can leverage AI in their roles:

• Analyzing quantitative usage data • Analyzing customer feedback • Analyzing NPS

responses • Optimizing the product roadmap • Prioritizing which features to build •
Generating user stories and personas • Backlog management • Creating in-app copy

FAQs
How Enhance Product Development? Decision making () Automation (AB Testing / User
story) Personalization (Larger amount information to taylor experience)

What type of Team do you need for AI? Depend whats your plan: Your own Models:
Bunch of Engineers & R&D Team Connect Open AI models, maybe not a big team

Ethical Considerations? Transparency: Communicate Presence of Ai to your users Bias

mitigation: Algorithms could be, identify & address BIAS early on

What type of Company? Every single Company should go in AI

📍
2. Adapting your software strategy in
the AI era
Why AI Tools are so transformative?
Nuevas posibilidades como:

● Image Generation (Content & Social Media)

● Code Generation (Copilots)
● Voice Generation (VOice Assistants)

El Iphone revoluciono la sociedad:

● Sparked innovation in mobile technology and app development

● Paved the way for social media and mobile commerce
● Fundamentally reshaped the way we live

AI tendrá el mismo impacto transformador AI will impact PM

 ● Change the way PMs analyze quantitative & qualitative data
● Shape the in product experience and users ability to self-serve support, purchases,
etc.

Forming the Right Strategy for AI

Three questions to consider when forming your AI strategy:

1. What areas of your product need to change? No mores customer Journey or

more clicks or steps Step 1: Use Product Analytics to identify areas of friction
Booking Search Function: Assistant to Plan a Trip
2. What areas of your product will remain? ¿En que parte el usuario no quiere
necesariamente escribir una pregunta? Personalizing easy to understand
dashboard/images/Messages (Summarize)
3. What completely new additions to your product are now possible? Create a
completely personalized onboarding AI will let users get to value faster and have a
better in-product experience than ever before Duolingo Max: Conversational
Character Assistant, serve the learning experience

Laying a foundation with AI principles

Set your own principles, How utilize AI for your own work and your own customers: Pendo
Principles as a starting point:

1. Customer Centric Approach

2. Transparency and Open Communication
3. Data Governance
4. Optionality & Customization
5. Compliance with Legal & Regulatory Frameworks
6. Fairness & Equity
7. Thought Leadership
8. Setting the tone from the top

Building AI-powered features:

Framework Four levels of AI:

● Level 1: There is no involvement from AI and everything is manual. Todo Manual

● Level 2: The process is still manual and driven by the human, but AI is there to
assist, provide insights, and give recommendations. Amazon Reviews
● Level 3: AI drives the process and the human is there to fine-tune, edit, decline, and
approve things. Netflix onboarding experience (recommendations)
● Level 4: There is no human involvement and the process is fully automated by AI,
like a self-driving car. Waymo Autonomous Vehicle

PM is the key to success to AI Features

Reason #1 Moving from level 2 to level 3 is a product question, not a machine learning
question the decision to take the AI-powered feature from level 2 to level 3 was driven by
PMs and UX Designers, based on customer feedback. Reason #2 Great AI does not
necessarily equal a great Product Si la AI es re precisa y potente pero la experiencia de
usuario es muy pobre

Reason #3 In the end, the product manager controls de roadmap PM debe buscar
innovación, 3 ways:

● Create a balanced roadmap (balancing proof of concepts & delivery as well work/Low
& High Risks Projects)
● Plan to be surprised and change your roadmap
● Foster an Innovation Culture: Explore new Ideas

3 Recomendaciones:

● Avoid Moonshots, Focus on finding ways to provide value to your customers

● Create Working Groups with engineers, data scientists and designers
● Make Room for innovation, embrace a culture of experimentation with AI
● Leverage feedback to inform how you'll tweak the algorithm to provide even more
value to customers.

3. 📍AI and the product-led organization

6 Characteristics of Product Led Organization
Reúne a toda la empresa en torno al producto. Enabling each team to leverage to better
support and retain customers, control costs and achieve the most important outcomes.
Friction areas are eliminated and the product is more engaged for the users.

