PLP & Type-Ahead Preliminary Research

Project Work Recommendations

Search / Type Ahead

  1. Suggest 'search queries' and not (just) auto-results of categories and products
  2. Improved sorting of suggestions
  3. Synonymsโ€‹ and typo-correction for queries
  4. Only show products if query contains a full/partial code, or includes a low-level categoryโ€‹
  5. Search queries to include search and/or purchase history (cookie, KOB-specific)
  6. Make search query persistโ€‹ after submission (both main and sticky)
  7. Emphasize the search suggestions, not query (bolding)

Out of Scope

  • Cirrus Signal-boosting + Endeca (Phase 10)โ€‹

PLP Recommendations

  1. Filter by availability ('In My Store' and 'In My DC' into a filterโ€‹)
  2. Autoswitch the results layout to match the user's query (Grid/List)
  3. Add search suggestionsโ€‹
  4. Default to SKU-level presentation over Family (Thumbnails should match variation searched for)
  5. Display variation information (color/finish, sizes)
  6. Change pagination to 'Load More'
  7. Display category-specific attributesโ€‹
  8. Revise list-item information heirarchy (product codes, primary/secondary info, compare, documents, purchase history, select indicators only when filtered, etc. mouse over? gallery images)
  9. Create "in-cart" indicatorโ€‹/mechanism
  10. Personalize experience according to KOB
  11. Trim filters to 10
  12. Remove page refresh on filteringโ€‹
  13. Add 'Previously Purchased' as a filterโ€‹
  14. Faceted Search
  15. Add category-specific attrbitute sorting
  16. Improve filter usability ('fixed filters'?)
  17. Add image references to visual filters (ex: thumbnails for colors)
  18. Fix bug: Selections and Filters should match
  19. Enhance 'Inline-banner' cartridge
  20. Display contextual search snippetsโ€‹




Questions

We want to have these questions answered before we begin design. (Answer each question for PLP and Type Ahead unless otherwise noted).

  1. (Competitive Analysis) How do our competitors solve the problem(s)?
    1. See the Competitive Analysis section
    2. What will be different about our product compared with current similar solutions?
  2. (Heuristic Evaluation) What are the best practices? Are we following them?
  3. (PLP-only) Do people use the PLP for certain types of product (categories) more than others?
    1. eg. faucet and fixtures, not commodities
  4. (Task Analysis) What task(s) are our users trying to accomplish?
  5. (Overall sentiment) Is the current experience good? Bad? What does it do well? (What do customers like about it?)
  6. (Incremental Improvement, Future Enhancements) How can it be improved? (What have customers asked for?)
  7. What changes in user behavior will indicate that we have solved a real problem that adds value to our users?
    1. How will we measure success?
  8. Does relevancy factors in whether customers purchase the product before?

PLP-Specific

Informational Questions

Sort & Result

  1. What sort methods are preferable? (Frequently Purchased, Price, Relevance, etc.)

Product List Items

  1. Do types of products factor into what information is important? (eg. 'Size' for a fitting, 'Finish' or 'Center Size' for a faucet)
  2. Most important ↔ least important information
    1. Codes
      1. Do customers ever look at product codes (the alt code) while on search results? Why?
    2. Actions
      1. What actions need to be available (Add to Cart, Add to List, Quick View Details, Change Finish/Variation, Compare, etc.)?
    3. Availability
      1. What availability messaging is important?
    4. Indicators
      1. What indicators important? (My List, Purchase History, In cart, etc.)


  • Ask Kelly why we use family SKU?
  • Analytics for PLP or type-ahead?



Persona

Who are our users?

  • User type: T3/T4 
  • Page(s): PLP 
  • Frequency: Weekly use
  • Minimum: 12 times
  • Timespan: Over the last 6 months

Questions:

  • What types of customers are using Type-Ahead? ... PLP? (roles, job types, industry... etc)
  • What are the goals they're trying to achieve?
  • What's motivating them to use these features to achieve their goals?



Customer Feedback

What do our users say about the current experience?

Usabilla & UX Feedback Database

Search Survey Results

  1. Users expect to see accurate search results based on their query
    recSveBT9q9XyChCRrecAuosUMg4VGFcjG5b69ec8086b5fe795d44550e
    5c6aba98fe27fd1fe23fed905c3fd48fa779384ea3736d4a5c3767f5b07f245796542ded
    5c375190a779381da12df8f65c35f15ea779380cb44b852a5c0ea531c198dc049d1fbc14
    5c0ab084c1222b0c093a7b615c082b77c1222b24062540675bf5b83061a1442772139b2f
    5bdb71b005d00e6d3d3fe6445bdb4dbbc770925a0b1ed6165bd7a147c770921e941c6b93
    5bc5f6683c6088773867bc2a5bbe2abec778dd0a4d08aacd5b196744037fc1102d6655c9
    5b19668a227687190a7c4f7d5b157eb79bc4e66510482038
  2. Users expect relevant refiners
    recLSkfxbZlomqB89rectBmVKaIKBIGsZKrecEQAsgrElRxkT5s
    recw5sWrlt00hsYFerecyrdFW6jsy8BDm9recnwn2vS8X6T5ffE
    recWbDE8zsE4txkyIrecLMH0EKHy14NbT5recCZxuBQPYdgwOf4
    recr48brpbYAw8ENQrecHxTgo1PENQYjuDrecVNtWwbqePs6MoP
    reckVnGfAcRJQtmnW5c3778d1a779380a466434065c27c46698cd564848002de8
  3. Users need to see product attributes on search results
    recxaD4H27vb3xdI8reclGdoLVkEL6v2r25c65768d00caca60de0ec9a8
    5bbce5d9c778dd75ca0926e45b4e23e6de1c4a1c113a9f0d5aeb30e8db23987a14528ee0
    5acfbbd06f3b5a63543cf418
  4. Search Range doesn't make sense
    recluRLifxax4bCqFreciLTahxKHjkQtSFrecfS5rCxAPoKThfK
    recMyXEyDWHUaoU8xrecd02XIUt8f8F8inrecYodX7pWFRMbj99
    recmN4l7LUccPYxq35c943f598183874a2a18a0eb
  5. Users want type-ahead
    rec9cHU0IcfBQsaYrrecdxRf8P2xNu1V6C5c35e72fa779381b5545baf3
    5b7b1e0dd5cedc1aef26bd4e5b6097f75349930a5d31f45b5b36553fde1c4a26f46933a5
    5af0d15cdb23986bb308f01a
  6. Users don't want a page refresh after a refiner selection
    recagnGPMei67PP4BrecGV69FwyoGwwpdareclSjtisgUhIY75Y
    recQcEg1RjpFuR3nPrecHdrDVeib0E8R5m
  7. How important are product images?
    5c643dc4b8a9ee67007403d55c07f9d60f37bb130d031aa45be97d5bf5133319f027d7f1
    5af100e4db239829c65b7537
  8. Users want related items suggested in type-ahead
    rec9cHU0IcfBQsaYr5c37896ab07f243e5711198d5c375190a779381da12df8f6
  9. Want to see whole products (entire toilet, not just bowl)
    recHPl0x8NEIturPcreco5Zx2oUnSktnVo
  10. Search clearing/not clearing on submission
    rechfi39IGb21qCLU5c2682953c608818dc0381c1
  11. Users want results based on "their store"
    5b4b561a94085a512b0eadf05ad8a4f7ad834d51244aa16a
  12. Users want to see purchase history in PLP
    recBHIkmoF8G2VGId5bfc59a398133701db3e4c41
  13. Users want to see all available product results
    recDJT71UTYow1aFsrecr48brpbYAw8ENQ
  14. Users want their results auto corrected automatically "pepe" to "pipe"
    5c54b1f0172c7465c8760b82
  15. Type Ahead flashes too quickly to be helpful
    rec9cHU0IcfBQsaYr
  16. Users know how to use type-ahead
    recdxRf8P2xNu1V6C
  17. Users want to switch between list and grid
    5b85943ed5cedc053e584ddb
  18. Product titles are unhelpful
    recjEsOf1A9g7Zb1U
  19. Users want to add products to cart from PLP
    recLSkfxbZlomqB89
  20. Users want to sort by price
    5bf6f62ea3855b4d19378104
  21. Users need pricing on PLP
    recpu70MOMNZk1iEU


