Every startup wants to get to the elusive product-market fit.

But unlike PMs at established companies, startup founders often have an extra layer of ambiguity to navigate — zeroing in on the target customer.

In this episode, Ian O’Donnell (Director of Design & Validation at Menlo Labs) shares his practical process to help startup founders get to product-market fit. He shares how he conducts product discovery calls and quantitative ad testing alongside founders to hone in on the target industry, user and buyer.

All of Ian’s advice is grounded in first-hand startup stories, in which he’s helped to shape their trajectory.

Chapters:

00:00Introduction to Product Discovery and Ian O'Donnell

01:56Ian's Journey into Design and Innovation

09:02The Role of Qualitative and Quantitative Research

10:41Case Study: Squint's Journey in Manufacturing

24:02Deep Dive into Manufacturing Needs and Insights

34:31Adoption Challenges and the Role of AI

36:49Squint's Journey to Success

37:55The Menlo Labs Experience

43:56Challenges in Manager Training

47:10Exploring New Directions

55:06Navigating Customer Discovery

01:01:56Advice for Founders

Where to find Ian:

• Menlo Labs

• Menlo Labs’ Fellows Program

• Ian’s LinkedIn

What can product management learn from video games?

How to make dying more fun.

In this episode, we talk to a household name in tech that you might already know — Aakash Gupta, founder of product-growth.com.

Aakash has done it all: from founding his own companies, to working in game development, to becoming VP of product at a unicorn. He shares several case studies on how to do low-lift product discovery both as a founder and as an established PM.

Because he knows what PMs are going through, Aakash has pragmatic advice on how to talk to customers as an overburdened, burnt out PM. :)

We also talk about how we can apply video game psychology (e.g. dying gameplay) to everyday product development. Take a listen!

Chapters

00:00 Introduction to Product Discovery and Akash Gupta

01:36 Akash's Journey into Product Management

05:03 Defining Moments in Akash's Product Leadership

09:09 User-Centric Product Discovery at Rap to Beats

18:23 Impactful Features and Gamification in Product Development

31:43 Insights from Epic Games and the Psychology of Gaming

40:50 Enhancing Player Experience through AI and Matchmaking

43:52 The Psychology of Failure and Game Design

47:29 The Importance of Playtesting in Game Development

51:04 Navigating User Research in Large Organizations

54:13 Leveraging Data for Product Discovery

58:21 Building Effective Customer Feedback Loops

01:01:43 The Role of Research in Product Management

01:05:01 Democratizing User Research for Product Managers

01:09:18 Balancing Analytics and Customer Conversations

01:13:49 Creating a Continuous Feedback Loop with Customers

Where to find Aakash:

• LinkedIn

• Aakash’s newsletter at Product-growth.com

• Aakash’s podcast

• Book: The Ultimate Guide to Getting a PM Job

Sometimes you can play all your cards right, and the product will still fail.

We kick off 2025 with a New Years episode of the Product Discovery Show. In this episode, Stacey Feero, Director of Product at AgencyAnalytics, talks about a 0-to-1 product she built for Canada’s largest retailer.

Stacey walks through how she identified the cause behind the product not taking off (spoiler alert: user insights), how she iterated on it, how it got to $20M ARR, and how it still got shut down. She shares valuable lessons throughout that journey on managing stakeholder expectations and matching the product to user mental models.

Chapters:

00:00 Introduction to Product Leadership

06:05 Transitioning into Product Management

13:03 Challenges in User Experience and Market Fit

19:59 Adapting to User Needs and Market Changes

25:48 Strategic Marketing and Product Positioning

31:39 Navigating Product Development Timelines

36:53 The Importance of Timing in Product Success

44:25 Transforming User Mindsets for Product Success

50:00 Lessons from Product Launch Failures

56:46 The Role of Curiosity in Product Management

Where to find Stacey:

• LinkedIn

• No Fear Product Leadership - Coaching

• Stacey’s Substack newsletter

What do you mean, that adding screens sometimes drives more user growth?

In this episode of the Product Discovery Show, Maria Cuasay, Director of Product at Ancestry, shares how she did product discovery to discover why her growth experiments weren't working.

Maria shares several case studies: one where she optimized the checkout flow at Ancestry and adding more screens actually solved more user pain points. She also shares an immersive way of de-risking physical mail marketing at Opencare. Throughout the episode, she shares practical tips for those looking to launch successful products.

