Fresco

Context & Background

Fresco are a B2B2C company who provide a smart kitchen platform for connect kitchen appliances. The side users mostly see are white-label apps allowing the user to control their appliance.

This project will be focused on one particular connected appliance, it’s companion white-label app and the users who own both. I will refer to these as “Smart Kitchen Appliance” and “Kitchen Appliance Company” to preserve anonymity.

My Role & Collaboration

At the time of this project, I was the UX Researcher at Fresco.

I created the research plan, recruited the participants, moderated the sessions, analysed the data and reported the insights.

The UX Lead, Marc Brophy, drafted the proposal for this research and helped by giving feedback, prioritisation, direction and reviewing any documents I had during this project.

The research proposal & plan were reviewed by the director of UX of the Kitchen Appliance Company.

Timeline

The research study spanned 2 months in total.

Research Statement and Goals

The problem statement for this research:

*"We have some early adopters for connected cooking using the Smart Kitchen Appliance and app. However, the JTBD for connected cooking are vague. We do not know why exactly users are ‘hiring’ the Smart Kitchen Appliance and app for connected cooking.Further research with Smart Kitchen Appliance users can help clarify JTBD for connected appliances and the onboarding journey."*

We recently launched the Smart Kitchen Appliance and user base of people who owned one was slowly growing.

We had insights from home cooks without connected appliances, however we lacked understanding of the benefits these new connected users were experiencing.

Conducting this research would reveal the advantages of connected cooking, allowing us to create Job stories that guide product development, identify user problems, enhance their experience, and target our marketing efforts effectively.

Research Methodology

For this study, I used the Jobs-To-Be-Done (JTBD) framework. For the methodology I conducted interviews. We wanted to get a broad understanding of our users, and it is also generally recommended for JTBD.

To mitigate the Hawthorne effect as much as possible, there were no stakeholders present at the interviews and it was just the participant and myself. Instead, these interviews were recorded, which were then shared with the team.

The interviews lasted roughly an hour each, allowing time for chatting and building rapport with the participants before having a deep dive into their motivations and more.

Recruitment Criteria and Process

Who Was Recruited

The pool of users to recruit from was quite small as the smart appliance had only been released for 5 months and in a small quantity. Therefore I cast a wide net for recruitment, reaching out to anyone who had paired their Smart Kitchen Appliance to the app. I had a goal of 10 users, which I ended up recruiting, alongside an extra 2 to account for no-shows.

How Recruitment Was Done

I was recruiting users from the app and segmentation was done via our analytics tool Mixpanel. Once I had the cohort defined, the emails were exported and participants were recruited via MailChimp.

People tend to block things that look like ads or spam, therefore the email was formatted to look like a generic email you’d personally send or receive from someone you know. This has been effective in past campaigns.

The cohort received the following recruitment email and linked a survey asking them for their availability:

Sample Questions

Below are some questions taken from the interview script alongside the rationale behind the questions. The order of these questions was later reshuffled to flow better and be more chronological in their user journey as to promote user recall.

The interview script was based on a combination of some general known unknowns we had as a team and on JTBD model’s “Unmet Goals”, “Constraints”, “Catalysts” and “Choice set”.

Analysis and Synthesis Process

I analysed the data by listening back to the recordings and writing out the main points of their responses into a spreadsheet.

I synthesised the answers the answers to understand the users’:

• Unmet goals — what kind of change our users are trying to make in their lives and how the connected kitchen experiences fits into that.

• Constraints — what is preventing them from fulfilling these goals.

• Catalysts — events that change or create unmet goals, constraints and choice set.

• Choice set — what our users consider consider ‘hiring’ to overcome their constraints and get closer to achieving their unmet goals.

Afterwards these were used to create high and low level job stories. This was then followed by a general thematic synthesis of the answers given to identify any other themes of interest.

Outputs and Deliverables

The learnings from the interviews could be condensed to one high-level job story:

• When I need to cook, I want to save time, so I can do other activities.

At Fresco, we believed that the value of connected cooking was connected step-by-step recipes, ones that hold your hand with each step and send the right settings to your appliance at the tap of the button.

However, this research showed that this was not the core value for Smart Kitchen Appliance, but rather their goal was to get some of their time back which cooking takes. The value we provided was giving them the ability to control and monitor their cooking progress using the app whilst doing household chores or leisurely activities outside of the kitchen.

They ‘hire’ the Smart Kitchen Appliance to help them get their time back when cooking.

*"I like that I don't have to be in the kitchen. We have young kids, you know, and often if they don't need help with homework or whatever it is they're immersed in at that moment. Or if I'm doing laundry, it allows me the ability to multitask. Versus a stove where I have to kind of watch it and adjust the temperature accordingly or watch to make sure it doesn't over spill or it's not burning or something. With the app, I could be doing laundry and, and controlling the cooking at the same time."*

Quote from one of the participants

Impact

The impact of the research wasn’t immediate as there were other factors also influencing the direction of the company and product, however, the JTBD research had influenced many design decisions going forth. Since the time the research had been completed:

• The status of smart appliances has gone from a pill to an entire screen, which helps greatly with controlling and monitoring.

• The status of the appliance is clearer and the user has more control over what kind of modes they can start.

• The team is currently working on re-introducing push notifications which did not make the cut when the app was being rebuilt, but now is being prioritised.

• Lastly, we now include monitoring your appliance into the definition of our most important metric: how many “smart” cooking sessions were done in a week. Including users who just monitor their appliance has boosted our ‘north star’ metric by 18%.

Reflections

I have since carried out further research studies and improved on my process, but this one was an important one for me in terms of what I learned from our users and in how I can approach this process better next time.

In reflection, there’s a lot I would have differently such as:

• Do a more thorough screening.

• Use Calendly or Google Calendar appointments for recruitment.

• Livestream interviews rather than record them.

• Ask stakeholders to help take notes.

• Analyse and synthesise data as you go until you reach a saturation point.

• Present research findings to the team rather than just posting them on Slack.