Usabilla Search Query

  • Missing product images
  • Appropriate refiners
  • Search not returning proper results (2)
  • Search range can limit results
  • Show size in product description (2)
  • Product results limited before login
  • Search by model number
  • Boxed items should show quantities

The search bar is used significantly more than category navigation (84% vs 16%)

All Search Types

36 - Customers are unable to find searched products that are hidden behind an account log in

18 - Customers are not finding (all) products that match a searched keyword input

15 - Customers have a difficult experience finding products because of missing or inaccurate data (product descriptions, images, etc.)

12 - Customers have a difficult experience finding products when searching for specific sizes

6 - Customers are not able to find products by searching a manufacturer's part number

5 - Customers are unable to narrow products within a desired price

4 - Refiners don't provide a consistent experience

3 - Search type-ahead and auto-fill are trumping customers search input

2 - Customers have a difficult experience finding products because of site slowness

2 - Customers have a hard time finding a specific product (size, material) when it's grouped in a family

103 - Total

Search by Category Navigation

10 - Customers are unable to find searched products that are hidden behind an account log in

6 - Customers have a difficult experience finding products because of missing or inaccurate data (product descriptions, images, etc.)

5 - Customers are not finding (all) products that match a searched keyword input

4 - Refiners don't provide a consistent experience

4 - Customers have a difficult experience finding products when searching for specific sizes

4 - Customers are unable to narrow products within a desired price

1 - Customers have a hard time finding a specific product (size, material) when it's grouped in a family

1 - Customers are not able to find products by searching a manufacturer's part number

1 - Search type-ahead and auto-fill are trumping customers search input

1 - Customers have a difficult experience finding products because of site slowness

37 - Total

Search by Search Bar

26 - Customers are unable to find searched products that are hidden behind an account log in

12 - Customers are not finding (all) products that match a searched keyword input

9 - Customers have a difficult experience finding products because of missing or inaccurate data (product descriptions, images, etc.)

8 - Customers have a difficult experience finding products when searching for specific sizes

5 - Customers are not able to find products by searching a manufacturer's part number

2 - Search type-ahead and auto-fill are trumping customers search input

1 - Customers have a hard time finding a specific product (size, material) when it's grouped in a family

1 - Customers are unable to narrow products within a desired price

1 - Customers have a difficult experience finding products because of site slowness

65 - Total



Competitive Analysis

How do our competitors solve the problem(s)

Home Depot

https://www.homedepot.com/b/Bath-Bathroom-Faucets-Bathroom-Sink-Faucets-Single-Handle-Bathroom-Sink-Faucets/N-5yc1vZbrg2?NCNI-5&searchRedirect=single+handle+bathroom+sink+faucet&semanticToken=200r200000+%3E++st%3A%7Bsingle+handle+bathroom+sink+faucet%7D%3Ast+cnn%3A%7B2%3A0%7D+cnr%3A%7B8%3A0%7D+cnb%3A%7B9%3A0%7D+ct%3A%7Bsingle+handle+bathroom+sink+faucet%7D%3Act+tgr%3A%7BNo+stage+info%7D+qu%3A%7Bsingle+handle+bathroom+sink+faucet%7D%3Aqu

Grainger

  • Suggestions
  • Product categories
  • Brands
  • Product results where you can directly add to cart

PLP/Search Results Page

  • Results are filterable
  • Previously Purchased
  • Categories
  • Availability
  • Brand
  • Price
  • Shop by
  • Ratings
  • Product specific filters
  • Sort by: top sellers, brands, price
  • Help us improve this page
  • Check boxes for filters
  • List view or table view
  • Can search individual filters
  • View more available after six options
  • View more at bottom


HD Supply

Search

  • Suggestions in different departments
  • Suggestions based on words
  • Recent searchers
  • Show recommended products
  • Add directly to cart

PLP/Search Results Page

  • Can Print and email from this page
  • Sort by: relevance, brand, price, top sellers, country of origin
  • List vs Grid view (default)
  • Compare feature
  • Relevant check box filters
  • Show more for filters are shown with five or more items listed
  • Most filters are collapsed
  • Has pagination
  • Has cartridges under the results: Customers who views items you browsed also viewed, related searches, related products

Motion Industries

Search

  • Product categories
  • Brands
  • Shows previously results
  • Product results where you can directly add to cart

PLP/Search Results Page

  • Can search within results
  • PLP has categories

Home Depot

PLP/Search Results Page

  • Sticky and expandable filter section


Best Buy

Search

  • Left column hover affects products shown in right hand column



Heuristic Evaluation

What are best practices? Are we following them?

See our guidelines page for a starting point on best practices.

Product List Findings

  1. Do types of products factor into what information is important? (eg. 'Size' for a fitting, 'Finish' or 'Center Size' for a faucet)
    • Yes, present the user with sufficient product information (category specific) to adequately assess a product’s suitability (to their unique needs), and enable the user to get an overview of the product list (i.e. their available options) and compare products of interest to one another. - Baymard
  2. What is the most important ↔ least important information?
    • In order to provide the user with direct insight into product aspects they’ve expressed an interest in, sites must dynamically include product specifications (i.e., the “product aspects”) based on the user’s applied filters and sorting (i.e., the “user’s interests”) - Baymard
  3. What product codes are important?
  4. What actions need to be available (Add to Cart, Add to List, Quick View Details, Change Finish/Variation, Compare, etc.)?
    • Common repetitive elements such as “Compare features”, “Wish list”, “Buy buttons”, and “Quick View” don’t contribute any unique insights about each product. These elements may therefore instead be displayed on hover or relegated to the product page, to reduce product list clutter and allow the user to focus on the product details that are unique to each list item. The first and foremost solution to this issue is to thoroughly evaluate which list item elements (text and buttons) are important to users in each specific context while keeping in mind that choosing one list item element will inevitably mean leaving out other elements. - Baymard
    • Quick Views are often symptom treatments for poor product list designs. The subjects were only appreciative of Quick Views in the instances where they solved a problem that was ultimately rooted in an inferior list item design. The key challenge is helping users 1) distinguish Quick Views from product pages, and 2) make sure they don’t accidentally open one when they wanted the other. - Baymard
      • Visually enhance the product page link on Quick View
      • Make the Quick View a gallery, not an intermediary product page
      • Don’t place the Quick View link in the middle of the thumbnail
  5. What availability messaging is important?
    • The anticipated delivery date, or delivery date range, should be provided to users directly in the product list. This allows users scanning the product list to see when they’ll actually be able to receive the product. Items that are truly unavailable (i.e., discontinued) should either be removed from product lists or it should be clear to users that the item isn’t available to purchase and heavily demoted in the product list. - Baymard
  6. What indicators are important? (Lead Law, Loyalty, My List, Purchase History, etc.)