Companies that embrace Product Led Organization

● 28% increase in total active users

● 30% more qualified leads
● 15% increase in net revenue retention
Six characteristics of product-led organizations:

1. Align each function around the product Customer succes team builds onboarding
inside the product and marketing team leverages in-app messaging for cross sell and
upsell
2. Make decisions with data over gut feel AI & Machine learning tools can better
process and organize data, enabling product-led companies to make more and faster
data driven decisions. Team across organization can easily access product data and
leverage it in their work
3. Use the product as a marketing channel Leverage the product as an ultimate
marketing channel: Event promotion, Social Proof, Launches, Product Updates,
Sunsets
4. Have amazing onboarding Tailor in-app onboarding to the specific workflows that
are most important for a users role
5. Help users help themselves In-product chatbots can interact with users to address
common pain points, offer recommendations, and provide personalized support at
scale. Embed support documentation and resources in the products navigation or a
widget. Use in-app messages to direct users to these areas.
6. Collect and use feedback from customers Proactively collect feedback with in app
surveys and polls Let users submit feedback from directly within the product at any
time.

3 benefits of AI in a product-led organization:

• Getting smarter Foundation of data clean accurate and feasible access, Helping separate
noise from signal and analyze data faster than a human on their own. • Helping humans be
more effective Replace repetitive human workflows, tasks to the product: onboarding,
support, sales interaction, etc. • Improving product delivery Automatically create launch
plans, in app guides, materials to support frequent releases, Finding trends in qualitative
data to help shape feature development and improvements Suggesting ways to take action
and quickly iterate on new products and features.

Leveraging AI across a product-led organization

Marketing

Fuel personalized experiences at scale Use AI tools to analyze product usage and customer
feedback to build hyper-targeted campaigns. Utilize Product Usage data and Customer
Sentiment Insights, AI can speed up this analysis and automatically suggest which users are
potential power users.

Sales

Leverage Product usage data of Free & Trial Users to determine wich ones are most likely to
upgrade. Leverage AI to more effectively identify patterns that signal high engagement and
intent. Use Ai t generate personalized outreach copy
Customer Success

Use data to guide their engagements and take a proactive approach to customer health.
Identify where customers are struggling & collect feedback to better understand those
behaviors. Use AI to analyze more product usage data and customer feedback than ever
before. AI tool can auto-generate in-app guide campaigns to drive conversion, expansion or
adoption.

4. 📍Powering product-led growth with

AI
Product-led growth A go-to-market business strategy that places a company’s software
product at the center of the buying journey.

Six principles of product-led growth:

1. Providing a free user experience No better sales vehicle that the product itself, 2
common types:
● Free Trials: Full functionality of the pay product for a limited time (14-30 days)
● Freemiun: Always free but with limited version of the product in terms of functionality
and support
● Product Tours: Lets user explore your solution without ever accessing it or sign-up

AI can help in determining the right set of features for your freemium product. Balance of
features give away of the free product offering, AI can analyze data of your paid product
identify patterns and recommend which products include in your free product. 2. Delivering
an “aha” moment as soon as possible

Immediately let the user discover the aha moment, multiple ways but a particularly useful
one: In-app guide.

With AI, you can make "aha" moments happen in more ways and identify and test them at a
much faster pace
analyze product usage and behavior data to suggest aha moment, test this flows to
determine if usage of those features correlate with higher engagement or even conversion

3. Committing to best-in-class usability Optimize UX & Usability, collect Feedback

from users to know what it’s working and what is not in the time and place that
matters at most, to have an engagement product. AI could help:

○ Analyze user behavior to identify friction points

○ Automatically suggest resources based on user behavior data
○ Track what resources users select to improve its recommendations in the
future
 4. Delighting users to encourage stickiness Best products are those wich Keep
users coming back for more, robust feedback collection and quantitative usage data
are crucial to decide what features and functionalities prioritize on the roadmap. AI
could help:

○ Sort through and draw insights from more feedback data

○ Quickly categorize feedback into themes and identify where to focus efforts
5. Example: Instead of a Multistep friction to input or select
dates/locations/preferences/pricing or other filters with NPL allow users to simple
type a short description of their desired trip

6. Making purchasing feel like the natural next step Let users know what they are
missing without the paid version, and make it easy and natural to convert when they
are ready, effective tactic in-app notifications CTAs to take users direct to a page to
purchase. In-app notifications/messages analyze usage data AI could help:

○ Analyze data to identify the right moments for in-app messages

○ Automatically generate in-app guides to be reviewed, edited and set live.
○ Generate personalized product and feature recommendations
7. AI can help you run more tests than ever before about which pricing, packaging, and
coupons work best personalize dynamic pricing based on users location, company
size, time of the year, etc.