Chapters:

00:00Introduction to Product Discovery

01:10Maria's Evolution as a Product Leader

03:17Defining Moments in Maria's Career

07:44Building vs. Scaling: Maria's Preference

09:47The Role of Product Discovery at Ancestry

10:37Understanding Ancestry's Business Model

11:29Optimizing the Checkout Flow

12:55The Importance of Qualitative Insights

19:14Transforming Insights into Action

27:23Lessons Learned from Experimentation

28:57Reducing Mental Friction for Better Conversion

30:36Building Reputation and Expanding Remit

31:50Innovative User Insights at OpenCare

34:58De-risking Direct Mail Marketing

39:15Rapid Iteration in User Research

42:33The Importance of Clear Copywriting

46:05Combining Quantitative and Qualitative Data

50:59Structuring Insights with Opportunity Solution Tree

54:45The Balance of Discovery and Execution

Where to find Maria:

• LinkedIn

• Medium

Need examples? All the Stories copycats that rolled out and rolled back (on LinkedIn, Twitter, soon to be Tiktok). 

Here’s why they are bound to fail: Stories feature is copied wholesale, when there were no complementary benefits it brought to the platform and the end user. 

Before we do a deep-dive into the specifics, a story:

A taco place was located in a quaint and hip part of town. One day, the owner decided to buy a neon “OPEN” sign to draw in more customers at night. It worked. Most of the other stores on the same street would close at 9pm, but new customers driving by saw the “OPEN” neon sign at the taco place and would stop to grab a bite or two. Business started booming at night and transformed the taco place.

The taco place owner told this trick to his 2 friends, a bar owner and a burger joint owner. Both were ecstatic, and decided to buy neon “OPEN” signs of their own to try out. 

Months go by, and neither of them saw any bumps in customers. After a year, they dejectedly took the signs down. 

Why didn’t the “OPEN” neon sign work for either the bar or the burger joint? 

It turns out that the bar was located in the party district, with its streets already filled with neon signs. The “OPEN” sign at the bar just kind of blended in, and any pedestrian walking down the street would not notice it against the backdrop of screaming lights.

And the burger joint? It’s not open past 8pm, so the sign was only on when there’s sunlight out, making it barely noticeable. 

The moral of the story is that what works for one business might not work for others. Each business has its unique customers, customer needs and situational contexts. Borrowing hacks wholesale would not guarantee success. 

This is unfortunately what I’ve noticed with more and more tech companies -- to copy features from competitors without an intuitive hypothesis for why this feature would be useful for its users. What we end up seeing are the same features copy-pasted across our social media apps, rolled out for 6-12 months to monitor results, and rolled back when there is less traction than desired.

Instead, the Stories feature thrived on platforms where it provided a complementary job-to-be-done to the core offering, and as a result strategically transformed the platform in the eyes of its users. 

Here’s a breakdown of why I think the Stories feature worked for some platforms and not others:

Snapchat:

The original creator of Stories, Snapchat understood its users extremely well. The job-to-be-done (JTBD) for its core messaging function was to send goofy photos of yourself to friends in a 1:1 or 1:few format. But to keep your friends updated, you need to initiate a photo post in each chat / group chat. It becomes cumbersome if you have lots of friend circles to keep updated.

Enter the Stories feature, where you can post ephemeral content that shows up to all your friends. You can keep them updated on what you’re doing or feeling without having to initiate the post for every chat. Alternatively, if you want to make Stories viewable by anyone, it becomes a ‘Day in the life’ type of ephemeral diary that anyone can access.

What you can see with Snapchat Stories is it’s complementary to the original JTBD of its core messaging feature. It transformed Snapchat from a messaging platform into a social media platform, without taking away from the core of what makes it special.

Complementary Job-to-be-done (JTBD): give all my friends or connections a candid glimpse into my life, without the mental burden of judgment or misuse

Platform strategy shift: Messaging platform → messaging + social media platform

Instagram: 

Instagram copied Stories from Snapchat, which exploded in popularity. By happenstance, the JTBD of Instagram Stories plugged a hole left behind by its core feed. The core feed emphasized perfection and aesthetics, which made creating posts burdensome. Creators made 1-2 posts a day after hours of agonizing over the perfect raw photo, filter, captions and hashtags.

Stories, on the other hand, removed the barrier for creating perfect content. Because Stories disappear in 24 hours, users created more genuine snapshots of their lives in-between perfect posts. Again, Stories were an excellent complement to the original JTBD of perfect pictures, transforming the platform with an added air of authenticity.

Complementary JTBD: Showcase more authentic snapshots of my life, without the burdens of creating perfection

Platform strategy shift: Perfection → perfection + authenticity

Twitter Fleets:

Gotta give Twitter some credit here -- they rolled out Fleets with a hypothesis and did a number of iterations to address the core problems. However, none of the hypotheses turned out to be true. 