Visual Findings

  1. How important are large product images?
    • Best practices: Users lean heavily on list item thumbnails during product finding for visual recognition, and tend to either ignore or get confused by list items without them. Thumbnails need to be large enough to provide users with sufficient visual information about each product. This is especially true when the product is visually driven (product aesthetics). - Baymard
  2. What preference do our users have of List view vs. Grid view?
    • Best practices: Grid View is better for visually-driven product verticals, while a List View is typically more appropriate for spec-driven ones. If the site’s catalog has a mix of the two types of product verticals, both view types should be implemented and (automatically) switched between depending on the user’s current product scope. - Baymard
  3. How many products should be 'in view' on screen?
    • Search should therefore load 25-75 products per “page” by default. For product categories, the default number of list items to load per “page” is 50-150 products (fetched in ~20-product sets using the dynamic loading technique described earlier), depending on the type of product being displayed. Aesthetically-driven product verticals should be closer to 150, while spec-driven ones should move toward the lower end of the range. - Baymard
    • To accomplish this with search a “Load more” button should be used, ideally with a dynamic number of results returned based on relevancy scores. On a category page A combination of a “Load more” button and lazy-loading should be used. Display 10–30 products on initial page load and then lazy-load another 10–30 products until reaching 50–100 products – then display a “Load more” button. Once clicked, another 10–30 products are loaded in, resuming the lazy-loading until the next 50–100 products have loaded in, at which point the “Load more” button once again appears. - Baymard



Discovery Meetings

UX Question Ideation

February 28, 2019

 Click here to see Assumption Chart

Assumption DescriptionCertaintyRiskImportanceFeedback
Users want results based on "their store"-2F

Users don't want a page refresh after a refiner selection-2Uncertain

Users want to add products to cart from PLP-2Bโš ๏ธ
Users want to sort by price-5Dโš ๏ธ
Users want to see purchase history in PLP-5Cโš ๏ธ
Users expect relevant refiners-5Bโš ๏ธ


Users need pricing on PLP-5Aโš ๏ธ
Users want to know what is immediately available (1-hour pickup)-1Bโš ๏ธ
Users expect to see accurate search results based on their query-4Aโš ๏ธ


Users need to see product attributes on search results-5Aโš ๏ธ
Users want availability in their search results-5Aโš ๏ธ
Users want their results auto corrected automatically "pepe" to "pipe"2Iโš ๏ธ
Users want to see all available product results1Hโš ๏ธ
Users want to add a product to cart from type-ahead3Hโš ๏ธ
Users want related items suggested in type-ahead4Gโš ๏ธ
Users want results tailored to their account role5Gโš ๏ธ
Users want results based on their KOB2Gโš ๏ธ
Users want type-ahead1Fโš ๏ธ


Users want to see search history2Fโš ๏ธ
Users know how to use type-ahead4Fโš ๏ธ
Users care to see family level results when searching without refinery keywords00

Users want to switch between list and grid3E

Users want conversational search results/task oriented "Fix faucet"5D

User want to search "slang" or "descriptive" works (Ex: Big Pipe)3D

User want product docs available on search results1D

User want personalized search results (My list matches)1D

User want visual cues for relevance2B

Interview w/ Product

March 11, 2019

  1. What does success look like for these projects? How will we measure it?
    1. Nicole: Success looks like (addressing customer problems)
      • Search history
        • Storing and tracking what customers have searched for in the past (previous search terms)
      • Extensibility
        • We want to design it so we can scale it with additional features in the future
      • Personalization
        • FY2020
        • Be able to personalize experience based on what you've searched for
      • Signal boosting
        • Machine learning
      • Dimensions
        • Brands
        • Categories
        • Products
    2. Noah: Success looks like
      1. Two users:
        1. Users trying to find a particular product to purchase
          1. They dont necessarily know what the product is, but they know what they're looking for (3/4" copper 90)
          2. Success for this user:
            1. to add the correct product to their cart as quickly as possible
              1. need to be able to validate that product is correct
        2. Users doing research, may be looking for specs
          1. Maybe looking for a 'toilet' or a 'sink'
  2. What are the main things we know about Type Ahead/PLP that you recommend we address in this project?
  3. Who are our target user types and what are the specific tasks we want to design around/take into account?
  4. Is there anything else we should know before we begin research?
    1. Merchandisers will know how customers search
    2. Customer KOBs search differently
      1. HVAC
        1. searches by SKUs almost as a rule
        2. Use wild card
      2. FFS uses fee codes

Type Ahead

  • Nicole: Maybe not business a problem, but industry standards problem
  • Nicole: Type Ahead problem statement: Typeahead Redesign Problem Statement
  • Matt: We want to start with industry standards, but filter that through our customer's expectations (some of our customers know our SKUs)

Discussion w/ Nicole & Jessica on Type Ahead Personalization

March 13, 2019

  • Customers are organized into 'profiles' by
    • KOB segments, and
    • Region
    • Products are often buried based on a customer's region (relevancy ranking)
    • Segementing effects price column
  • What impact does Personalization have on Type Ahead?
    • Fire & Fab
      • Never wants to see a toilet or faucet or fans, or lighting in their search
        • Ex: We would want to bury top-level Plumbing results, it wouldn't appear to, say, page 13
      • Only want to see valves, pipes and fittings
      • Collection is excluded from Type Ahead, but Collections are something F&F use extensively
        • They buy a line or products, and products are often organized by 'class'
      • Fire-lock is the #1 collection for F&F
  • Navigation ranks by of results returned

  • Someone searches for 'schedule c' but you know they are a plumber, so you are going to want to return plastic pipe and tubing that fits that

Type-Ahead Discovery

March 20, 2019

PLP Lucidchart

UX Working Session Meeting Notes

March 27, 2019

Attendees: Matt, Farid, Lina, Dan, Cibonay, Laurie, Nicole, Noah, Percy, Liz, Tim, Drew

Type Ahead

  1. Main Search vs Sticky Search
    1. The sticky nav currently does persist the query. However, it does not mirror the main search, resulting in someone typing "faucet" in the sticky search and then typing "chrome" in the main search, they will not be searching for "chrome faucet". Suggestion for the search query 
  2. Liz: As you get more into the personalization you are going to need the personalization for the different types of KOBs.
  3. Make search query persist - (1, #10)
    1. Note: The sticky nav currently does persist the query. However, it does not mirror the main search, resulting in someone typing "faucet" in the sticky search and then typing "chrome" in the main search, they will not be searching for "chrome faucet"
  4. Account for typos - (1)
    1. Additional Suggestion: Make suggested search queries for autocomplete, where hovering over suggestions repopulates the input field with that query.
  5. Account for synonyms - (1)
    1. Note: Autocomplete does not have synonyms — regular search does have synonyms.
  6. Only show products if query has a sku, or is low-level category search
    1. Note: Ranking or relevance order should be looked at when autocomplete suggestions appear. Currently they are not available.

PLP

  • Noah: The list view is a familiar view for our customers.

