From Finders to Builders

Founder / Host

Meister

Senior UX Researcher

Kiwi.com

UX Research Lead

Rover.com

Glovo

UX Researcher

SendinBlue

User Researcher

Studio Zabaione

UX Researcher / Facilitator 

Vermilion

Product Manager

UX Research Lead | ReOps Specialist | Speaker

Being the first researcher at a company can be daunting. It doesn't have to be.

Setting everything up from scratch is not an easy feat, researchers have to wear many hats. 

Not only do they have to get buy-in from stakeholders, showcase the power of research and actually deliver top-notch work but they also need to build the foundations for their team to thrive. 

I'm Julian, a UX Researcher specialising in building research practices from scratch.

Are you trying to build yours? Let's talk.

Languages ENG | ES | ITA

This is not what "using AI for research" is about...

Julian Della Mattia

If somebody is conducting any research at your company, you need to have strong foundations in place. It will save you time, it will save you headaches and will save you money in the long run.

I'm talking about systems for finding the right participants, a way of properly storing your research output, getting the tools you actually need, creating templates, documentation... Basically building your practice from scratch and turning it into a well-oiled machine.

Many research teams are still rather small and researchers often have to start from zero. Not only they have to get buy-in from stakeholders, showcase the power of research and actually deliver top-notch work but they also need to build the foundations for their team to thrive. And this task can be daunting.

This is where I can help you out.

I'm Julian, a UX Researcher specialized in Research Operations and I help researchers in startups and scale-ups build their research infrastructure.

I've been working in tech for 7 years and after being the first researcher in different companies and helping +24 clients with their first ever research projects, I've developed the skills and the knowledge for building a UXR practice and its infrastructure from the ground up.

Through 1on1 or group coaching, I can help you or your team develop a plan for building your research practice from 0.

Should we talk?

Book an Intro call: https://calendly.com/julianxthe180/30min

Write me: julian@the180agency.com

Or visit www.the180agency.com

Languages ENG | ES | ITA

Senior Product Researcher at DuckDuckGo | Speaker | Host @ Finders to Builders

Not everybody around your organisation will consume research output the same way.

PMs will consume information differently and will look at different details than say, the marketing team.

Devs and designers might approach stuff with a different mental model.

Some people are in a hurry, others take their time to read.

That's why, if you want your research to be consumed widely, you should have layered approach.

What do each of these roles care about? What should I leave in, what should I leave out?

I like thinking about it as building with LEGO blocks.

How can you split your work into building blocks that you can use, shuffle and reuse to craft different artefacts that cater to different audiences?

Senior Product Researcher | Speaker | Host @ Finders to Builders

I used to do report presentations that were the cure to insomnia.

I spent too much time in the things that got ME excited, ME interested, not the audience.

No rhythm, no tension, no connection.

The result?

Stakeholders yawning and waiting for it to be over, just to leave my report gathering dust inside GDrive.

If you want your presentations to be a home run, focus on the things that matter to the audience.

Tell them what did you find out, why they should care and what you recommend to do next and see your insights take off.

🤖 What's the difference between AI-user answers vs human answers?

What do you need to consider when using AI for something like this?

This and more in the latest episode of Finders to Builders on AI-generated users.

One of the biggest issues when we deliver research results is misalignment with stakeholders expectations.

This leads to reactions like "Oh, we knew that", "Good but how do I put that into practice?" or "Results aren't really what we were looking for".

And then your research goes to die a slow death inside the drawer.

How can you overcome that? 

I'll tell you what works for me: create a mini sample of what results could look like.

Write hypothetical but realistic results and phrase them the same way you'd phrase the end result.

Create categories, sections, whatever you would do, just in a shorter format.
Then share it for feedback.

"We might get something like this, is this in line with what you'd expect?"
And then tweak accordingly.

By getting alignment on the end photo, you ensure that you'll deliver something that will likely get picked up and acted on.

Any plans for January? Why don't we start the year by discussing AI in user research at promptUX Conference?

See you in Berlin!

Look forward to a hands-on session, where Julian Della Mattia will explore how #AI is reshaping research practices, enhancing our ability to gather, analyse and understand data, while also debunking some smoke and mirrors surrounding AI in our field.

Get your Tickets now for the promptUX Conference and get ready with new ideas and insights for 2026!

🤖 Synthetic users. The future of research or just part of the AI hype?

After talking about this for some weeks now, I finally sat down and recorded this episode!

I ran some experiments building my own synthetic users and comparing their answers to human answers.
Are synthetic users useful? How? When? Should I consider using it for my research?

The answers to this and more in the latest episode of Finders to Builders.

Link in the comments.

