AI storyboards w customer journey – jak unikać błędów prototypowania #EN117

TL;DR

• Trzy kluczowe błędy prototypowania: iterowanie w wysokiej jakości, skupienie tylko na zadaniach, pomijanie kontekstu klienta
• Pixar approach: rok testowania konceptów w low-fidelity storyboardach przed drogą animacją
• Customer journey w 4 etapach: discovery, onboarding, habit building, mastery – każdy wymaga innego podejścia
• Core loop jako fundament: Rock Band potrzebował 6 miesięcy tuningu podstawowej pętli zaangażowania
• AI democratyzuje storyboards: można tworzyć bez umiejętności rysowania dzięki promptom
• 30-dniowe storyboards: kluczowe dla debugowania problemów z retencją i długoterminowym zaangażowaniem
• Case study mental health startup: pivot dzięki testom z klientami „żyjącymi w przyszłości” doprowadził do sukcesu

Amy Jo Kim, współzałożycielka Game Thinking, przedstawiła podczas webinaru rewolucyjne podejście do prototypowania customer journey. Jak sama mówi, jej doświadczenie obejmuje prace nad grami takimi jak The Sims, Rock Band, Covet Fashion oraz współpracę z firmami jak Netflix, Procter and Gamble czy Replica. Jest również edukatorką na Stanford i współzałożycielką USC game department.

Ewolucja prototypowania – od tradycyjnych metod do AI

Kim rozpoczęła od przeglądu obecnych metod prototypowania, które wymieniali uczestnicy webinaru: Figma, Protopie, Miro, post-it notes, pen and paper, a nawet Blender. Szczególnie popularne staje się „vibe coding” – iteracyjne kodowanie w dialogu z narzędziami AI, jak zauważyła Kim po Game Developers Conference.

Dla zespołów marketingowych znane jest customer journey mapping, które mapuje podróż od awareness przez service do loyalty – co Kim określa jako „prototypowanie messaging”. Jednak żadna z tych metod nie pomaga w debugowaniu engagement i retention, co jest jej specjalnością.

Trzy błędy prototypowania, które rujnują produkty

Kim zidentyfikowała trzy fundamentalne błędy podczas prezentacji:

Pierwszy to iterowanie w wysokiej jakości. Jak wyjaśnia: „If you iterate in high fidelity, you’re going to get focused on and get feedback on all the details, the visuals, the graphics, the fonts, the layout. But you often miss getting feedback on the concepts because people get immediately engaged with the details.”

Drugi błąd to fokus wyłącznie na wykonywaniu zadań. Kim tłumaczy: „As a UX designer, this is what I learned in my classic training to focus on map out how a user completes a task… But journeys are not one task. And focusing on and storyboarding just one task… is a great start. But if you’re dealing with engagement and retention, you might think you have it figured out when you’re actually missing the bigger picture.”

Trzeci błąd to pomijanie kontekstu klienta. Jak mówi Kim: „Good design is all about creating something that’s specific and unique to the customer’s experience and habits and needs. And if you build something without customer context, you’ll often make something that doesn’t fit neatly into customers lives.”

Lekcje od Pixar – dlaczego low-fidelity działa

Kim odwołuje się do sukcesu Pixar: „Pixar had an absolutely unprecedented string of hits when they first came out. Different than any other studio in history. How’d they do that?” Odpowiedź: „They iterate with low fidelity storyboards during pre production they figure out the character development, the plot points, they test it all in low fidelity before going to animation and production.”

Kluczowe jest to, że animacja jest bardzo kosztowna. Jak tłumaczy: „Even though it’s getting better with AI, it’s still very expensive, especially for a feature film. So if you animate something and then figure out your story’s wrong, you are hosed.”

Customer journey przez pryzmat game design

Kim przedstawiła strukturę opartą na czterech etapach:

Discovery – „when they’re a visitor and they’re just figuring out what is this all about. So for Rock Band, it might be going to someone’s happy hour or Saturday night gathering and seeing it in action.”

Onboarding – „for newbies. We often think a lot about onboarding, which is important, but it’s only the thing that happens at the beginning.” Kim podkreśla: „You should actually work on your core loop earlier and do onboarding sooner toward launch.”

Habit building – „Most people don’t even really think about or model this, but this is your core loop. This is what’s the thing that pulls people back to your product or game again and again?”

Mastery – „for the smaller percentage of people that master your game or product, really get to know the systems, and then want something more.”

Core loop – fundament zaangażowania na przykładzie Rock Band

Kim ujawnia szczegóły procesu tworzenia Rock Band: „Rock Band has a core loop, a core habit loop. That’s the thing that your player does over and over again… And in Rock Band, the core loop is play a song.”

Proces był znacznie bardziej złożony: „It took us six really hard months of testing and tuning to make all those instruments work together smoothly online, which was the challenge.” Następnie: „We then, once we got the basics, introduced layers of skill Building feedback into the song itself.”

Kim testowała różne koncepty: „We had a bunch of different ideas for art, from gritty and realistic to like over the top cartoony. And we tested them with sketches first before going to full on animation.”

Zespół miał luksus posiadania artystów: „At Rock Band, we had the luxury of having a bunch of sketch artists who could just whip things up in real time in meetings. Most people don’t. Most teams don’t.”

AI jako demokratyzacja storyboardów

Kim przedstawia rozwiązanie: „Well, Genai to the rescue. If you can’t draw now, you can prompt.” Podstawowy prompt obejmuje: „make a black and white line drawing. You introduce your protagonist, you put your person in a setting in their life and you have them interacting with your product through a device.”