Search / Type Ahead Recommendations Working List

  1. Suggested Search queries and not just auto-results (Categories/Products)โ€‹
  2. Better sorting on the list items displayedโ€‹
  3. Account for synonymsโ€‹ and typos
  4. Only show products if query has a SKU, or is low-level category searchโ€‹
  5. Improve relevancy to include search/purchase history / KOB
    1. Limited to the cookie session
  6. Make search query persistโ€‹
    1. Both main and sticky nav
  7. Prioritize the search suggestions, not query (bolding)

Out of Scope

  1. Cirrus Signal-boosting + Endeca (Phase 10)โ€‹

PLP Recommendations Working List

  1. Pull 'In My Store' out of search range box, and into its own clear/easily identifiable filterโ€‹
  2. Show Grid/List based on rulesโ€‹
  3. Add search suggestionsโ€‹
  4. Default to a SKU-level presentation over a family presentation
  5. Display the variation information (if color/finish, show thumbnails, if sizes, TBD), 'More Options' not enoughโ€‹
  6. 'Load More' over paginationโ€‹
  7. Display category-specific attributesโ€‹
  8. Remove indicators unless the user is filtering on it (ex: Lead law)โ€‹
  9. Are Documents Relevant on PLP?โ€‹
  10. Consider mouse over on product list items for additional information or images, discuss as an alternative to quick viewโ€‹
  11. Remove 'Compare'โ€‹
  12. Add purchase history indicatorsโ€‹
  13. Create "in-cart" indicatorโ€‹
  14. Add gallery for products that have important, multiple imagesโ€‹
  15. Minimize relevancy of product codeโ€‹
  16. Personalize / Using Oracle to understand how KOB behavesโ€‹
  17. Hide more than 10 filters in a listโ€‹
  18. Remove page refresh on filteringโ€‹
  19. What experience do we want to focus on?โ€‹
  20. Add previously purchased as a filterโ€‹
  21. Filter by status: Available / In My Storeโ€‹
  22. Faceted Search
  23. Consider category-specific sorting
  24. Consider fixed filters
  25. Consider graphical filters where appropriate
  26. Selections and Filters should match (bug)
  27. Enhance Inline-banner
  28. Display contextual search snippetsโ€‹

Search Results

  1. Show Grid/List based on product type (1, 2, 3)
    1. Note: Liz said that some products will have accessories incorporated within the category
  2. Add search suggestions (1)
    1. Liz: Make sure we give suggestions relevant to that user's KOB profile
  3. 'Load More' over pagination (1)
    1. Consider auto-loading next page/set of results
    2. Note: No real preference either way but agreed to move forward with it
  4. SKIPPED TEMP: Consolidate items into family whenever possible (1, #1)

List Items

  1. Display the variation information (if color/finish, show thumbnails, if sizes, TBD), 'More Options' not sufficient (123, #1)
    1. Note: We were in agreement to move forward
  2. Display category-specific attributes
    1. Note:  
    2. Remove indicators unless the user is filtering on it (ex: Lead law)
    3. Show specific category attributes based on the user's filters
    4. #3 Customer Feedback
  3. Display contextual search snippets
    1. Matt: This is where we would add additional information on the search results
  4. Remove Documents
    1. Matt: Documents were originally added because customers valued looking up warranties. 
  5. Remove 'Compare'
    1. Note
  6. Add purchase history indicators
    1. Note
  7. Create "in-cart" indicator
    1. Note: Suggestion to look at this feature not only on PLP but also on PDP
  8. Add gallery for products that have important, multiple images
    1. Note: As long as it doesn't add to the page height
  9. Consider mouse over on product list items for additional information or images, discuss as an alternative to quick view
    1. Note
  10. Minimize relevancy of product code
    1. Note: Tim asked for us to consider what elements can be personalized for our KOBs as we solution
    2. Type-ahead HVAC

Filtering

  1. Pull 'In My Store' out of search range box, and into its own clear/easily identifiable filter (1, #4)
    1. Note: Concern for removing the top search range box. Tim likes the idea on concept but concerned with going to "All Products" as there are regional products that need to be filtered as well.
    2. Note: When filtering by "In My Store", consider having remaining products not in your store communicated to the user so they know there are more products available that match their search query.
    3. Note: Have "In My Store" selected setting stay active throughout the site.
  2. Hide more than 10 filters in a list (#2)
    1. Note:
  3. Remove page refresh on filtering (#6)
    1. Note: Noah said that removing a page refresh is a huge win. We need to talk to Isaiah to see if there are technical restrictions on this.
  4. Add previously purchased as a filter (#2)
    1. Note:
  5. Filter by status (#3)
    1. Note

Sorting

  • Consider category-specific sorting
    • Note


Research Plan - Working Session

4/11/19 9:00 AM

Attendees: Matthew EvertFarid Mokraoui (Deactivated)Dan FlynnDrew PalkoCibonay Russell (Deactivated)Lina Angel (Deactivated)

Goal: To prepare list of questions we want answered to begin the Define stage of design.

To understand how users search, and

To understand what they search for

Research Problem:

  • Understand how users find products (products findability).

Objectives:

  • Understand how our users are searching for products.
  • Identify why they are searching for products and the motivation behind it.
  • Identify what products they are searching for.

Questions:

  • What are they searching that is causing null search results?
  • Do they know the products that they are searching for? Are they knowledgeable about the product or just aware of the product?
  • How often are searches performed for yourself or for someone else?
  • Think about the last time you purchased a product for the first time, how did you search it for it?
  • What terms are being used?
  • How do different KOBs search?
  • What products do you search for that give you a hard time?
  • Can analytics measure the different paths users take during search?



Findings

  • Always Persist the User's Search Query on the Results Page - Baymard #346
  • Clear Search Settings if the Same Query Is Re-Submitted - Baymard #350

Search Range

List Layout

  • Number of Items
    • Prominently Display the Total Number of Results - Baymard #377 / Show Total Number of Items in the Currently Displayed Product List - Baymard #531
  • List Style
    • Use grid view for visually driven products, list view for spec-driven ones - Baymard #533
    • Thumbnail Sizes Should Reflect the User’s Need for Visual Product Information - Baymard #456
    • Auto-Switch the Results Layout to Match the User’s Query - Baymard #379
  • Pagination
    • Search Results Should Use a ‘Load More’ Approach - Baymard #381
      • Search should, therefore, load 25-75 products per “page” by default. For product categories, the default number of list items to load per “page” is 50-150 products (fetched in ~20-product sets using the dynamic loading technique described earlier), depending on the type of product being displayed. Aesthetically-driven product verticals should be closer to 150, while spec-driven ones should move toward the lower end of the range. - Baymard #502
      • To accomplish this with search a “Load more” button should be used, ideally with a dynamic number of results returned based on relevancy scores. On a category page, a combination of a “Load more” button and lazy-loading should be used. Display 10–30 products on initial page load and then lazy-load another 10–30 products until reaching 50–100 products – then display a “Load more” button. Once clicked, another 10–30 products are loaded in, resuming the lazy-loading until the next 50–100 products have loaded in, at which point the “Load more” button once again appears. - Baymard #501
  •  Loading
  • Additional Content

Sorting

Filtering

  • Have Faceted Search - Baymard #376
  • Categories in the Sidebar Should Be Selectable and Nested Visually - Baymard #468
  • Truncate Filters with 10+ Values - Baymard #470
    • Avoid displaying long lists of sizes that visually overwhelm the user and make them think too hard. (Validately Usability Test Findings)
  • Show Number of Matches for Each Filtering Value - Baymard #467
  • Hide Dead-End Filtering Values - Baymard #475
  • Style Filtering Values as Checkboxes and Links - Baymard #479
  • Use Caution With Drop-Down Filtering - Baymard #465
  • If You Have Multiple Product Statuses, Allow Users to Filter by Them All - Baymard #417
  • Allow User-Defined Ranges for All Numeric Filtering Values - Baymard #416

List Items

General

  • Provide the Right Amount of Product Information in a List Item - Baymard #440<<<
  • Include Universal and Category-specific attributes
    • Universal
      • (Price and... ) other key Universal Attributes are: product title or type, thumbnail, user ratings, and variations. For lists of search results, contextual search snippets would fall in this category too. - Baymard Article
  • Include Product Attributes Consistently Across List Items - Baymard #443
  • Combine SKUs Unless Visually Distinct - Baymard #421 <<<
  • Consider Indicating Sub-Values of Selected Hierarchical Filters - Baymard #485
  • Highlight Any Product Already in the User's Cart - Baymard #484
  • Avoid Repetitive List Item Elements - Baymard #530
  • Consider Showing Additional List Item Information on Hover - Baymard #526
  • Ensure Pathing Is Clear and Consistent on List Items - Baymard #552

Brand

  • Consider Including Logo Thumbnails When Brands Are of High Importance - Baymard #444

Image

  • Best practices: Users lean heavily on list item thumbnails during product finding for visual recognition and tend to either ignore or get confused by list items without them. Thumbnails need to be large enough to provide users with sufficient visual information about each product. This is especially true when the product is visually driven (product aesthetics). - Baymard #455
  • Product Thumbnails Should Match the Variation Searched For - Baymard #386
  • List items without thumbnails were often completely ignored, as most of the subjects perceived these products as “incomplete.” - Baymard Article

Title

  • Display the full title
  • Different Product Attributes in the List Item Should Be Visually Distinct - Baymard #448 <<< Webdisplaynames
  • When name, product type, number of items, materials, etc. are all presented as one long product “title” – users have a much more difficult time scanning the list items - Baymard Article

Codes

  • ???