This Wed I'm getting together with the one and only Christopher Nguyen to discuss building research practices, Vietnamese coffee and lettuce.

Don't miss out on this epic event.

I help UX designers go from Fuzziness to Focused (uxplaybook.org), now Freedom (backlog.design)

Design can’t live without research.

Yet most companies don’t have a research practice.

Without research, designers are flying blind.

Which leads to:

→ Surface-level insights
→ Lack of decision clarity
→ More + more iterations
→ Building the wrong thing
→ Very low feature adoption

We’ve all been there.

I invited Julian Della Mattia on the upcoming The Design of Everyday People to chat about how designers can set up a research practice in their orgs.

We’re diving into everything ResearchOps:

1. The foundations & infrastructure
2. Operationalizing research
3. Evangelizing research

And all of your questions!

Julian has built research practices from scratch at startups and scale-ups, hosts the Finders to Builders podcast, and is a senior product researcher at DuckDuckGo.

Setting up your research team?

This convo will make it 10x easier.

You be my note-taker.
I’ll be your interviewer.

📌 Drop your questions below
🎙️ Join us Wednesday, Dec 3rd

User Insights at DuckDuckGo | Speaker | Host of Finders to Builders Podcast

🤖 Have you tried using it AI-generated (aka synthetic) users already?

I ran some experiments with it, creating my own set of AI-generated users using different LLMs + real data and then conducting interviews with them vs real human users to compare.

So, what's my conclusion? Here's what I believe AI-generated user are good and bad for.

GOOD FOR
⏲ Early stage testing
AI-generated users can quickly validate very rough ideas, flows, or prototypes before involving real participants. They help teams spot obvious gaps, unclear instructions, or broken logic early on.

🧽 Polishing a protocol
- Before running a real study, AI-generated users can stress-test the research script, interview questions, or survey prompts. They can help ensure wording is clear, transitions make sense, and the structure is cohesive.

🔎 Hard-to-reach profiles
- When human participants are scarce, expensive, or slow to recruit, synthetic users can simulate these perspectives to keep progress moving. This is especially useful for niche audiences.

💪 Training less experienced researchers or PWDR (People Who Do Research)
- Non-researchers can practice interviewing, moderating sessions, or running usability tests with synthetic users in a low-stakes environment. It allows them to build confidence and technique before working with real participants. Probably my favourite use for AI-generated users.

BAD FOR
🔭 Deep exploration
- AI-generated users cannot replicate the nuanced, unpredictable, and often emotional reasoning that emerges in open-ended qualitative research. They’re not reliable for uncovering latent needs or motivations.

🌋 Researching topics with high unpredictability
- Areas where real-world variability, personal history, or complex contextual behavior matter cannot be faithfully simulated. Synthetic users tend to be more consistent and rational than humans, which limits realism.

🔒Trust & safety topics
- Sensitive areas require real human experiences and ethical oversight. AI-generated users cannot substitute for lived experience or safe handling of sensitive data.

💵 Demand or market fit
- These users are not suitable for predicting real human willingness to pay, true preferences, or adoption likelihood. They can’t model real market dynamics, emotional investment, or economic tradeoffs.

Want to learn more? Check out the last episode of Finders to Builders on Substack, Spotify and Apple podcasts.

UX Leader: "What we need is a seat at the table. it's not so hard to understand. Why don't you take us seriously?"

- "Sir, the restaurant is full and you have no reservation, I cannot simpl... ugh nevermind."

🤖 I just compared AI moderators vs. human moderators in user research interviews and I'm surprisingly optimistic about AI's potential here.

This isn't humans vs. machines. It's about knowing when to use which tool.

AI moderators are structured and linear. They follow the script, stay in their sandbox, and work best for well-defined research questions.

Human moderators dance through conversations. We leverage organizational context, build genuine rapport and know when to press deeper from different angles.

The magic happens when you combine both. Let AI cover ground quickly, then use human insight to explore the most promising areas deeply.

Don't default to "hurry mode" just because AI is fast. Use it strategically, not as a replacement for thorough research.

For solo researchers juggling multiple teams, this combination approach could be a game-changer.

Interesting in knowing more? Check out the latest episode of Finders to Builders.

taplio

Julian Della Mattia's Linkedin Analytics

Julian Della Mattia

Performance ScoreJulian Della Mattia's LinkedIn Performance

What is Julian talking about?

Julian Della Mattia's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Julian Della Mattia's Linkedin Analytics

Julian Della Mattia

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Julian Della Mattia's LinkedIn Performance

What is Julian talking about?

Julian Della Mattia's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Julian Della Mattia's Linkedin Analytics

Performance ScoreJulian Della Mattia's LinkedIn Performance

Julian Della Mattia's Best Posts (last 30 days)