8. Building in virality at every turn Helps the super users spread the word quickly &
easily, result is a greater momentum as more and more users discover the value of
your product and let the world know about it. AI could help:

○ Analyze behavior data to identify when users are most likely to share the
product.
○ Leverage AI to generate in-app content to prompt the user to share and
collaborate.

Examples of applying AI to the six PLG principles: • Analyzing product usage and conversion
data to identify aha moments • Creating hyper-personalized onboarding experiences and
pricing strategies • Extracting themes from feedback to ensure you’re building sticky features
• Generating in-app content to prompt users to share and collaborate
 📍
5. Making product development
smarter
Product Management Life Cycle:
A cyclical framework for how great products are built and improved.

Launch: Build a launch plan to make sure customers adopt and continue using new
functionality Evaluate: Use quantitative and qualitative data to evaluate the success of a
launch Iterate: Determine how you can improve what you've built.

Phases of the Product Management Life Cycle:

• Phase 0: Define a Business Outcome

Start with a Clear business Objective that you're looking to achieve AI no puede definir el
Business Outcome

• Phase 1: Discover

Uncover the pain points of both your customers and the larger market.

AI will help save PMs time by synthesizing and identifying patterns across multiple data
sources. Generate recommendations and evidence to justify which user problem to focus on

Questions:
 ● Whats do typical user flows look like?
● What's stopping users who churn from achieving their goals?
● Is there a workflow where users who churn tend to get stuck?
● What are the trends in user feedback, and is that feedback getting acted upon?
● Do highly satisfied or "power users" tend to engage in certain behaviors?

• Phase 2: Validate

Determine the right solution to move forward with

Antes un PM tenía mucho trabajo operativo en el Discovery como: Setting Up 1o1 Interviews
→ Getting Feedback → Analyzing themes

AI powered tools can allow for quick analysis of data points across mediums and make
recommendations based on the findings. Ejemplos: Product Usage Data, In-app
polls/Surveys, Support Tickets, etc. AI can make building potential solutions and prototypes
faster and easier than ever before also AI enable testing of multiple prototypes at once

• Phase 3: Build

Partner with engineers and product designers to ensure your idea comes to life as intended
or determine how to pivot if its needed.

PM’s pueden ahora generar Nimble roadmaps can quickly incorporate changes this way
they can provide visibility in to progress and learning user a single source of truth that all
teams can access With AI, PMs can incorporate product testing into their roadmap earlier

PMs pueden crear múltiples documentos para Ingeniería y Diseño

User Stories (US) Communicate product or feature requirements from the perspective of
user value: As {Type of User}, I want to {action} so that {benefit}

Product Requirement Documents (PRDs) Outline what capabilities need to be included in a

release for design and development teams.

Acceptance Criteria: The conditions a product or feature need to fulfill so a user performing a
given task will accept it.

• Phase 4: Launch

Drive a controlled rollout of new functionality based on product usage and feedback.

Launch maybe be smaller but more frequent, teams must launch product & features much
more strategically today than the past PMs role during launch:

● Align sales and marketing in order to maximize reach to the target customers and
prospects
● Work with product marketing and provide guidance on timing and positioning
● Help decide which features to make paid vs. free
AI help to craft In-app personalized guides, walk throughs, In-app messaging, etc. Products
will undergo "smart" releases with a controlled rollout based on usage and feedback from
users. AI help to craft automatically generated dashboards & reports for tracking adoption,
monitoring usage, etc. AI will enable product-led growth mechanisms for anu product launch.

• Phase 5: Evaluate

Product Usage data

● Insights into user behavior

● See where users are getting stuck
● Learn what actions users are taking

Customer Feedback NPS and Support Tickets

● Understand whether you solved the problem identified during discovery

AI can auto-determine what is and isn't working about a new product or feature, and then
give recommendations on what to do next. With AI, PMs can more quickly and efficiently
evaluate the success of a release, accelerating their ability to identify and implement
improvements.

• Phase 6: Iterate

AI va a ayudar la iteración en cada paso del ciclo de vida de producto, PM of the future will
be measured by business outcomes achieved rather than features shipped.