[coming soon] Whirlpool Brazil

problem

Like many other product segments, home appliances may require replacement parts or additional accessories to enhance their functions. However, identifying the right parts can be difficult due to the complex alphanumeric codes used to differentiate between product lines and specifications. These codes typically indicate whether a shelf is compatible with a certain fridge model, for example.

With the COVID-19 pandemic and resulting financial uncertainties, the demand for product repairs instead of buying new appliances has increased considerably. Unfortunately, this has also led to an increase in the purchase of incorrect parts and subsequent returns, resulting in a high return rate of 6.70%. Our main challenge is to reduce this rate.

goals/definition of success

We aimed to improve the quality of purchases in terms of accuracy.

‍How to measure effectiveness/accuracy?

‍Percentage of incorrect purchases resulting in rate of return

Conversion rate to evaluate whether customers feel more secure when viewing products on the site, leading to more items added to their cart

CSAT (Customer Satisfaction Score) through satisfaction surveys of consumers who purchase parts

Use of VisualEye or RealEyes tools to assess the visual impact of implemented modules

Monitoring of hot spots and clicks on features for parts and accessories
‍

‍How to assess the user's post-purchase experience?
‍By monitoring reports and comments from NPS*, social networks (social listening), and user support channels.

* Although we monitored the NPS, the calculated number was not considered a key metric because the purchase of parts and accessories is often done together with other appliances. Additionally, the sample size can be very variable and may not represent the real audience.

my process

brands/channels

Among the brands that are part of Whirlpool in Brazil, we work directly in the e-commerce channels of the Brastemp and Consul brands, as well as in the Compra Certa shopping club. However, the improvements also need to be considered in conjunction with external stores and marketplaces, where the quality of information and content sent is also extremely relevant

teams I met with

Meetings with various departments of the company related to the journey of parts and accessories were essential in identifying actions that had already been taken, actions in progress, and opportunities that we could pursue together.

a very important initiative: Parts & Accessories forum

I suggested and organized a monthly forum of 1.5 hours for all teams to present their progress since the last meeting, followed by an open debate. This allowed us to help each other and compare our work on a monthly basis, each focusing on tackling the pains discovered in our own areas of expertise.

It also provided an opportunity to discuss real examples of what people said during calls via WhatsApp and telephone, bringing developers and managers closer to this reality.

bringing people closer: contacts with consumers

A few hours of conversation with people who bought parts and accessories was fundamental.

‍Focus group
‍We conducted two thematic conversation circles with people from various regions, ages, and experiences with different brands. The conversations were rich and fruitful.

To facilitate the discussion, I prepared a simple script that provided some structure while still allowing for flexibility. We not only addressed customers' experiences with buying parts and accessories for their home appliances but also brought in experiences with other brands and the Brastemp/Consul service.

data analysis

Touchpoints
‍Where do people buy parts and accessories on websites? What is the difference between these input channels? Do people search directly in the search field or search from the washing machine they have, for example? How is the heat map of the pages? Where do they click? How many images are displayed in each product carousel? Are the images clear on clear acrylic products, for example?

‍Conversion Data
‍To understand changes in behavior, I analyzed general e-commerce data for the years 2019, 2020, and 2021. I also analyzed the Black Friday periods, during which online sales are always boosted exponentially, both inside and outside our channels. I looked at absolute sales figures and tried to associate them with the rate of return.

UX analysis

Content
‍I attempted to answer questions such as: what text have we produced and how much of it is being read and understood (based on complaints)? What alternative approaches can we try? What visual media do we have, such as pictures or videos? Is the content consistent across different areas of the site?I used analysis frameworks such as PURE to identify and visualize difficulties in using the site to purchase parts and accessories for appliances.I conducted a heuristic analysis of the existing contact points.

‍Google Analytics + Hotjar
‍Using reports from the team's CRO, I was able to observe user journeys, percentages of sessions that interacted with our modules, and the product sales funnel. On Hotjar, I reviewed heat maps, videos of recorded sessions, click zones, and content surveys.

In the desk research stage, I researched public statistics, reports from consumer research organizations, academic research, and reports on the topic of safety in the installation of home appliances.

In addition, I tried to answer questions such as: what are our advantages and disadvantages in the digital experience in the area? what has the market been doing in relation to the sale of secondary/auxiliary products?

desk research

During the desk research stage, I reviewed public statistics, reports from consumer research organizations, academic research, and reports on the topic of safety in the installation of home appliances.Additionally, I addressed questions such as: What are our strengths and weaknesses in the digital experience arena? What has the market been doing in relation to the sale of secondary/auxiliary products?