Price

  • Display ‘Price Per Unit’ For Multi-Quantity Items - Baymard #446
  • Prices Should Be Easily Scannable - Baymard #452
  • The product price is a fundamental product attribute that ought to be included under all circumstances (and in the few cases where it can’t, some indication of why is typically warranted) - Baymard Article

Refiner

  • Always Indicate Essential Product Variations in the List Item - Baymard #450

Actions

  • Ability to Add to Cart (TBD by preliminary research)
  • Common repetitive elements such as “Compare features”, “Wishlist”, “Buy buttons”, and “Quick View” don’t contribute any unique insights about each product. These elements may therefore instead be displayed on hover or relegated to the product page, to reduce product list clutter and allow the user to focus on the product details that are unique to each list item. The first and foremost solution to this issue is to thoroughly evaluate which list item elements (text and buttons) are important to users in each specific context while keeping in mind that choosing one list item element will inevitably mean leaving out other elements. - Baymard #530
  • Quick Views are often symptom treatments for poor product list designs. The subjects were only appreciative of Quick Views in the instances where they solved a problem that was ultimately rooted in an inferior list item design. The key challenge is helping users 1) distinguish Quick Views from product pages, and 2) make sure they don’t accidentally open one when they wanted the other. - Baymard #529
    • Visually enhance the product page link on Quick View
    • Make the Quick View a gallery, not an intermediary product page
    • Don’t place the Quick View link in the middle of the thumbnail
  • Ensure Pathing Is Clear and Consistent on List Items - Baymard #552
  • Avoid Quick Views - Baymard #529

Availability

  • The anticipated delivery date, or delivery date range, should be provided to users directly in the product list. This allows users scanning the product list to see when they’ll actually be able to receive the product. Items that are truly unavailable (i.e., discontinued) should either be removed from product lists or it should be clear to users that the item isn’t available to purchase and heavily demoted in the product list. - Baymard #528
  • Display Accurate Availability & Delivery Indicators - Baymard #528 <<< Discontinued, Out of Stock Indicators

Description

  • Display Contextual Search Snippets - Baymard #382
  • Include List Item Information Based on Applied Filters and Sorting - Baymard #480

Documents

  • Remove documents as info element (TBD)

Indicators

  • What indicators are important? (Lead Law, Loyalty, My List, Purchase History, etc.)
  • Consider Sparsely Highlighting Unique Purchasing Attributes - Baymard #447

Specifications

  • In order to provide the user with direct insight into product aspects they’ve expressed an interest in, sites must dynamically include product specifications (i.e., the “product aspects”) based on the user’s applied filters and sorting (i.e., the “user’s interests”) - Baymard #480
  • Contextual search result information - Baymard #382
  • Numeric Product Specs Need Proper Labels and Units - Baymard #449

Misc

  • Update elements to UI Kit standards

Search Range

Validately Study on Search Range

https://validately.com/report/837dbcd9-a59a-4a07-8e70-bd0c739f5bb0

Top Problems
  1. Did not self-discover the Search Range tool. Found it only because the test questions prompted them to keep looking. (3 of 5 testers)
  2. Did not see the yellow bar explaining that no products were found nearby, so all products were displaying. (3 of 5 testers)
  3. Never found the Search Range - said he would go to the Locations Directory to find products nearby (1 of 5 testers)

Jira Tickets

https://ferguson.atlassian.net/browse/FET-628

Firstly, we are using sku.nonstock_stores & sku.stock_stores fields, using like this:
For example, for sku ABC:

  • sku.nonstock_stores = 123
  • sku.stock_stores = 456

Then we consider sku ABC is InStock in store 456, and NonStock in store 123.

And for the three Search Ranges:

  • Nearby Stores: show the products from the stores which associated with the selected branch in Job Board, and show both InStock & NonStock products.
  • In My Store: show the skus from the selected store in Job Board, only show InStock skus, and haven't calculated whether the InStock count is greater than 0.
  • All Products: All the catalog products should show, both InStock & NonStock products.

https://ferguson.atlassian.net/browse/FET-1247
https://ferguson.atlassian.net/browse/FET-525
https://ferguson.atlassian.net/browse/FET-537
https://ferguson.atlassian.net/browse/FDC-3098



Percy's PLP Pain Points

  1. Product List start too far down the page

    1. Depending on browser setting the user can only see one row of products if any at all

  2. The user cannot tell the difference between a family or sku

    1. This can lead to the user not knowing that some options are a group of products

    2. The ONLY indication is “more options shown” in tiny lettering in the corner

  3. Web display names are too long

    1. This leads to important information about the product being cut off

  4. Families do not show what they are differentiated by

    1. The user does not know if these products are grouped by size, color, material, etc unless they click into the product

  5. It is not clear to the user that they are signed in

    1. This can lead to the user thinking that we do not carry a product instead of it just being behind login

  6. The format should default be in list view and not column view

    1. Important information about the product should also be shown in extra space

  7. Most important refiners are not at the top of the refiner option

    1. The customer is less likely to go too far down to hunt them down

  8. Important refiners default to collapsed view

    1. Users have no idea how the can refine until the clink into it

  9. Refiner values can take up more than one line if they are too long

    1. This takes up valuable space where another refiner is needed

  10. Refiner lists are too long

    1. This causes choice paralysis for the user

  11. Search range taking up important space that should have products

    1. This should be moved to the side column above refiners

  12. A lot of empty space to the right and left

Sorting

https://ferguson.atlassian.net/browse/FET-534



Further Reading

https://baymard.com/blog/list-item-design-ecommerce

46% of e-commerce site have severe usability issues because they display too few product attributes in the product listings.

  1. 64% of e-commerce sites impede users ability to compare the product in their list due to inconsistencies in the listed product attributes.
  2. 50% of e-commerce sites make it needlessly difficult to scan and read the included product attributes, because it’s all just included as one long “product title”, cramming 2-5 lines full of technical specs.
  3. 76% of sites neglect the principles of progressive disclosure, and always display the same amount of information for all product list items, instead of progressively showing extra information for items the user has shown an interest in.

A prerequisite to optimizing how list item information is displayed is to even have the correct information available in the first place.

Research studies have shown that there are two general groups of attributes that should be included in all list items: Universal Attributes and Category-Specific Attributes.

  1. Universal Attributes are essential to virtually all products throughout the site and include: Price (and Price Per Unit for multi-quantity units), Product thumbnails, Product title or type, Product variations (colors, sizes, materials, capacity), and User ratings.
  2. Category-Specific Attributes are meanwhile unique to each product vertical and can therefore vary from category to category. For example, category-specific specs for toy products can be “Age Rating”, while for smartphones it might be specs like “Screen Size” and “Storage Capacity”, etc.

