Designing for discovery

Editor's Notes

• #2: travelmatch - startup\nchange how travel search works\nshare process & learnings\nwork in progress\nJonny, Rob\n
• #3: 70% of people know what they want to do on holidays, not where they want to go\n
• #4: travel sites structure\nusers need to fit discovery around the search technology\nwhere to? when? who?\npoorly designed forms, mandatory fields, poor results \nlittle discovery\n
• #5: Here’s the idea: \nWouldn’t it be awesome if you could find holidays that match your idea of a great time?\nWhat if you could combine any number of options that you are interested in?\n
• #6: Or perhaps a 5star holiday with infinity pool, golf and a beach bar. \nOr a 2-week break with babysitter, kid’s club, water slides no more than a two-hour flight away. \nThe sky’s the limit.\n
• #7: Core design principle:\nno mandatory information, but lots of equally valid choices.\nWhat we wanted to create is an online traveller agent who best serves users, not travel companies. \nWe want our customers to trust that we find them the best holiday possible.\n
• #8: Our second design principle:\nCompare prices and best value.\nDifferent people attach different values to different options. \nLet them decide what good value looks like.\nWe want to be transparent, and inspire trust that we actually find the best offers available from trusted travel agents.\n
• #9: early pitch deck\nknow scenario: 1+2+3+4\nEasily frustrated with a gazillion search windows open\nreading brochure text to try and find out if your key criteria are met\nchild care? golf course?\nTo design for discovery, not for existing technology structure, we...\nWe bring the fun back into finding that perfect holiday\n
• #10: Early concept illustration\nvisual drag + drop search\ncentered around postcard of dream holiday\n\n
• #11: Everyone loved the idea\nvery visual + appealing\nbuilt functioning prototype\nWe learned: while everyone loves this as a concept, difficult to use. \nno space for labelling \nChallenge to illustrate clearly\nThalasso Therapy.\n
• #12: secured funding\ndespite financial crisis\nshort hiatus while assembling a real team\nback end/ front end/ data/ natural language/ ux/ biz\n
• #13: start work in earnest. \nback to the drawing board. \nWe know the non-hierarchical search is a winner. \nwithout labeling too hard to use \nreal estate available on screen\n\nprototyped to create enough space + to make all choices available at all times\n\nsimultaneously give confidence that they know what to do with this very different way of searching. \n\nhow do users know how to open the search, how to use it know what they have searched for.\n
• #14: low-fi prototypes in user testing to address these questions. \nTo see how the postcard and search choices could work together with real time search results.\nCan people open it?\nDo they know what to do?\nDo they find what they want?\n\n
• #15: Search results: \nsummarize search/postcard\ngive plenty of contextual information to peak interest\nshow price\nshow pictures\nVariety of discovery routes:\ngive context of map\nvisual search alternative\n
• #16: To validate ideas and assumptions \n10 users in 1-2-1 research sessions\nWe learned:\nexcited about all the choice, \nthis new way of discovery was seriously challenging their mental model of how holiday search works. \nThey were looking for ‘the box’. \n\nWhat worked really well is all the contextual information about the holiday. \nThe reviews, maps, area guide, video, etc. \n\n
• #17: Along the way we identified a number of challenges\nwant to share these with you\n
• #18: We’re testing the capabilities of our Enterprise Search solution. \nWe are trying to break all the rules around travel search to create a fantastic customer experience from the outside in, not from the technology capabilities out.\n
• #19: Two words - realtime results\nnot ‘make a cuppa while you wait’ searches\nwe want results to update instantly on the same page\n
• #20: Getting people to the site\nSEO + SEM requirements\nNot always in line with customer experience goals, \ndespite Google’s notion that their raking reflects usability\n
• #21: If you want to give 150+ choices\nwhere do you display it all?\nHow does that work with aims for simplicity and ease of use?\nFirst time faced with trying to fit a search above the fold...\n768 height is still a majority requirement\nso you need to whittle down your requirements\n
• #22: startup challenge - the board\nfell in love with the simple pictorial search proof of concept\nhard to convince to let the idea develop\nnot always convinced of best practices\n
• #23: and finally, not to be underestimated\nstartups need to get their product to market \nif they want to stand a chance\ntime scales are tight\nwe’ve developed a well rounded eco system, but can only build the bare bones in time for a first launch\nWe are in soft launch right now, with weekly sprint releases\nGetting ready for a proper launch in Jan\n
• #24: What did we do?\nBrought together all our aspirations, research findings and challenges \ncreated a travel discovery engine (bing stole that)\n
• #25: Full screen search interface\nstructured the long tail of options into categories\nyou can select any number of combinations\n
• #26: We’ll create a visual string of your search query\nand update searches in real time\nupdate availability of options\nshows changes immediately on the page\n
• #27: continually improve\nwe’ve tried out UserTesting.com \nto get quick research results \nvariety of computer setups\nso doubled up as debugging / QA insight\n
• #28: As expected we learned users mental model of search was seriously challenged\nearly adopters / people like ourselves loved it\ntested well when pointed at the search\nbut for initial business metrics we need to convert\n\n
• #29: Small subset test: Ski\nroll back this search for a short while. \nWhile people love the interface once they try it out. \nHow to open the interface was the continued issue\nmore traditional looking faceted search\n
• #30: Conversion is up from 18% to \n30% in Google and 46% in Bing\n
• #31: Going forward:\nWe are getting ready for the proper launch\n
• #32: rolling out a hybrid for holidays\nmake adoption easier\nthen take our customers on journey to help with adoption\n
• #33: Struggling to find a succinct way to describe our USPs. \nDestination agnostic travel search + price comparison\nRolls off the tongue\ntweet answers to....\n
• #34: We closely watch our data\nRob will tell you a little about what he has been working on\n
• #35: \n
• #36: \n