Kroki procesu według Kim:

• Zdefiniuj protagonistę
• Opisz journey stages – „gdzie są i jakiego urządzenia używają podczas discovery, onboardingu, 21 dni po rozpoczęciu”
• Stwórz story beats – „nie każdy ekran, ale duże action points”

Praktyczna wskazówka

Kim dzieli się „hot tip”: „don’t try and put thought bubbles and dialogue bubbles in into your storyboards. Add those in Canva or Google Slides later because it’s really hard to get the AI to make those.”

30-dniowe concept storyboards w praktyce

Kim wprowadza koncept: „These are lo fi sketches that visualize a month in the life or even two or three months in the life of your customer using your product. The key thing about these storyboards is that your product isn’t the star, your customer is the star your product is a supporting player.”

Zastosowania według Kim:

• Testowanie 7-dniowego trialu prowadzącego do konwersji
• Mapowanie 30-dniowej retencji
• Rozwiązywanie długoterminowych problemów

Kim wspomina pracę z Netflix: „I worked with Netflix on a really tricky two month drop off that all their analytics couldn’t figure out and solve. We figured it out with some targeted qual.”

Case study: pivot startupu mental health

Kim przedstawia szczegółowy przypadek: „Recently I was working with a hot mental health startup. They had a great idea. There was some internal discussion about if it was the right model.”

Proces znalezienia właściwych klientów: „First, we found customers among their target audience who are living in the future… people that are on the forward edge of trying to do the thing that your product do.”

Dla mental health app: „we found young adults who are already trying to use AI tools to supplement therapy… There were people that weren’t doing therapy and just trying to replace it with AI. But then there was this segment of people who had a therapist, but it wasn’t as often as they’d like.”

Kluczowe odkrycie: „we found that a lot of people wake up in the middle of the night, their thoughts are racing, there’s nobody they can talk to, they don’t want to text their therapist and bother them, but they really need support.”

Największe zaskoczenie w badaniach: „it turned out they wanted something quite different than what we were showing them… they wanted their therapist to be in charge, but they wanted the AI coach to sit in on their therapy sessions, take notes, and be able to share a synthesis of what went on during the two or four weeks in between sessions with the therapist.”

Rezultat: „If you ever hear I’ve been looking for something like that from your carefully selected target audience, you are on the right track… And they’re now rapidly growing.”

Narzędzie Journeymaker – automatyzacja procesu badawczego

Kim zademonstrował swoje narzędzie składające się z czterech modułów:

Screener Builder – „you differentiate between your large future market, meaning your tam… And your high need beachhead, meaning the first 20 to 50 people that need your product or want to play your game yesterday.”

Jobs Finder – „you can upload unstructured research notes, say the answers from your screener questions. Interview notes that you took transcripts of interviews you did with 5 10, 15 subjects.”

Persona Designer – Kim podkreśla: „A year ago I would have been horrified by this because I have a big problem with synthetic research, synthetic Personas… This isn’t for being done with your research. This is for ideation. This is for planning your research.”

Storyboard Maker – generuje story beats dla każdego etapu journey z konkretnymi pytaniami o triggery, aktywności, feedback i progress mechanics.

Framework story beats w Journeymaker

Kim szczegółowo opisuje strukturę story beats:

First impression i setup:

• „What’s the first impression they’re going to have?”
• „If they want to learn more, how do they do it?”
• „If they’re evaluating whether they want to purchase or download something, how do they do it?”

Onboarding beats:

• „How do they get started?”
• „What’s the first activity they do?”
• „How do they know if they’re making progress?”
• „How do they know if they’ve completed onboarding?”

Habit loop beats:

• „What triggers them to return to the app or game?”
• „What activity do they do when they’re there?”
• „What kind of feedback do they get to know if they’re on the right track?”

Mastery beats:

• „How can you unlock content or something else?”
• „What can you earn, possibly invites?”
• „Are there milestones that people can achieve along the way?”

Praktyczne wdrożenie według Kim

Kim oferuje warsztat: „I am doing a one day workshop on April 26th… from 9 to 11:30am Pacific. That’s the hands on workshop. You’re going to build your own concept storyboards.”

Cennik Journeymaker: „$50 a month” dla individual users, „$300 a month” dla teams z „multiplayer and team management features.”

Ważne zastrzeżenie Kim: „All AI, it gets you 70, 80% of the way there with a first pass draft for you to then edit.”

Przyszłość narzędzia

Kim wspomina o rozwoju: „We’re looking at, you know, all the natural tools for importing and exporting, but it’s very likely that lots of integrations will be a big part of Journeymaker’s offering in the next year.”

Jeden z klientów powiedział: „man, if you integrated with jira, this thing would really take off” – co Kim ma na roadmapie.

Kluczowe takeaways według Kim

Na koniec Kim podsumowuje: „You can prototype to test and tune retention using 30 day concept storyboards… You can drive re engagement in your product by designing a sticky habit loop around your customers existing habits. You can make storyboards much faster and easier if you learn how to use Genai tools.”

Możliwość democratyzacji: „Being able to do in hours what normally takes weeks and skills you don’t have is just incredibly democratizing. AI for the good.”

Ten wpis jest częścią mojej kolekcji notatek z ciekawych podcastów, webinarów i innych treści, które uważam za wartościowe i do których sam chcę wracać. Jeśli chcesz sprawdzić oryginalne źródło, znajdziesz je tutaj: Prototype your Customer Journey with AI Storyboards