Including all applicable Universal Attributes and 1-3 Category-Specific Attributes listed for each product list item, was found to strike the best balance between providing the user with just enough information to accurately evaluate and compare the listed products, while at the same time avoiding information overload and intimidating product listings.

The list item is a balancing act: at one hand users need enough information to to be able to make an informed decision about which products to open and which to skip – but at the other hand users need to establish an overview of the products available to them. So while too little information can be highly problematic, so can too much. (Display too little information is the most common issue.)

During testing we’ve observed a simple technique to perform very well for mitigating this list item balancing dilemma: utilizing the mouse hover state to display additional information or images. During our usability study of 19 leading e-commerce sites, this extra layer of hover-enabled textual or visual information was observed to drastically reduce wasteful “pogo-sticking”, where the user heads back and forth between the product list and product pages only to discard irrelevant products.

It is however vital that only secondary product attributes are relegated to a hover-state since users will not be able to compare these specs directly (and won’t be available on touch devices without additional implementation). This should therefore be considered a bonus layer of information provided to the user to help them better determine which of the product list items that initially sparked their interest they should pursue further.

In essence the findings can be summarized quite simply: First, e-commerce sites must ensure their product list items include all Essential Attributes (price, thumbnail, product title or type, variations, and user ratings) along with 1-3 Category-Specific Attributes. Second, with the right list item information in place, e-commerce sites must ensure that the design of this informationallows for:

  1. Comparison through Information Consistency – both in terms of what information is included and the sequence in which it is displayed.
  2. Scannability through Separate Entities – avoid embedding specs in titles; instead style attributes as separate entities.
  3. Overview through Progressive Disclosure – consider utilizing mouse hover to show additional (secondary) item information and possible a new thumbnail (e.g. different product angle).


https://baymard.com/blog/product-listing-information

Users select and reject products based upon the information available about those items in the product list. It, therefore, isn’t all that surprising that during our large scale usability study of Product Lists & Filtering we found poorly selected list item information to be one of the most severe usability issues related to product list navigation.

During testing, we observe that too little or low-relevance list item information is highly problematic because users are unable to adequately assess the listed products when they lack essential information about those items. This led the test subjects to dismiss perfectly relevant products and forced them into unnecessary pogo-sticking between product list and product page – where they had to go back and forth, opening the product page of every single product listing just to learn the very most basic and core attributes of that product; a tiresome exercise that often resulted in the subjects abandoning the site as the friction in locating relevant products simply became too high. Clearly, displaying the right amount and the right kind of information in each list item is paramount to improving the user’s product finding experience.

Achieving a good signal-to-noise ratio in the list item information is key to improving the user’s ability to find the products they are looking for – yet it clearly isn’t easy to achieve. It requires considered curation of the product attributes displayed.

Product list real estate shouldn’t be squandered. Each and every element in the list item should be carefully chosen and presented to give users the best possible conditions for finding and choosing the products that are just right for them. A successful product list design must fulfill two requirements:

  • Present the user with sufficient product information to adequately assess a product’s suitability (to their unique needs and desires),
  • Enable the user to get an overview of the product list as a whole (i.e. the options available to them), and compare products of interest to one another.

The first requires sufficient product information to be shown for each product while the latter requires a sufficient amount of products being displayed on the user’s screen. This creates a dilemma, as having too much product information will take up real estate resulting in less products per page, which was found to discourage product list exploration during testing. Yet having too many products will make it difficult to display adequate information in each list item, resulting in extensive pogo-sticking.

Product list design therefore isn’t so much a matter of “less is more” as it is one of “just enough is more.” Figuring out which product attributes should be present in the list item and which shouldn’t is naturally highly site-specific and, as such, there are no hard-set rules. A good list item helps the user with just enough information to accurately evaluate which of the items in the list are of relevance to them, and – just as importantly – which items can be skipped. The list item information should furthermore enable the user to compare the relevant items to one another. It’s essentially an exercise in maximizing information scent and product comparability without bloating the product list.

Now, from our research studies we have found that there are two general groups of attributes that should be included in list items: Universal Attributes and Category-Specific Attributes.

Product variations, such as different colors, sizes, materials, looks, etc., is another essential attribute to indicate directly in the list item. Without them, the subjects often rejected products that actually matched what they were looking for simply because they couldn’t see that multiple variations of the item were available and therefore only judged the product on the default variation displayed, without ever realizing that multiple (highly relevant) product variations existed and would be available from the product page.

It’s not all product variations that should be displayed, however. For example, a desk available in different sizes would typically need an indication in the list item explaining that multiple sizes exist, whereas it’s probably less relevant for a shoe list item to indicate that multiple sizes exist as users will assume so due to the nature of the product.

The relevance of products in certain technical categories will be determined almost entirely by the product’s compatibility with another product. This “compatibility information” should therefore be included in the list items so users can identify the products that are relevant to them without having to open the product page of every single item in the list. Taking things a step further, if the user already has the compatibility-dependent product in their cart, the product list items should ideally state whether they are compatible with the item in the user’s cart.

Category-specific attributes are a smart way to provide 1-3 extra pieces of product information that is uniquely relevant to that product’s vertical and thus uniquely helpful to the user in determining whether the item is of relevance to them or not.

E-Commerce User Experience - Vol. 2: Homepages and Category Pages

Product listing pages need to clearly show users what’s available. If users aren’t interested based on a product listing page, they’ll never view the product page. Users need to be able to easily identify and understand the product. Clear, concise product names that include important characteristics and clear product images that are large enough to easily identify the product are crucial. If customers know what they're looking for, they should be able to recognize the product from the text or graphics on its product listing page. If customers aren't sure of what they want, the product list should supply enough information for them to decide which items to look at more closely. When sites provided clear product names and images, some users who already knew what they wanted to buy were able to add items to the cart without bothering to view full product descriptions. 


Comparing Products

The best way to allow product comparison is to provide comparable information about similar products on the site. Sites which provided different details about similar products confused and frustrated users who were trying to decipher the differences among their options. However, providing comparable information on product pages still requires users to jump back and forth between product pages in order to make comparisons. Some users took notes to handle such situations, and others maneuvered to open multiple screens, windows or tabs in order to view products side by side or quickly switch views between products. Good comparison tools let users quickly see the differences between products, lining up information in a table. Successful comparison tools allowed users to quickly and easily select items for comparison, presented equivalent information about each item, and allowed users to remove items, add items to the cart, or return to product detail pages from the tool.

34. Use short, descriptive product names on product listing pages
35. List the most important information first in product names
36. Show prices
37. Indicate if styles, colors or options other than those pictured are available.
39. Show availability information on product listing pages. Users should know as soon as possible if an item is available, back-ordered, sold out, or unavailable.
41. Make it clear if multiple pages of products are available.
42. Provide a clear link to additional product information from product listing pages, particularly if the page offers Buy buttons.

48. Use download time to help determine the best number of products to display per page.

54. Show visited links.

55. Indicate how many products are available across how many pages, and how many items are listed per page.

57. Indicate which page users are currently on as well as the total number of pages.

60. Allow users to sort in both directions (for instance, low to high and high to low) as appropriate.

61. Remove sorting tools if only one item is shown.

63. Provide a way to compare the details of similar items.

Type Ahead

Autocomplete Introduction & Benchmark Analysis

https://baymard.com/premium/topics/257

Also known as predictive search

Having autocomplete suggestions appear as users type their search query isn’t about speeding up the typing process. That’s a relatively small side benefit. The true value and impact on the user’s search experience comes from how autocomplete suggestions can assist and guide users toward better search queries. When autocomplete suggestions are done well they inspire users about the types of queries to use, teach them correct domain terminology, help them avoid typos, and assist them in selecting the right scope to search within.

