Cocchi story Vol.6

7 min readAug 31, 2020

This is our documentation of creating the service. We are going to cover here “Business Model (LIFE)”, “Mid-fidelity Concept Movie (FUN)”, “Prototype”, “Machine Learning”, “Findings from User Interview” and “Group User Testing in Japan” (10/Aug/2020–23/Aug/2020)

You can read our article about the first group user testing on Vol.5.

Cocchi story Vol.5

This is our documentation of creating the service. We are going to cover here “Low-fidelity Concept…

medium.com

Business Model (LIFE)

Currently, we are thinking of two ways of monetizing for LIFE (second phase). The first one is the franchise model and the second one is having a subscription fee from users and link one person or community with another one based on their preference.

Mid-fidelity Concept Movie (FUN)

We acted by ourselves to sharpen our concept movie before shooting the final concept movie and edited it without any sound.

Mid-fidelity video without sound

Prototype

Working prototype

We added 4 functions as follows:

Added an option of choosing “Want to visit” or “Already visited”. Because users would add photos before going to a place or taking a photo of a place to share with friends to go together later.
Added copy function to be able to move the post from a group page to a personal page and vice versa. Relocate add button from the center bottom to the right bottom as a hover circle button.
Added a hover circle camera button on the top page to be able to take a photo quickly and save it into the draft folder.
Added interaction of location picker UI to be able to choose the place even users are not located in the place where users want to add.

User Interface

We made mockups to test our UI by using Figma.

Machine Learning

We applied machine learning to our photos in Cocchi to understand the accuracy of it by using PyTorch (open source machine learning library). The result was not really accurate such as identifying a restaurant as a tobacco shop, therefore, we need to train our own machine learning model to identify accurately.

Findings from User Interview

We interviewed 3 female university students this time and changed our hypothesis because our 1st hypothesis is verified based on previous user interviews.

Hypothesis

This time, we wanted to understand whether female university students want to take a photo and save location on Cocchi or not.

Findings

Our potential users like to browse friends’ stories on Instagram and find yummy-looking dishes.
Interviewees didn’t understand the difference between personal and group pages.
Too many pins on the map and difficult to select one pin.
They plot pins on the map when they travel to do sightseeing efficiently.
They recognize Google map as a navigation tool not for saving place on it.
Want to pick multiple pins to compare the distance of the two locations.
They said that it is good to remember the place where they found on the street by chance but currently they are not taking a photo nor take a memo.
They search by name of the dish on Instagram because of too much information on it, therefore, there are big demands of organizing information potentially.
It’s important to be able to see different information at a glance on one page to compare where to go.

Next actions

Ask our potentials users whether they want to have video posts on Cocchi or not.
Prepare some default channels on personal pages to understand more clearly the difference between group pages.
Add filter function on “all page” to be able to see only necessary pins and compare with other categories.
Need to do more interview to understand whether our core target doesn’t need pin on the map for cafe/restaurant or not.
Should emphasize that Cocchi is an app to stock community information in the concept movie.
Update our map to be able to pick multiple pins or change the color of the picked pin and make it be able to compare with other pinned.
Need to do more user interview and understand our 2nd hypothesis is not really big pain or not.

Group User Testing in Japan

We tested our service to a group consists of 3 people (female university students) in Japan from 12th August to 19th August.

KPIs

We set KPIs as follows; 3 sessions, 1 post per person in 8 days. The below table is the result. We achieved our KPIs but obviously, we can see that users used our app only first 3 days.

Findings

The main reason why they didn’t use our app so much is that they hang out only once a month currently due to COVID-19, the same as the previous test in Cost Rica. Therefore, it doesn’t really make sense to evaluate our service based on KPIs.
Two of them are in Nagoya and one of them is in Tokyo, therefore once all of the posts in Cocchi, the map shows a middle point in the map which is inconvenient. The distance between Nagoya and Tokyo is around 350 km so the map shows completely other prefecture where no one live in the group.
When users belong to a lot of groups, notifications would be annoying.

Next actions

Use the Product-Market Fit Survey (“Must-Have” score) instead of KPI. (You can check more detail in the below link if you are interested in it.)
Show current location on the map instead of the middle point of posts.
Add the function of switching off notification sound and just showing updated notification signs so that users would recognize updates from the members in the group.

Must-have score

We asked the testers “How would you feel if you could no longer use Cocchi?” on a scale of 1 to 4. (“1. Very disappointed”, “2. Somewhat disappointed”, “3. Not disappointed (it isn’t really that useful)”, “4. N/A - I no longer use it”)

Three all of them answered “2. Somewhat disappointed” and below are their reasons and comments.

I haven’t used it that much yet and having good expectation after actually using it.
I can say that other app will be more than enough for me, but Cocchi is also attractive. I think as I get used to it, I’ll find out more about it. Therefore, it is bad that I won’t be able to use it anymore.
I misunderstood that the app is for searching restaurants. I rethink properly how to use it after hearing about the app again, then I want to use it now and later. I just found the restaurant yesterday and I was going to put it up tomorrow when I have time. Won’t I be able to use it anymore?