‍Competitive Analysis
‍I analyzed several players in the Brazilian and foreign markets across different retail segments. Another important point to note is how Whirlpool has been dealing with this issue in other countries.Based on these analyses, meetings, and dynamics were organized to generate hypotheses, CSD, and prioritization matrices.

flows & hypotheses

We discovered some complex scenarios that needed to be addressed. Our current flows were designed to identify areas where we could take decisive action to validate the hypotheses we had about potential solutions. These included:

Customers not knowing what kind of installation kit to choose

Insufficient content to guide customers in their decision-making process

UI and usability issues with the sales module on product pages

A general lack of knowledge among the public regarding parts and accessories for household appliances, including part names, locations, and installation procedures, which are often very specific
‍

One major user pain point we identified was that customers often receive finished products at home, but are unable to install them immediately due to inadequate or missing installation kits. In other cases, they may have doubts about which type of kit to choose.

To address these issues, we see the following opportunities:

Increase product visibility on websites

Reinforce ways to check product compatibility

Highlight the importance of installation kits in the journey of stoves, cooktops, and gas ovens

Assist users in choosing the correct kit for their needs.

UX metrics

How will we evaluate the results of the proposed changes to the digital experience?

NPS

Customer contact rate with the company

Return fee

Conversion rate

The aim is to monitor these segmented metrics only for purchases of Parts and Accessories within Whirlpool, across all brands. By tracking these metrics, we can gain insights not only into the impact on sales but also into whether we are reducing the number of interactions and actions required both by customers and the company. For instance, if the contact rate of customers who made purchases on the site does not drop significantly, it may indicate that we are still not addressing issues related to content and usability.

constraints

Everyone faces limitations.

Metrics separation
‍Most proposed UX metrics require interventions in certain systems to be obtained. Considering that there are several integrated sales systems involving other service, logistics, and telesales companies, it may take a while to obtain them.

‍The codes are really hard ¯\(ツ)/¯
Currently, there is no other way to identify appliances for service and purchase of accessories other than through the model code (such as CFS5VAR, for example). The problem is that locating this code on the product labels or invoice requires extra attention, and any mistake or the absence of labels can generate even greater frustration when a person is already dealing with an essential household product that stopped working and trying to solve an unforeseen event that already causes stress.

implemented solutions

Materializing what we discovered

solution #1

Product identifier on the category page

“My range is old, can I find parts for it here?”

‍On the parts and accessories category page, I designed a new version of a common e-commerce catalog page. In this new version, I updated the catalog UI with attention to the metrics of the previous version, bringing a functionality already present in other regions of the sites to identify appliance models.

This functionality, which we call Product Identifier, allows you to find parts and accessories by searching for the exact model of the product in the user's home. This significantly reduces the chance of making a wrong purchase.

Customers not knowing what kind of installation kit to choose

Insufficient content to guide customers in their decision-making process

UI and usability issues with the sales module on product pages

A general lack of knowledge among the public regarding parts and accessories for household appliances, including part names, locations, and installation procedures, which are often very specific
‍

Results:
+10% of viewing products in the Consul funnel
‍+6% for viewing products in the Brastemp funnel
‍40% of visits to the page interact with the feature

Available at:www.brastemp.com.br/pecas-de-reposicaowww.consul.com.br/pecas-de-reposicao

solution #2

Compatibility check on product pages

‍“Could this be the burner for my range?”A verification field has been implemented on the first fold of product pages in the parts and accessories category. This field includes an autocomplete mechanism to quickly show only compatible parts with the model of the user's home appliance.

‍Results:
+4% conversion on the entire Brastemp site passed through here in just over 2 months
‍+54% of people interacted with the feature in Consul in just 30 days

Available at:www.brastemp.com.br/productpagewww.consul.com.br/productpage‍

We have also developed a more complex module to show the catalog of home appliances that are compatible with the parts/accessories in context on the page.

This module integrates the product identifier shown earlier, through an explicit call-to-action: "Don't know your product code? Find out here."

Results:
+6,2% on conversion rate
‍+2% on page revenue value

solution #3

Content review about gas installation

The content originally presented on the gas stoves page was expanded and moved to the installation kits page, with some changes in terminology and writing based on input from the engineering team and user feedback

‍To improve usability, we added an example of the expected installation for each kit, as well as information on what is prohibited for each product. We also prioritized key information, such as which kit to choose in the case of piped natural gas.

