Interaction Design & Strategy Fail for a Live Events Startup
Lesson Learned:
Don’t fall prey to the siren song of aesthetics.
Interaction Design & Strategy Fail for a Live Events Startup
In this article…
Project Overview
Type of Project:
Interaction Design (Prototype Creation)
+ MBA Consulting Practicum
Roles:
Interaction Designer (intern)
Consulting Team Lead
Client:
Undisclosed due to NDA
Date:
December 2015
Location:
Honolulu, HI
Project Goal:
Conduct market research & identify market gaps and opportunities in order to build a user-centered service & prototype.
Project Details
To put it bluntly:
This company was destined to fail.
That said, this company was also not inherently “bad.” When we started this project, their mission was bold: to be an industry leader for the future of interest-based social engagement, focused on connecting locals and tourists alike with upcoming events, festivals, and merchant promotions. Not only did they want to host and present upcoming event information, they wanted to go a step further, using web crawlers and metadata to consolidate all event data, essentially becoming the premier place to search for events and learn about what was happening in specific locales. Hawaii, given its proximity to both Western and Eastern markets, was the perfectly placed test testing ground and local launch market to test the global viability of this tool.
Unfortunately, the company’s vision wasn’t enough. The company never made it past early stage funding. You might then be wondering (validly) why the heck I’m featuring this, of all possible case studies I could feature. I’m sharing it here for two primary reasons:
It underscores the importance of minimum viable products in the startup space.
While MVPs have become a bit demonized in recent years (and rightly so), this lesson is proving even more relevant in the age of the recent AI startup boom.The company (and I) fell prey to what I now call “the siren song of aesthetics,” which is, the irresistible temptation to see mockups or prototypes with high visual fidelity as soon as possible, at the expense of clarifying product objectives and functionality.
Updated: Dec 2024
a retrospective, 9 years later
How we got involved
In a collaboration between my MBA program and the Blue Startups incubator program, my team worked with this startup company by offering pro-bono consulting. Given this company’s early stage of identifying key partnerships, and securing seed funding, we were asked to help them with market differentiation and pitch preparation. We immediately began conducting secondary and primary research, identifying market gaps & opportunities, and created pitch decks for this team to secure the funding necessary for this effort.
I wasn’t even technically supposed to do interaction design, but when the scrappy startup founders learned I was proficient with prototyping tools (InVision, at the time). I found myself neck deep in design reviews and prototype creation (on top of survey design and interview script writing). Since we were already working for free (well, for MBA course credit) the founders leaned in, and took whatever labor we could provide.
What our team provided
In order to get a sense of the current operating environment, I orchestrated a three-stage market research plan, which included:
Literature review/audit of economic data: We combed through small business administration and tourism board data, to get a full perspective of the Hawaii events economy.
Stakeholder interviews: We interviewed prospective business partners in a variety of industries (arts & entertainment, transportation, tourism, hospitality) to get a better sense of their primary pain points with event marketing.
Competitive analysis: I conducted a heuristic evaluation of the existing online ticketing platforms (an adjacent space) to get a better sense of their limitations and offering gaps.
Metadata audit: Since the model of this company depended on two elements: organic event input and web-crawling, I reviewed common event metadata to help inform the information architecture for auto-generated events.
What we learned
Once we pooled this data together, we learned a few interesting things about this space:
While Hawaii was a uniquely useful market for recruiting and research for a global audience, it was actually an outlier environment for launching the app: the share of tourist activity was outsized compared to non-tourist (local) activity, and tourists had a low likelihood of installing a new application just for Hawaii events.
Browsing events is a tedious process. While the event ticketing space was saturated, and Facebook Events held the market share of event-data for free events, switching between two separate sources of information required greater interaction cost.
Current offerings did not address users’ core need: a single source of information for finding events (paid or free) based on interests or location. There were certainly web-crawlers… but they sucked. They were dubious sources of information and people didn’t trust what they found.
Metadata was problematically inconsistent, and poorly labeled, making the web-crawler approach complex and likely more labor-intensive than previously estimated (more on this later).
Our recommendations at the time focused on ways to quickly gain market share and offer a superior product, even without the auto-generated events.
Establish strong governance for user-generated data with labeled fields and structured data (in the form of categorization) to ensure a functionally more usable event database. Facebook Events had much looser categorization which was functionally harder to use at the time.
Identify key business partnerships and hosting all new event and ticketing information for free for a period of time. This would decrease switching costs and create cross-promotion opportunities, leverage credibility of established local organizations, and ultimately attract enough early adopters to reach critical mass.
The “siren song of aesthetics”
While none of our suggestions had anything to do with the aesthetics, this company was so fervently chasing seed funding for this effort, that the highest priority interaction design efforts focused on mockups and demos for investor meetings. Whenever features and functionality was discussed, the refrain was “Once we get funding, we can find all the backend support we need.”
I get it. It’s hard to “wow” an audience when all you have is a wireframe and a dream, and it’s hard to hire a developer if you need funding. It’s a chicken-and-egg problem that won’t solve itself.
However, the issue was actually not a lack of developer knowledge - it was an issue of falling in love with an idea, and sailing toward mirage of a solution. Aesthetics may be a way to sell an idea, but they can only work if the idea is grounded in something that brings user value.
There are a couple of ways this company could’ve launched an MVP to test the hypothesis, without hiring a full-stack developer, and without spending so much time on aesthetics.
Prioritize a small subset of information sources to create a reliable events database. Rather than scrape the entire Internet, identify high priority sources of events (1-2 online ticket retailers and Facebook) and integrate based on specific metadata from those sources only.
Create an invite-only, web-based community and ask key business partners to post their own upcoming events. The invite-only approach can create perceived exclusivity, focus on community-building, and establish a localized repository of events.
Ultimately, people didn’t just need more events on an app. Web-crawlers would have succeeded in event quantity, but they would not have delivered on event quality. What they really needed was reliability and convenience.
Now, to offer a 2024 reprise on this:
When I first wrote about this case study, I lamented about this company’s need to adhere to their web-crawler strategy, and how it impacted their speed-to-market. This is still true, but at the same time, this startup was both too early and also too late.
It was too early because 2015 was in the infancy of LLM development and machine learning was still a new concept. Thanks to thousands of underpaid human data-labelers, LLMs now can process information like this with some ease.
However, it was too late, because the founders were trying to create a social media company in a five-year time period where Facebook (now Meta) was aggressively overpaying to acquire and outcompete social media companies. While the company’s founders were transparent in their goals to build quickly and get acquired, they were likely more focused on the appeal of acquisition than they were on the delivery of value for their customers.
Is this harsh? Maybe. If the founders are here reading this, I’m sorry to be this blunt. However, this is an unfortunate reality of tech startup culture and is endemic even now, in the age of AI startups. It’s not enough to have shiny curated demos. Products will fail if the fundamental architecture and definition of user value is as shallow as the vision for them.
•••