Scannelier Mobile App

Target Group

I targeted people between 30 and 40 years old who drink wine occasionally. I recruited five participants whose interviews were conducted individually and via telepresence (due to the COVID-19 social distancing requirement).

Script

Friendly welcome: explanation of what I’m doing and how this interview helps me to understand the problem I wish to fix.

Background/follow-up questions:

• Who do you live with?
• What is your occupation?
• How is a typical week day for you?
• What do you usually do on the weekends?

Use of technology questions:

• How do you feel about technology in general?
• Do you use apps on your smartphone, tablet, computer to help you with tasks?
• Do you ask Google/Siri/Alexa for advice, definitions, addresses, etc?
• Have you ever bought something online?

Wine acquisition and consumption questions:

• Do you enjoy wine?
• Who do you drink it with and when?
• Do you have a prefered brand, grape, or type?
• When drinking wine, do you eat something?

Problem specific questions:

• Why do you prefer a brand, grape or type of wine between others?
• Can you tell if a wine is good or if you will like it from just looking at the bottle?
• What things do you have in mind when buying wine for a special occasion?
• What are your thoughts when doing a wine and food pairing?
• Do you use or know a tool/trick/person to help you choose wine?

Questions

1. Genre
2. Age
3. Marital status
4. How often do you enjoy a glass of wine?
5. When you drink wine, you do it because...
6. Do you usually combine types/grapes of wine while drinking?
7. Do you have wine glasses at home?
8. Are you willing to try different types and grapes of wine?
9. Do you buy most of the wine you drink?
10. Where do you buy wine most commonly?
11. Do you know which wine you need or want beforehand?
12. How many bottles do you get in the same purchase?
13. What is the most important thing you consider when buying wine?
14. Did you know that not all types of wine are served at the same temperature, and is decisive to maximize its flavour?
15. How much do you know about wine strains? (1 to 5)
16. Do you know how to prepare a food pairing with the wine you decided to drink?
17. If you answered "Yes" to the previous question: What do you usually choose first when defining your pairing?
18. Would you like to learn or have a guide on how to pair wines and meals?
19. When you drink without eating food (probably in a social gathering, party, etc.), you choose the type of wine, mainly taking into account...
20. You know or think a wine is good because...

Primary persona: Carolina

Secondary persona: Diego

HMW

• How might we make the selection process enjoyable?
• How might we instruct/teach users on how to do a wine and food pairing?
• How might we give users understandable descriptions about each wine?
• How might we provide comprehensive guidelines for wine grapes?
• How might we let users do an easy comparison?
• How might we guarantee users that they will find (or not) the wine at their local store?
• How might we teach common wine storage, etiquette and serving tips?
• How might we allow users rate and review wines they have tasted?
• How might we have users bookmark their favourite wines?
• How might we help users identify wines according to their taste preferences?
• How might we suggest other wines based on the ones they enjoy the most?
• How might we partner up our solution with shopping apps and other groceries distributors to integrate services?
• How might we make users check our solution before shopping for wine?

Expanded HMW

• In what ways could we predict/know the user's purchase frequency?
• How might we help users remember when to buy wine?
• What would happen if we suggest users related products to his wine selection (like glassware, appetizers, etc.)?
• In what ways could we provide wine fun facts to spark a conversation about?
• What's stopping us from selling wine directly instead of just helping users choose?
• How might we let users know if there’s a deal or special price on wines or related products?
• What would happen if we provide users with a price comparison tool?
• In what ways might we make users recommend our solution to others?

Vivino’s answers to my HMW questions

Q: How might we make the selection process enjoyable?

A: The app encourages users to scan a bottle’s label to obtain all the information about that wine. This includes average price, full and detailed description, food pairings, etc. Since the app also works as a social network, users can score and review every wine they taste. Additionally, there are challenges that users achieve by scanning certain wines and completing tasks; that’s an interesting use of gamification.

Q: How might we instruct/teach users on how to do a wine and food pairing?

A: Users can scan or search for a wine, then they can scroll to the suggested pairings section. Additionally, users can make a reverse search by defining a food type to receive wine recommendations.

Q: How might we give users understandable descriptions about each wine?