Twitter’s Fleets was originally rolled out to “address some of the anxieties that hold people back from Tweeting” (source). However, two issues with that: 1) given the design of the Twitter Home Feed, Tweets are pretty much ephemeral to the audience, and 2) anyone who has used Fleets can tell you that creating Tweets is much easier than creating Fleets. With Tweets all you need are 140 characters of a thought. With Fleets you need to select multimedia + overlay your thoughts over them. So the original hypothesis that Fleets are lower barriers to entry = false. 

It’s no surprise that those who ended up using Fleets were existing Twitter power users. The JTBD that those power users found was that Fleets allowed them to showcase immersive multimedia. That could be a complementary JTBD to the core offering, but I imagine Twitter’s currently focused on content generation for new users (rather than giving power users more ways to express themselves). 

From there, Fleets underwent a number of iterations. My theory is that through user research, Twitter team found that new users weren’t using Fleets because of the higher barrier to creation, and that new users have a visibility problem -- they don’t have enough followers who’d be interested in their Fleets. So Twitter underwent a number of iterations, by expanding Fleet viewership to those who don’t follow you & embedding ‘Related Fleets’ into Tweets. Imagine how irrelevant Fleets became for viewers as a result!

It was no surprise that Fleets was rolled back in July so Twitter can focus on other experiments that gained better traction, like Spaces. 

Hypothesized Complementary JTBD: Lower the mental barrier for creating content

Why it didn’t work: Higher mental barrier for creating a Fleet compared to a Tweet; low visibility; low relevance. Overall, no complementary JTBD with its core offering.

Iterations Twitter did: 

• Making Fleets viewable by a wider audience, including those who don’t follow you

• Giving Fleets more entry points and relevance, through Related Fleets

LinkedIn Stories:

LinkedIn rolled out their version of Stories in February 2020, with the original hypothesis to reduce the barrier of creating content. Similar to Twitter, the mental barrier hypothesis is unfounded, given that content on the LinkedIn Feed is already pretty ephemeral and easy to create.

However, what’s unique about LinkedIn is that if you’ve invested time and effort into creating multimedia content, you will want that content to be a permanent part of your professional brand. We see that already with coaches and gurus alike creating videos on Tiktok for brand-building, and are getting massive traction. 

LinkedIn could have iterated on its Stories feature by 1) making them permanent, and 2) pinning Stories to the top of both Home and Profile feeds for more visibility. However, it looks like LinkedIn is taking an equally interesting pivot into Tiktok territory. I look forward to their future tests with engaging videos and multimedia as its next iteration. 

Hypothesized Complementary JTBD: informal way to share multimedia updates to your professional network, without any mental burdens (due to ephemerality)

Why it didn’t work: people actually want their Stories content to be brand-building and permanent

Future Iterations: Tiktok/‘Reels’ feature instead of Stories

Tiktok Stories:

Oh Tiktok, what are you trying to do? 

Tiktok recently rolled out its own version of Stories and I am unsure about its hypothesis. Similar to LinkedIn and Twitter, Tiktok core videos are already 1) easy to create, and 2) basically ephemeral. 

As an avid Tiktok consumer, I subscribe to 300+ creators, and have only seen at most 3 creators use Stories at any given time. The value proposition of Tiktok Stories is unclear to both creators and consumers, and the only time where I’ve seen some value was for creators notifying consumers of an upcoming video. (If notifications were the original intent for the Stories feature, then a notifications/community feed would be a better solution).

Tiktok has since embedded Stories from people I don’t follow into my Home Feed, making it a terrible experience all around. Because there is no complementary benefit to the core videos, I am skeptical that Stories will become useful long-term. 

Hypothesized Complementary JTBD: To reduce the barrier for creating content

How creators are actually using it: Notifications to viewers about upcoming content

My verdict: Low usefulness and relevance, unknown complementary benefit to its core offering

Summary

Social media platforms are in the attention economy - they are incentivized to capture more and more eyeballs. However, product development is a lot more nuanced than copying and pasting features. There needs to be a complementary (non-competing) JTBD that the new feature brings to the core offering, which leads to a strategic transformation of the platform. 

The best way to identify these strategic JTBDs is to deeply understand your users. Customer insights, combined with product intuition, make product development more powerful. Through constant learning and iterations, we can hone our original thesis and make features more useful. Without a deep understanding of the users and their needs, these features will never get a fighting chance to succeed. 

Subscribe for more armchair thought pieces on consumer psychology and building products for funky human brains.