Autocomplete query suggestions functionality is found at 90% of the e-commerce sites and should be considered a “web convention” for search fields.

Webinar Video

When users encounter an autocomplete dropdown, they usually perceive the autocomplete suggestions as personal suggestions to them specifically and generally trust the site to know better. Because of this, autocompletes can be very powerful.

Autocomplete offers 4 key components: reassurance, exactitude, spelling, and guidance.

  • Reassure: This is a site where you'll find what you need
  • Exactitude: If user doesn't know the exact reference/code the auto-complete will show them more options 
  • Spelling: Typos corrections / suggestions
    9% of all users have typos in their search queries. (ex: is it "macbook" or "mac book"?) Autocomplete immediately resolves those common low-level typos, especially phoenetic typos.
    Offering guidance via domain terminology. (ex: it's not a "macbook power charger" it's a "macbook power adapter")
  • Guidance:
    • Autocomplete slows down users - but that's a good thing! It increases search success rate.
    • One of the issues is that it alters what the user searches for, so there are some pitfalls:
      • Autocomplete has to first and foremost suggest search queries, not only products or only categories
    • Before we start trying to fix the Type-ahead we need to understand the actual behavior (system logic):
      • Mistake to take search logs directly for autocomplete suggestions datasets without cleaning it first.
    • When users perceive the autocomplete is inaccurate, it becomes noise that they don't use, and we lose all the benefits

Not only suggest queries & scopes like Amazon

Auxiliary data should be styled differently than query suggestions

Show less than 10 suggestions (Hicks & Mille's law) - If you have a scroll in the suggestions results then something it's "wrong"

Highlight differences, not what users typed "Camera" they know they need a camera, so highlight the options suggested instead of the word "Camera"

Padding between options (Hit the right option) and with the "environment"

Auto-results is different than auto-complete (Don't show products only)

Never Mix Auto-results and auto-suggestions (DIFFERENCE BETWEEN PLP & TYPE-AHEAD/AUTO-COMPLETE): If users click goes to a PDP or PLP? Visually differentiate Product Suggestions from Query Suggestions.


Type-Ahead Notes (Move this)

  • We don't appear to be assisting via typos or domain terminology synonyms
  • If we're going to show products, it has to be because the search query matches at a very-low category level.
  • Avoid pluralizations, duplicate, and wrong suggestions



Decibel Insight

www.ferguson.com/category

All Products CATEGORIES Bathroom Plumbing ing Janitorial Hangers, Struts & Fasteners MORE CATEGORIES Safety & Security Paint, Equipment & Supplies Building Supplies You CUSTOMER SERVICE Contact Us Locations Returns and Cancellations Warranty Info SHOP BY ROOM Kitchen Plumbing gs Heating & Cooling Lighting & Fans Plumbing Parts & Supplies Appliances Hardware Water Heaters s Tools (4491) C] Victaulic (4392) C] Zurn (2901) C] American Standard (2789) Show More Refinements Need Help? For help with orders or products call (757) 312-9304. For technical help, call: 888-222-1785. CORPORATE INFORMATION About Us Careers Ferguson Cares Press Room Social Media Ferguson plc Information Gauges & Thermometers Casting & Cast Iron Boxes, Frames & Covers Irrigation & Drainage Home Décor Electrical Adhesives & Sealants Sign up to receive email about product iaunches, events, deais & more! RESOURCES Brands We Carry Corporate Freight Ideas & Learning Center Legislation Center Supplier Diversity Trade Talk Vendor Relations Video Library Website Information Website Tutorials FERGUSON BRANDS FNW Valve@ Mirabelle@ Monogram Brass@ PROFLSQ:J. PROSELECT@ RAPTOR@ Product Westcre Clicks Min 1 Copyright 0 2018 Ferguson Enterprises Inc. All rights reserved. Terms of Site Use I Terms of Sale I Privacy & Security I Sitemap Max 134 " style="width: 3.5416in; height: 6.5104in;">

  • Duration: 1m 17s
  • Click/touch Hesitation: 44.2s
  • Mouse Distance: 3,792 px
  • Scroll Distance: 3,070 px

Validately

Find-ability: Pipe Fittings

https://validately.com/report/e43a1a15-93db-4fc3-89a8-eacf98eaa92c

Top Problems

  • 5 of 5 users found it easy to search for copper tees and elbows using the search, but were confused by the "3/4 x 3/4 x 1/2" size requirement, sometimes settling for "3/4" or "3/4 x 1/2 x 3/4".
  • 4 of 5 were slowed (and sometimes overwhelmed) by the number of sizes in the filters and on the PDP.
  • 4 of 5 users easily browsed to the correct product families by choosing "Pipe Fittings" as their first category. The one who started browsing in "Plumbing" spent 8 frustrating minutes trying to find copper tees and eventually gave up by choosing a tee in another material.
  • 1 of 5 was frustrated by missing images on the results page.

Usability Insights & Recommendations

  1. Avoid displaying long lists of sizes that visually overwhelm the user and make them think too hard.
  2. For products with many sizes and/or multiple dimensions, consider a different selection method that steps the user through the choices and/or helps them to focus on one choice at a time.
  3. Don't make the user repeat a complicated choice. Example: These testers worked hard to find the exact size for copper tees using the filters, but still had to scan a list of 150+ options when they got to the PDP.
  4. Create a path to Pipe Fittings from the main Plumbing category
  5. Place an ADD TO CART option at the top of the PDP when the user has searched for a specific size or has narrowed the search results with filters. Or, open the "Buying Options" section of the PDP to reveal the ADD TO CART option for that specific size.
  6. Consult site analytics for the effect of line items vs. product quantity display in the cart icon on all customers vs customers with large orders.
  7. Create an aid to help non-trade pros to understand how fittings are sized

Water Heaters

https://validately.com/report/5e357a3b-acf0-4d1a-b37f-6bf6edf16544

  • Filters appear to be inconsistent between search and browse paths when users do not explore "correct" categories for their product and fail to get to the lowest level of the taxonomy. How can we help users who don't see the filters at all (or don't get to the necessary level)? Consistency of labels in the filters is also needed
  • One tester had great difficulty saving her tankless and conventional water heater selections to show her customer, because she expected "Favorites" or "My List" on the Search Results. She repeatedly clicked the grid/list view icon to save the product, never understanding what that icon did.

Find HVAC Repair Parts

https://validately.com/report/b22a7e5c-c254-4862-880e-b798da3e3ca4

Product Names Are Crucial for Find-ability

  1. 3 of 5 testers expected to see defining product attributes in the product name so they could choose accurately from the Results. They did not look for refiners immediately, but clicked back and forth to product pages.
  2. 5 of 5 testers experienced inconsistent product naming conventions, which made their search harder.

Suggested Search Delights Users - or not

  • Users who notice the suggested search terms are excited by it's potential. They love it when it gives them good results, but are frustrated when it doesn't.

Browsing for HVAC Tape Is Very Difficult

  • 5 of 5 testers could not browse the categories to find HVAC tape, because they did not see or understand the Adhesives or Accessories categories.
  • 2 of 5 testers who persevered beyond reasonable expectations did find the Adhesives or Accessories category, but they settled for duct tape instead of metal HVAC tape, because they couldn't filter by material.

Filters are helpful when seen

  • Most users are focused on the content area of the Results page. When they realize filters are available, they find them helpful. For the HVAC tape product, filters for width, length, and material were only available if they used the Adhesives category. Searching "hvac tape" offers only a Brand filter.