A: Vivino offers a thorough description of every wine, from straight facts (regional style, grapes, alcohol level, etc) to highlights about how other users are consuming it. There are descriptions for each grape and style, plus some fun facts. The app even provides full details on the winery (location map, other wines produced, rating), and a vintage comparison.

Q: How might we provide comprehensive guidelines for wine grapes?

A: Each wine type and grape has a screen where users get a complete explanation. The history, acidity level, color, body, food pairing, and countries where used are some of the information given. Since names are in foreighn languages, there are tips on how to pronounce them correctly.

Q: How might we allow users rate and review wines they have tasted?

A: As the information is crowdsourced, users define the taste characteristics, vote on the taste notes, and can go deeper on reviews.

Q:How might we have users bookmark their favourite wines?

A: Every scanned bottle label is stored in “My wines” collection. Users can also add any wine from the typed search results. Once a wine is stored in “My wines”, you can add to it a place (GPS enabled), the price you paid, and a personal note. Also saved wines can be added to a wishlist or to the user’s cellar (another list intended for stored bottles that haven’t been tasted).

Q: How might we suggest other wines based on the ones they enjoy the most? How might we help users identify wines according to their taste preferences?

A: The app has a “Taste profile” that is built from the wines that the user has rated to provide personalized recommendations. Befriended users can see each other's taste profile and share recommendations too.

Q: In what ways could we provide wine fun facts to spark a conversation about?

A: At least three fun facts can be found on every wine screen. They are mostly related to the grape or type.

Q: What's stopping us from selling wine directly instead of just helping users choose?

A: Vivino sells wine and ships to 16 countries around the world. Their collection is huge, which offers options for every taste and budget. They even have special offers and discounts on bundles.

Q: How might we make users check our solution before shopping for wine?

A: The app encourages users to scan wine labels. This may suggest that Vivino is not interested in having users check the app before shopping wine, but at the precise moment of deciding. However, this might come naturally with engaged users: they will check it anyway thanks to the motivational tools and perks like the challenges, cellar list, and friends (social networking).

Q: In what ways could we predict/know the user's purchase frequency? How might we help users remember when to buy wine? How might we let users know if there’s a deal or special price on wines or related products?

A: The app has notifications that allow users to know when special offers and deals are available. This might be triggered by any of the user’s input.

Vivino’s unanswered (or partially answered) HMW questions

Q: How might we let users do an easy comparison? What would happen if we provide users with a price comparison tool?

A: There is no way of comparing wines. The only way to see more than one at the same time is through the website, by opening on separate browser tabs, yet this is still not a proper way to compare.

Q: How might we teach common wine storage, etiquette and serving tips?

A: The app does give serving tips regarding the temperature and preparation. However, storing and etiquette tips are not given. There is an almost hidden “Wine News” section in the website where there are tons of articles with serving tips, etiquette and such. Very interesting, but underestimated.

Q: How might we guarantee users that they will find (or not) the wine at their local store?

A: If Vivino doesn’t sell the wine the user selected, it will show links to online stores that do. However, the suggested links are not related to the user’s location, so shipping is not guaranteed.

Q: How might we partner up our solution with shopping apps and other groceries distributors to integrate services?

A: Vivino has a merchant partner program, which allows wine producers and merchants to sell their wines through their plattform. Additionally, the affiliate program is directed for partners who are interested in advertising the Vivino brand. However, there is no visible connection with shopping apps, supermarkets or such.

Q: What would happen if we suggest users related products to his wine selection (like glassware, appetizers, etc.)?

A: This happens in the “Wine news” articles page.

Q: In what ways might we make users recommend our solution to others?

A: The classic share button is available on the wine screens. Users may their prefered social networks, email clients, etc.

Substitute

My users can’t even relate to an enthusiast level of expertise with wines. They enjoy and drink wine, but not as frequently (twice per month) and enough to build a taste profile based solely on their experience. Rating the boldness, tannicity, dryness, etc. of a wine is something beyond my user’s knowledge and disposition.

Instead of this, users should be able to build their taste profile by asking their preference of flavours commonly found in fruits, vegetables, meats, etc. This is a valid approach used by experts and it’s more relatable to my users. I also confirmed this with a sommelier, who states that the best way to know which wines to recommend to someone depends on knowing its palate and personal taste. This covers one of the key problems in the Point of View: the user’s taste.