Subscribe now

Share

One of the most critical aspects of doing research is stakeholder alignment during the scoping phase. This can go by a number of different names — research kickoff, research brainstorm, etc. — and generally happens before the research plan is drafted.

Stakeholder alignment also happens to be one of the more difficult things to do, because alignment of any kind requires getting the right people in a room together, setting the right context, gathering information from different brains, and (wait for it..) synergize.

Now throw in the COVID-19 remote-working curveball, and you have a pretty thorny problem to solve.

Thankfully, there are some pretty good remote collaboration tools out there. I’ve been a big fan of both Mural and Miro for this reason. For the sake of this blog post, I will be sharing my stakeholder alignment template for Mural.

The magic of Mural for alignment can be summarized by three things:

• send the board link to anyone and they can join without the friction of creating accounts

• anyone you invite can add text, move stickies etc. in real-time on the shared board

• the colourful design makes Mural a fun collaboration tool

I’m including below my play for doing remote stakeholder alignment. Hoping it could help teams out there that are facing remote-first challenges. Feel free to copy my Mural template and repurpose as you’d like!

>>> Download Mural Template here <<<

Here are some details about using the template:

Before the meeting:

• Schedule a Zoom/Hangout meeting with your stakeholders

• Create a Mural board like the one above, and include the link to the board in the Zoom chat / calendar invite

During the meeting:

• Give an overview of what will be covered in this alignment session, including goals, activities and expectations

• Gather materials to build shared context for your stakeholders before jumping into the alignment activity. Example materials include: Why are we prioritizing this research project? Where does this research project live in the quarterly roadmap? What existing data do we have for this project?

• Once you’ve gone through background materials, share the Mural board and ask all stakeholders to join the board as anonymous creatures

• Assign each stakeholder a colour. They’ll stick with their assigned colour for each of the steps

• Go through each section and follow the instructions for each step as shown on the board

After the meeting:

• Critically examine the information on the board to draft your research plan. For example, sometimes the questions brainstormed by the team are not framed as research questions; you’d want to reframe them as research questions for your plan

• Share your research plan with the team once completed for additional feedback

Feel free to let me know how your sessions go!

Follow me for more tools and best practices on UX.

Subscribe now

Limitations of Spoken User Interviews:

User interviews are one of the most flexible methods for UX Research, yet we are all familiar with their constraints. Namely, there’s always a wealth of information to dive into in a short period of time.

To illustrate, we’ll be using breakfast as an example for the rest of this post. Let’s pretend we work for a CPG company, looking at people’s attitudes and behaviours about breakfast.

Here’s what a typical spoken user interview may look like:

Interviewer: What do you like to eat for breakfast?

Participant: Eggs, oatmeal, and sometimes bread.

Interviewer: How do you like your eggs?

Participant: If there’s boiled eggs in the fridge, I’ll eat them. If there’s time, I’ll spice it up with an omelet. If there’s mounds of time, I’ll make a poached egg.

Interviewer: When do you choose oatmeal instead of eggs?

Participant: When I want to eat more healthily. Eggs can sometimes have too much cholesterol, and that’s something I’m watching out for.

Interviewer: What about bread?

Participant: When I know I’ll have a hard day ahead and need the carbs.

Interviewer: How do you prepare the bread?

Participant: If I want to be fancy, I might toast in the pan and put the egg on top. Otherwise I just eat it without any preparation.

What you’ll notice from the dialogue above is that the interviewer might want to dive into each breakfast item in depth, with follow-up questions for each.

With user interviews, they’re great for conversational storytelling, but really only one thread at a time. What’s worse, our brains are not built to hold a lot of information. That’s because our working memories are limited, we experience cognitive load when there’s more information than there are capabilities to process them.

First, a primer on how our brains and memories work. 3 terminologies:

• attention: the ‘spotlight’ we use to begin processing sensory information from the world

• long-term memory: where we hold long-term information from the past that we want to retrieve for usage

• working memory: temporary storage for information, which you can manipulate while in working memory

And 2 processes involved with these components:

• Encoding process: information comes in from the world. When I pay attention to it, the information gets held in working memory as I process and decide which parts to encode into long-term memory.

• Retrieval process: information from long-term memory is being retrieved and held in working memory as I decide what to do with it.

http://translatethebrain.com/professionaldevelopment

So working memory is very crucial for how we function. Problem though: working memory is temporary (15-30 seconds), and it’s limited in storage (holds ~7 items). 7 items, 15 seconds. Not a lot of processing power.

This is where cognitive load comes in: when we have more information than we have capabilities for processing.