Usability Insights & Recommendations

  1. Within a product category, use consistent attributes in the product names
  2. Within a product category, offer consistent filters whether the user searched or browsed to the category. Update: Tim Indoe explained that the filters are consistent, but only available at the lowest level; the users who didn't path through the correct subcategories did not see the lowest level and most helpful filters.
  3. Keep improving suggested search terms - they have great potential to delight the user and help them accomplish their tasks.
  4. Edit the "Adhesives" category in the taxonomy and on the landing page to say "Adhesives & Tapes" for better recognition.
  5. Check the code for display of checkboxes in the AMPS and HORSEPOWER filters for condenser motors - the selector boxes did not display until the user clicked the word, such as "1.9 amp"

Find HVAC Installation Supplies

https://validately.com/report/a844b7f0-90a8-4ca3-ad73-5ff37eb6ee8f

Product Names & Filters

  • Searching and browsing is harder when looking for items with more than 2 dimensions. Many product names have room for only 2 dimensions and they are not always presented in the same relative position, which makes scanning the search results difficult.
  • The experience suffers when filters do not display or are missing options. Example: there are many 25-ft line sets in the catalog, but searching "copper lineset" gives incomplete options for length and line diameters

Spacing Matters

  • Size formatting matters - associate Laura Fields will be examining this example and others as she works on a size-formatting normalization process:
  • Example: filters show or don’t show based search box input. In this video clip Respondent 793984 searched "copper lineset 25ft" but did not see any 25-foot examples.

Offer Contextual Help for Refining Search Results

  • In this and other find-ability tests, we observe that most users focus only on the content area of the Search Results, scanning product names to find what they need. They don't look to the left for filters, or look there only after difficulty finding what they want.
  • However, after the first page scroll, most users will look at the top of the content area for help refining their product list if they don't see what they need. Consider other ways to help the user refine their search within the product listing area:
  1. Offer content-specific help if the user remains on the search results more than x seconds or pages through more than x pages of results. This could take the form of a message directing their attention to the existing filters.
  2. Create a new method of filtering within the results. Examples: search within results; let them input product dimensions in a form; sort by size or other attributes.
  3. If the user stays on the results pages for more than x seconds, trigger a Chat session with a product specialist.

Cirrus 10

Problem Themes

  • Support different ways users search for part numbers 

o    Partial Numbers: For example, if a part number is ABC123456, they will only type ABC123. 

o    Special Character vs One Word: For example 123-456, customers search for them as 123456.

  • Accessories are ranking too high
  • Searches related to measurements

o    The difference in the units of measure 

o    Variations in fractions 

o    Handling Fittings 

  • Problem with Diffuse Categories 

o    This problem occurs when categories are comprised of two different types of things

  • There’s a gap between user vocabulary when they search and the product catalog vocabulary
  • Products with images are not prioritized
  • Prioritize products that are of Ferguson Brand 

Existing Settings 

Search Controller Behavior 

  • The search controller is a piece of code/component that handles logic and querying of the search engine. When the customer submits a search term to get a search results page, the following actions and the logic is performed by the search controller. 
  1. A query is sent to Endeca using the ProductCode search interface. This search interface only covers product code/product number related fields. 
  2. If there’s an exact match from the ProductCode search interface, the user is redirected to the Product Page of the matching product found in Endeca. 
  3. If there’s no match in the ProductCode search interface, another query is sent to the “All” search interface. This search interface covers a lot more fields to search. 
  4. If there are results, then the search results page is rendered with the results. 
  5. If there are no matches from “All” search interface, the user is redirected to the zero results page. 

Solution Design

Concept Weighted Relevance (CWR)

Relevance Design Pattern 
  • Category – what a product is. 
  • Brand – whose name is on the product. 
  • Use – what the product is used for. 
  • Style – what the product is like. 
  • Options – the options such as color and size that can be specified at time of order. 
  • Precision Keywords – meta keywords that have been carefully selected to further describe the characteristics of the product. 
  • Name – the name of the product. 
  • Recall Keywords – meta keywords that have been selected to cast as wide a net as possible for concepts that are core to sales and marketing goals. 

Solutions for Part Number Searches 

  • Customers sometimes search in forms that are not exactly the same form as the one in the catalog. Here are variations: 
  1. Customers submit only the partial form of the part number.
  2. Some part number have hyphens or forward slashes, but users search for them without those characters and as a single alphanumeric word. 

Solutions for Measurement Searches 

  • Create equivalencies between these variations via thesaurus entries. 
  • Consider treating symbols for measurement to be searchable characters instead of Endeca ignoring them by default. However, evaluate if doing this has any negative impact. Consider these two characters:

o    “ stands for inch in 1” 

o    / used as part of fraction measurements like 1/2 inch 

Solutions for Accessories ranking too high

  • Reliance on product names
  • Identifying accessories

o    Inspect the leaf category name (or even any node in the category path). Look for hints of these words:

  • Parts 
  • Accessories 
  • Create a new property

o    a. Property Name: isAccessory 

o    b. Value: 0 if not accessory, 1 if accessory. 

  • In the relevance ranking strategy, we use a Static Module

o    static(isAccessory, ascending) 


Previous notes/research:

02/15/19 Quarterly Review Nicole mentioned: 
Type-ahead is the “Dropdown window that shows when searching”.

Changes:

  • Search History
  • Frequently Purchased
  • Add to Cart Option

Some highlights wrote back on 08/2018 after some PLP and PDP researching:

Changes needed:

  • PDP unorganized information 
  • PDP multiple sizes for a product difficult to use
  • PLP Filters
  • PLP list

B2B SITE

Features:

  • PRODUCT CATALOG:
    • Product Detail with rich descriptions, tech specs and image zoom, Manufacturing models, shipping weight, country of origin, etc. 
    • Imagery consistent and high-quality
    • Use product suggestions / required / related
    • Right product in few clicks as possible (Find-ability)
    • Search and Navigation: SEO and product find-ability using codes, keywords or descriptions
    • Product Categories and Robust filters: Price, model, brand, etc.
    • Detailed Availability and quantity available
  • ACCOUNT ADMIN: Enabling their customers to become more and more efficient (e.g.: Add Users, Pay Invoices)
  • UX: Simplify the buying process for the best possible UX
  • HELP: FAQ Questions

Crucial components of a B2B PDP/PLP https://bintime.com/blog/b2b-product-positioning 

  • Title: Clear and as close as possible to the manufacturer’s descriptions
  • Product Positioning should be clearly mapped out in a usable grid
  • Product Information: Feature-based or benefit-based? Include all the info available
  • Strong CTA
  • Great UX (Avoid broken links or complicated navigation)

Motivations

  • A fast and convenient way to order the product or service
  • Plenty of information 
  • An easy way to get in touch for a quote
  • Simplify the buying process for the best possible UX

Tips

  • Use product suggestions
  • Use Keywords research to discover what the audience is looking for
  • Imagery consistent and high-quality
  • FAQ Questions
  • MAKE SURE YOUR USP is present! (Unique Selling Proposition)

What Best B2B have? https://b2xpartners.com/review-of-the-best-b2b-ecommerce-websites/

General:

  • Invest in Customer Experience
  • Large complex catalogs but get their customer to the right product in few clicks as possible
  • Product content rich and detailed
  • Account Management features are enabling their customers to become more and more efficient

Catalog:

  1. Search and Navigation: SEO and product find-ability using codes, keywords or descriptions
  2. Product Categories and Robust filters: Price, model, brand, etc.
  3. Product Detail with rich descriptions, tech specs and image zoom, Manufacturing models, shipping weight, country of origin, etc. 
  4. Detailed Availability and quantity available
  5. Related Products and Products other customer bought.

Features to Implement in a B2B website https://b2xpartners.com/the-secret-to-b2b-e-commerce/

“B2B Customers come to your website because they have to. It is their job.” 

  • Search
  • Navigation
  • Cross-sell
  • User Administration
  • Custom Contracts

LINKS: