Unmet Needs, Opportunities, Recommendations, Next Steps

Viator’s strength — its deep directory of product listings — currently doubles as a weakness insofar as the site largely presents itself as a directory, which can be hard to relate to. When considering CX- and IxD-related concerns, top design opportunities to make the site more sticky are:

• Represent common traveler segments by developing content funnels around primary archetypes (personalization). From there, curate and promote content around 3 categories — food, geography (a.k.a. “geo-hubs”), and local resources — which are absent and should be front & center.

  • Taking food as an example, Figure 3 illustrates some progress since the time of this study, although IA issues persist.

• Help travelers be more intentional, goal-oriented, in control, and less overwhelmed through simplified product taxonomy, which places too much demand on working memory and some cognitive disabilities. Figure 4.

  • By extension, there are various opportunities to enhance:

    • Faceted search

    • Focused search: e.g., berrypicking 🍓

    • Exploratory search: e.g., information foraging

• Translate abstract and subjective desires (for authenticity, nostalgia, quintessence, adventure, etc.) into proposed solutions. Test accordingly.

• Partner with trusted travel bloggers and expand U.S. social media presence through ad campaigns.

• To make the site more inspirational, form a content strategy team that additionally offers a robust and streamlined IA practice.

So What? The Bottom Line and a Value Prop

The more the above recommendations are meaningfully addressed, the more visitors gradually wean off Google as they:

• Form new mental models (small-scale schemata) as they adapt to relevant and relatable content (personalization) in more appropriate informational chunks and locations (load, working memory, IA governance), gradually reconceiving how to interact with the site…

• …which, in turn, promotes new how-to behavior (procedural knowledge) by replacing Google with in-site design enhancements built from enhanced search options (e.g., berrypicking, foraging)…

• …which, in turn, fosters…

  • Visitors seeing themselves in the site, which relates to identity formation (personalization)

    • Herein lies the notion of “valueception”: Helping customers better attune to actual value

  • Brand loyalty, which is lacking in the U.S.

  • At a minimum, performing in-site searches; at a maximum, performing goal-directed in-site journeys

• …all of which mitigates high SEO/SEM costs over time and, in the process, makes the site stickier.

Impact and Business Outcomes

I wrote a short position paper on the above issues, which was triangulated with existing research and included a series of actionable recommendations, some of which were integrated into product roadmaps. Thanks in part to this work, and in conjunction with an expert review (formative qualitative audit) of booking processes I conducted, remedies to the more acute issues — personalization and IA governance — are currently underway.

For example, Figure 5 shows that content funnels organized by interest (personalization) were added to the site’s global navigation. While this is progress, however, placing the funnels as sequential menus within a mega-menu — which itself is easily overlooked — isn’t the design approach I had recommended. Moreover, the funnels take visitors to blog articles, not curated results, which is another shortcoming.

In another example, in Figure 6, recommendations about promoting content by geography (“geo-hubs”) also made it to the global nav. And this time, the upside is that the content funnels take you to results, not blog posts.

Down the road, generative AI may further aid these types of content curation.

Constraints of Primary Research

• Little root cause analysis evident in primary research

  • e.g., scant data re: satisfaction with tools and sites used

• Non-standardized mod guide

• A ~50/50 split between granular data and composited (generic) data sacrificed several qualities of synthesis. For instance, it prevented the development of practical definitions of abstractions (e.g., authenticity, adventure) and limited more nuanced interconnectedness of overlapping topics.

Constraint of Secondary Research

• Secondary Observer Effect (a.k.a. Evaluator Effect): idiosyncratic or subliminal variations across researchers

Retrospective

• What did I learn?

  • A considerable amount about travel styles!

  • Enhanced coding and filtering prowess

  • Meta-studies are not so far removed from literature reviews in the sense that you’re simultaneously deploying several zones of focus as you move from explorations to evaluations to conclusions to recommendations.

• What could have gone better?

  • The work was monastic in nature. The non-collaborative nature of the effort led to occasional analysis paralysis.

  • Grounded theory is not to be underestimated, i.e., it took longer than expected. I was also taking a challenging graduate seminar at the time, which additionally reduced the pace I wanted to move at.

• What would I have done differently?

  • Leaned on 1-2 colleagues for more frequent reality checks.

  • If I had the time and 12-25 participants, I would’ve aimed for the gold standard: qualitative interviews with a standardized methodology and mod guide.

    • Had this been a usability study of certain aspects of the website, and not a meta-study about travel, I originally imagined some form of Kano modeling may have been useful here (customer delight vs. implementation investment). I’ve since been convinced that likely would’ve been folly and I would’ve instead used conjoint analysis and MaxDiff.

• What went well? What will I keep doing?

  • I had time for quality coding and synthesis, which helped give the work the attention it deserved. In turn, this contributed to interventions currently in development.

    • In contrast to Miro, Mural deserves a shoutout for its superior coding functionality.