Cognitive load applies to user interviews, to both interviewer and participant. As participant, they are actively retrieving information from long-term memory, holding within working memory, deciding how to share with you the interviewer, and doing all of this within a short period of time.

Likewise, for interviewer, we also have a lot of cognitive load. We have to pay attention to what’s being shared with us, hold that information in our working memory, decide what to encode into long-term memory, and also decide what to probe deeper into based on what people are saying.

As you can tell, lots of information, not a lot of space in our minds to process it all.

Fortunately though, we have evolved cognitive aids to help us. Cognitive aids are things like checklists, glossaries, concept maps, etc. They allow us to offload some of the demands on working memory onto a more physical form — so we don’t have to hold it all up in our brains.

Introducing: Plotting

With this pre-amble on working memory and cognitive load, how can we make user interviews less cognitively demanding and more visible?

Introducing Plotting. Essentially, plotting is a combination of concept mapping and the think-out-loud protocol.

The goal of plotting is to gauge at people’s mental models through conversation and visual representation.

Let’s use our breakfast example to look at how plotting would work. The first step is to capture all the different breakfast items the participant likes to eat. Eggs, salad, fruit, bread… all the raw elements are captured onto individual sticky notes.

This capturing can happen by the participant or by the interviewer, noting down the different items. This is the first phase.

The second phase is where the interactivity happens. As we are doing this capturing of raw elements, we are doing this in a shared document that can be viewed by both the interviewer and interviewee. With remote interviews, my favourite tools for capturing these on a shared board are Mural and Miro.

Then the interviewer can ask the participant to plot each raw element against a visual framework predefined by the interviewer. Before the interview, the interviewer can decide on a visual framework to use for a particular research question, and the participant can interactively plot out the pieces against the framework the researcher has chosen.

The participant is doing this real-time and thinking out loud, so the interviewer can follow along on the thought process. The goal is for the interviewer to gain a more nuanced understanding of how the concepts are structured and related in the participant’s mind — their mental model.

Here are 4 types of visual frameworks the researcher might have pre-defined:

1) Categorical Plotting, similar to card sorting. For this type of plotting, the participant is moving content pieces into categories that are either pre-defined (“closed”) or created by the participant real-time (“open”). The goal is to show how items or concepts relate in participants’ mind, and why.

2) Dimensional Plotting, where the raw elements are placed against a continuous dimension and thinking out loud. This helps us surface how concepts rank in people’s minds.

3) Multi-Dimensional Plotting, where concepts are mapped against multiple pre-defined dimensions. This helps us see interactions of raw elements with different dimensions, to see clustering patterns.

You might see clusters forming. For example, there are 2 clusters of breakfast items consumed several times a week — one cluster that the participant feels good about, one that they don’t feel as good about. After deep-diving with the participant, perhaps the insight here is that the happy cluster is considered a ‘treat’ when there’s time to prep these items, while the not-so-happy cluster is considered more practical grab-and-go options.

4) Story Plotting, similar to storyboarding and journey-mapping. Participant is capturing key actions and events in an experience, and plotting out the emotional dimensions along the way. This helps us see key events along an experience, as well as how the participant feels throughout.

If you’re wondering about the difference between Plotting versus Card Sorting, I see their difference as being two-fold. First, plotting is more living within the problem space of the double diamond, while Card Sorting is more in the solution space. Additionally, their goals are different. Plotting is to inform exploratory research, while the goal for Card Sorting is to inform information architecture.

How do I Use Plotting?

Generally, you can use plotting when you’re doing exploratory research to understand the problem space, when you have multiple items to probe deeper into, and when you want to look at relational connections between items in people’s minds.

There are a number of benefits to plotting. You get to save time by zooming into specific parts of the board with participants (instead of describing the topics to dive into more deeply). Also, there is a lot more interactivity and participants tend to be more engaged than straight-up talking for an hour. Additionally, you can get a full range of the items to investigate for any given topic. Finally, it makes synthesis a little easier, with the frameworks already pre-made. You’re essentially looking at frameworks across participants to look for patterns and themes.

In terms of how to use plotting, it looks fairly similar to general research methods:

1. Define the research question you want to investigate

2. Establish the framework: decide on the visual framework and what it would look like (e.g. categorical plotting, multi-dimensional plotting)

3. Pre-plan interview questions: figure out the best phrasing of the series of questions to the participant

4. Conduct the sessions

5. Synthesize by looking for similar themes and patterns across participant frameworks

Overall, plotting is a flexible, visual way of conducting user interviews. It overcomes some of the limitations of our working memory, and makes interviews more fun and engaging. Give it a try and let me know how it goes!

Subscribe now