I helped to scale the way Meta identifies risks in elections, globally.  I am unable to share precise details about the methodology and variables used, but the truncated presentation below paints the picture with a broad brush. 
This research required quantitative modeling that was validated through qualitative interviews with users from 23 countries, along with both internal and external subject matter experts.
I was the research lead on this project and this system is now being used by Meta to identify risks during the busy 2024 election cycle. 
In 2022, it became clear to me that election workers in Brazil and the US were under threat.  I advocated for creating a guide to help them lock down their social media accounts, protecting them from unwanted contact, hacking, and potential doxing. 
I interviewed users in Nigeria, the US, Brazil, and the Philippines to see whether they understood product safety interventions we might launch around national elections in their countries.  
I also consulted with a team to build profile frames indicating that someone had voted, which raised civic awareness and encouraged people to vote.
Problem:  ACME Co.'s first product was a mobile app that made it easy for content creators to make short-form video recipes.  They wanted to know the pain points, motivation, fears, and excitement food creators experience when producing video recipes so they could prioritize features for their product launch.
Timeframe:  One week
Methods:  Seven 30 minute structured interviews, coding and analysis in ATLAS.ti
Please see the presentation below for results, findings, and recommendations:
PROJECT:  EBAY'S Punitive Feedback System Hurts Users
Problem:  EBay’s feedback system lets buyers and sellers rate each other, adding credibility and trust to the platform.  There are issues with the current system, however, because buyers and sellers can see the feedback left for them by the other party, which creates an incentive to wait until they have received positive feedback before leaving theirs.  I conducted this study to test the notion that the system is unfair and creates an incentive to punitively leave negative feedback if it has first been left by the other party.  
Timeframe:  Two weeks
Methods:  Open-source research on a sample of auctions and statistical analysis using t-tests and chi-square tests in SPSS
Results:  In transactions involving negative feedback, when a buyer or seller left negative feedback for the other party first, they received negative feedback in return 100% of the time compared to 57% of the time if they left positive feedback first.  This difference was statistically significant. 
Findings and Recommendations: 
Open feedback encourages retaliation.  To encourage a less punitive and more honest feedback system, eBay should institute and incentivize a system in which buyers and sellers leave blind feedback.
Feedback should be time-limited.  Allowing only 30 days for parties to leave blind feedback for each other would deny buyers and sellers the opportunity to wait each other out. 
In the End:  This study was published in the journal Nebraska Anthropologist and a copy was sent to a personal contact at eBay.  EBay did revamp their feedback system in the early 2010s, but it is unclear whether this research contributed to the change – the revised system still does not include blind feedback.  Other platforms appear to have learned the lessons I related above, though.  AirBnB, for example, uses blind feedback in its system. 
Download a copy of this paper here.​​​​
PROJECT:  All Hands on Deck - Managing Crisis in Live Event Production
Problem:  The COVID-19 pandemic has uprooted the concert and live event industry.  While the pandemic was the longest and most impactful disruption they have faced, this was not the first time a major event negatively influenced their market.  The 9/11 terrorist attacks, 2017 Las Vegas concert mass shooting, and Pulse Nightclub attack all had lasting material impacts on the industry.  How can live event producers manage shocks to their systems?      
Timeframe:  Two months
Methods:  Semi-structured long-form interviews with a ticketing executive, grounded analysis using ATLAS.ti, design thinking
Findings and Recommendations: 
Disruptions can lead to “massive” innovations and “creative destruction” is real.  Established companies can become agile in the face seemingly insurmountable change if they are willing to sacrifice time-honored policies and procedures.  With the right commitment from the top, organizational change can and does happen in large companies, so why wait for the sky to fall before implementing an innovative culture?  Create plans to respond to disruptions before you encounter them.  
Established hierarchies upend in times of crisis.  In high-stress tactical situations, the most skilled person is often given control, regardless of their title or place in the org chart.  Try to identify these specialists before a crisis emerges and build a support plan for them, should something arise.  They will be working outside their comfort zone and relying too heavily on them is unfair and may have unintended consequences.    
Contingency planning minimizes burnout.  During a crisis, it is not uncommon for employees to make heroic efforts at work.  In the long term, this level of effort is unsustainable, which leads to burnout.  Having a plan for unexpected edge case scenarios can minimize this risk by giving people an agreed upon order upon which to fall back.  This is especially important for key employees in critical positions, who may find themselves quickly overburdened in their already high-effort roles. 
In the End:  The company has implemented changes in their crisis and contingency planning and has hired security and pandemic consultants to help them develop processes for responding to the next crisis.  
See slides five through 13 in the presentation below for a brief readout of these findings.
​​​​​​Project:  Service Industry Communication Roadblocks Stifle Innovation 
Problem:  A bar and restaurant group with a loyal crowd of twenty and early thirty-somethings was losing clientele as they "aged out" of the bar scene.  The businesses were clearly not attracting new clientele, but the owner held the attitude that “We’ve always done it this way, so why change?”  This frustrated employees, managers, and remaining patrons and lost the owner money, morale, and experienced staff.    
Timeframe:  Two months
Methods:  Behavior observation, long-form interviews with staff and management, design thinking
Findings and Recommendations: 
Trust goes both ways, and it comes through execution.  Bartenders have ideas to make their establishments better, but they often do not feel heard.  We know that owners value their patrons and want to keep what they have without alienating people, you can build trust with staff by soliciting ideas and instituting low-risk changes.  This gives bartenders a sense of ownership they crave. 
To get better ideas from staff, take the "high" out of "hierarchy."  Managers often act as gatekeepers for busy owners, but how can ownership be sure they are not missing out on good ideas?  At what point should a bar solicit ideas from its staff?  Do your staff members need to fight to get time with you, or to get an idea implemented?  Why?  What does someone need to do to convince an owner that their idea is worth acting on?
Bartenders are trend vampires - listen to their expertise.  Bartenders are on the front lines every day.  They understand the people coming into their establishments and have as much financial interest in being busy as owners do.  While an owner may be worried leaving faithful patrons behind, the truth is, those people may already be gone.  Hiring staff who are in tune with trends and able to maximize can help ownership stay current and profitable. 
In the End:  I completed this project in the Spring of 2020, just before the COVID-19 pandemic effectively shut down the service sector.  The manager I worked with for most of the project left the industry but she will use our findings if she returns.  I will use these findings in my future work because they are generalizable to many service-oriented workplaces.  ​​​​​​​
Back to Top