Combine

The target group doesn’t have a wine cellar, nor will go hunting for a specific wine sold at specialized stores. They want to know if the wine they have in front of them is good enough. This is a critical point for users, and needs to be as efficient and straightforward as possible.

Carolina would be pleased to use the bottle label scanner that Vivino features. With just one hand and without typing a letter she would find how compatible she is with that bottle in front of her. However, Diego will need a search field to type in the brand and type of the wine in his phone/computer screen.

Combining these two methods will provide alternatives that users will use depending on their context of use, which directly affects the usability of the product I‘m designing.

Adapt

Vivino has strong social functionalities: follow users, connect their Facebook contacts, make posts and comments, and share wine recommendations. These functionalities may be very interesting for at least aficionados who consider wine a hobby, but as previously defined, those are not my users.

However, the recommendations functionality is worth adapting into a lighter and indirect version. Users shouldn’t have to engage into a strong social interaction to see which related/similar wines were enjoyed by other users. In a few words, a “Other users enjoyed these similar wines” module.

Modify

According to the above (SCA) it’s a good idea to magnify the use of the wine scanner and search. It should become clear for users that this is the principal function of the product, which provides nested affordances to make a clear identification of wine and check how recommendable it is for the user.

Considering the context of use, it’s better to minify Vivino’s social networking functionalities. The key concept to distill is “safety in numbers”, to make users feel more confident by showing how many others have tasted/enjoyed the wine they have just scanned/searched. This will reinforce the app’s credibility and give a sense of humanity behind it.

Put to Another Use

Although is not the desired intention of the solution to design, this product can also be used in supermarkets and wine stores. A standalone version could be used in POP interactive displays, encouraging users to scan the bottles they are interested in and see how well they match with their taste. Also, wine brands can use this app to attract new consumers through precise recommendations within their own product catalog.

Eliminate

In general terms, there’s a lot I could simplify from Vivino’s app to make a more relevant product for my users. Buying through the app or where to buy are not useful functionalities that don’t aim to solve the biggest problem and pain point in the customer journey map. My users are not facing a huge assortment of options at their local grocery store or delivery app. It all comes down to “I like these flavours and will drink wine in these conditions, what’s the best you have for me now?”.

Rearrange

Vivino is not designed to present users with a specific task, and the UI demonstrates there are 4 main uses for the app (top lists, search, friends, my profile). And that’s fine for users who want an app that helps them manage their hobby and learn a lot from it. But my users need constraints that allow them to focus on the most important part and functionality of the app: find the right wine. A nested affordance (aka. Wizard pattern) approach provides a much more efficient step-by-step process that suits better the task my users need to perform.

Assumptions

1. My personas don’t have time to waste while doing grocery shopping. Installing an app, setting it up and getting it to work is too much to ask.
2. Users of my app will scan/search a wine or two, but then they will forget or lose interest and go back to safe/random choices.
3. The target group is a very small niche, making the app obsolete in a short amount of time.
4. My users are avoiding supermarkets and grocery stores due to the health risks of COVID-19. This changes the primary persona’s channel and context.
5. Some users will find it too difficult to scan a bottle’s labels. Accessibility or skill barriers won’t let them operate the main functionality thus rendering it unusable.
6. The taste profile works for some users, but others will receive inaccurate wine matches and suggestions.
7. Inadvertently, I will end up wasting the user's time by making them scan and search through options.

Challenge

1. How can I make the setup and scanning processes as efficient as possible? Do users have to create an account in order to use the app? Is it necessary altogether?
2. How can I keep users interested in my app? Am I designing a discardable tool that works only a few times?
3. Am I narrowing too much the features and functionalities for the sake of efficiency and cognitive load optimization? Can I include more features without affecting the user flow and making it more attractive to a bigger audience? How can I make it attractive to extreme users (wine connoisseurs and complete clueless users)?
4. If users can’t have a bottle physically in front of them, how can they perform a label scan? How can I bypass/replace the scan feature to show wine matches and suggestions to users? Do they always have to scan or search for a wine to remember the last one they enjoyed?
5. How can I make the scan and search features as accessible as possible? Which disability types are the most affected by accessibility issues when using the camera function of a mobile device?
6. How can I guarantee that the taste profile survey will provide accurate and enough data from users?
7. How can I let the user know that they have scanned/searched enough options? How and when should I recommend them to settle with the best match so far?

Ideas

1. The taste profile survey must be as short as possible (10 - 15 questions max.). The survey should reveal each question individually and transparently showing the user’s progress. Illustrations, plain language and affordable UI interactions must be in place to make it a positive experience. Allow social networks and third-party accounts sign-up can also help by providing user’s data (only the reasonable amount) without them having to type it in. I must keep in mind that once the app has been set up (taste profile complete) it becomes a tool for “I-want-to-know” micro-moment.
2. The app’s database must be broad enough to always deliver results to scan/search queries. This will prevent users from deserting and promote exploration. I should reward users with gamification patterns (badges, levels, challenges) and exclusive discounts (partner up with supermarkets, stores and wine brands).
3. I should include small cards with interesting content like fun facts and pro tips that encourage deeper learning about the scan/search result. These must not obstruct the main content.
4. Users should always be presented with the option to scan or search a wine label. Also, a previously scanned and/or favourite wines list should be available, so users can review their progress and guide future purchases.
5. The UI must make it easy to select and operate any of them with one finger. For the search field, predictive search (or autocomplete) must be available. I must keep the UI in line with a very simple conceptual model, and also seek to comply (at a high level) with accessibility standards. I will have to test, since early prototypes, the product with users with disabilities.
6. I should seek for an organization or company of wine experts that provides me with consultancy to build the taste profile survey and endorsement. If I show users this professional background, they will trust the app more.
7. I should provide a comparison tool, so that users can spot differences between the latest scanned/searched wines. I must define a match score that measures how compatible the scanned wine is with the user. I will also display a “best match so far” alert when the user has found the highest level of match score per session.

Mapped ideas

Now

1. Label scanner (using device’s rear camera) to identify a wine option at the supermarket or grocery store.

2. Predictive search field for regular type-in search functionality as an alternative to label scanning.

6. The label scanner tool has an auto-shutter functionality that automatically submits the image to bring back the wine screen.

8. The task of scanning/searching a wine will unfold like a “wizard” setup. Nested affordances will be gradually displayed to guarantee that users achieve their goal in terms of a micro-moment.

9. Reward users with users after completing certain challenges. These are easy and/or frequent tasks like: achieving a certain number of scans/searches, finding a 50% match score wine, scanning/searching 10 wines of the same type/brand, etc.

13. Recommend to a friend button that sends a message (email, SMS, QR, WhatsApp, Messenger) to another user with the app download link. “Wine App helped me choose the wines I truly enjoy! Install the free app here. Cheers.”

15. Deliver notifications for special offers and ask to favourite a previously scanned/searched wine.

18. A previously scanned and favourite list for users to look back into.

19. Number of users that scanned and enjoyed a wine.

Wow

3. Before scan or search, ask users if there’s a food pairing and range number of guests for better suggestions.

5. Match score displayed after a scan/search has been performed. Tooltip will provide information to explain that the score is based on the user’s taste profile and expert advice/endorsement.

7. Other users enjoyed these similar wines list to display no more than 3 items that have a similar match score and type.

11. Best match so far alert displays when the user has scanned/searched more than one wine, showing the wine and score for that matter.

12. Share button for every scanned/searched wine to create a social media post (Facebook, Twitter, Instagram, etc.). The post includes a link to a static page with the wine’s data and promoting the app’s features.

14. Add to compare tool to allow users identify main differences between the last 2 or 3 scanned/searched wines.

17. If the scanned/searched option has a low match score, suggest a wine type/grape to guide the user to a better score.

20. Plain language and comprehensive descriptions in the wine screen must be provided and prioritized: a) Name/brand, type/grape, year, match score (%), discount coupon (if available). b) Taste characteristics, pairings, users enjoyed this wine (number), nationality, related/similar wines, fun fact, serving tip.

How

4. First time users will have to build their taste profile by completing a survey that asks for their preference about different types of food and aromas.

16. Partner up with a wine brand, supermarket or distributor to provide exclusive and immediate discounts.

Necessary

10. One-thumb-one-eye operability throughout the UI that promotes an efficient interaction in spite of the user’s external environment full of distractions. High accessibility must be met too.