Our focus was on educating users about important points in the purchase journey, as well as providing useful information for everyday life after the purchase is made.

Results:
+46% page read time increase in 30 days
‍bigger hot areas in pages heatmaps

solution #4

Page layout review

At the Brastemp e-commerce, I conducted a layout review and proposed several options be tested via A/B tests. These options were tailored to the types of products and contexts.

Rearranging modules on pages

Improving page patterns

Implementing a new, flexible, and rewritten technical specifications module

Highlighting an exploded view of the appliance, showing the location of each part

Including a video about the product and its installation on the page

Creating a help module with support links

solution #5

Content review

What is being said about the products? What are people missing? How can we make it more explicit in which scenario it is necessary to request the conversion of the range to another type of gas? After a content review, with an incredible help of a UX Writer, we were able to standardize and rewrite the content, change some terms, and even product names. We were also discussing the clarity of the images at this time.

solution #6

Warning: Installation kit change

The importance of changing hoses and gas valves has emerged as a crucial demand on the information side. It has been statistically and qualitatively proven in interviews that many people keep using the same kit for several years without changing it. This is a serious issue that can lead to risks arising from a gas leak. The maximum recommended period is up to 5 years.

In addition to an INMETRO (Brazilian Standards Institution) recommendation, it is easy to buy kits that are much cheaper than the originals but lack approval and are made of low-quality materials with thin hoses, for example.

To address this issue, we created and tested the performance of a quick popup using some ready-made website components that highlight the importance of having an adequate kit before the person goes to the cart.

Results:
+91% of addition of installation kits in just 10 days

solution #7

New layout for sales add-on component

The main component for showcasing parts and accessories on the website (add-on) had an outdated layout that no longer met our needs. Over time, the module became more complex and added more important text, becoming cluttered. Still, some information was not being conveyed, and we lacked images of the products.

To address these issues, I redesigned the add-on. I added the ability to insert more product information in a modal triggered by a click. I also included a horizontal carousel to showcase more products without overwhelming the rest of the page. This helped bring attention to the module. After reviews and usability tests, we were able to make further improvements to the design.

Results:
+33% purchases from the module in Compra Certa
‍+530% purchases from the module in Consul

validating and approving

In just over three hours of remote or face-to-face usability tests, we gained several insights and observations regarding our initiatives. Some aspects were great, some were even better, some were bad, and some were even worse.

We were able to make changes to some aspects, but not to others. Many technical aspects were discussed and involved with the work of other teams. All the insights were catalogued, and some changes were made before the solutions were released.

And about the general metrics?

Over the course of eight months, our team achieved incredible results. Throughout this journey, my team was super helpful, available to help, and open to dialogue. Our UI and UX decisions for the ecommerce channels were made thoughtfully and through extensive discussion.

We can't attribute all of our positive statistics solely to our UI/UX interventions, as these are complex issues that involve a broad range of factors. Several logistics, pricing, registration, and business initiatives were also implemented during this period.

As I anticipated, accurately measuring the impact of changes to metrics such as NPS, bounce rates, and contact rates proved to be quite difficult.

Results:
-29% rate of return
‍-16% of reduction in the number of NPS detractors in the digital experience category
‍
‍+ conversion rates:
‍+8,86% Brastemp
‍+16,12% Consul
‍+2,65% Compra Certa

Takeaways

‍The feeling of insecurity can persist
‍Some people still report feeling insecure when buying parts and accessories. Therefore, some initiatives have been suggested and prioritized:
‍Many things turned out well, but it's a continuous learning process! 🙌

Conduct page quizzes via Hotjar to ask what people would like to see on our product pages. For example, would they prefer to see videos, more photos, 3D models, more text, or measurements?

Produce informative videos with people demonstrating how to install the part or accessory.

Improve product images by exploring new ways of photographing and displaying transparent products.

Consider hiring a UX Writer to ensure that the way we name our products is clear and understandable to the public. For example, is "trivet" or "burner" the best way to describe the product?

‍Customization on each channel
‍The behavior of parts consumers varied across the Brastemp, Consul, and Compra Certa channels, even for the same functionalities. While this had positive impacts in all channels, the results show the need of designing segmented parts and consumer personas for each channel, this will provide a more assertive performance by the team in the coming cycles