We create methodologies and open-source tools for performing social science in the digital world: recruiting representative survey panels and evaluating the behavioral impact of ad campaigns, apps, and websites.

Recruit

Leverage the targeting power of digital advertising to recruit a custom, stratified pool of respondents. Virtual Lab creates custom audiences and conversion events to optimize targeting, enabling you to stratify by any variable for which you can gather data.

Optimize

Virtual Lab creates separate ad sets for each stratum and optimizes ad spending to ensure your over-represented strata and your under-represented strata become properly-represented strata.

Survey

Survey respondents via web surveys, chatbots, or any digital experience you can connect. Virtual Lab comes with a lightweight web survey optimized for poor connections and a chatbot that works with Facebook Messenger and Whatsapp. Follow up after hours, days, weeks or months for seamless longitudinal studies or experience sampling.

Retarget

Divide your respondents into balanced treatment groups and test the effect of your ad by actually advertising to them in the real world, on their Facebook timelines, Instagram feeds, Google search results, or the websites they browse.

Monitor

Monitor responses in realtime via the Virtual Lab dashboard and download current data in CSV format at any time. See when respondents join, how long they spend, and how many questions they answer.

Secure

Open source code ensures full transparency and auditing of security practices. Host Virtual Lab yourself and own your own data or let us take care of the hosting on GDPR compliant, secure Google Cloud infrastructure. All network communication is encrypted by default with TLS and all data at rest stored on encrypted cloud block storage.

Want more info?

Check out our slidedeck

Use Virtual Lab in your next project:

info@vlab.digital

Virtual Lab was built to power studies by us, researchers and academics trying to come up with more efficient models of running studies. Here's how it's been used, both by us and by external research teams:

Gender Attitudes in India

Do short videoclips delivered through social media change gender attitudes and promote information-seeking behaviors in Urban India?

This study evaluates a series of online videos supported and produced by two edutainment champions in the field: Population Foundation of India (PFI) and the WEvolve initiative. Respondents are randomly assigned to either a control group or one of three treatment arms, where they are sent videos from PFI, WEvolve, or both.

Respondents are sent videos, one-at-a-time, over the course of several days. All video plays are recorded on the Virtual Lab platform. Baseline questions on knowledge and attitudes are asked before the videos, one week after the videos are shown, and again after 3 months.

Respondents are invited to share the videos with their friends as well as sent links to webpages of NGOs working on gender-related topics. Clicks to the websites are tracked and a measurement of time visited calculated allowing the researchers to measure quasi-behavioral outcomes.

Donati D., Orozco V., and Rao N. (working paper)

Covid-19 and Stereotypes in Italy

In addition to its impacts on public health and the global economic system, the novel coronavirus has led to increasing concerns about racial attitudes and stereotypes and their impacts on political attitudes and sentiments towards global cooperation. These stereotypes may also translate into more general geo-political attitudes towards individual countries, such as China, who was at the center of the outbreak.

Using the geographical targeting in Facebook's ad platform, respondents were collected across 542 municipalities and 90 provinces in Italy. Surveys began in February, 2020 and were continued through out the crisis, with a few questions being asked to each respondent at weekly or bi-weekly intervals.

Asking questions via Messenger enabled the researchers to have continued, constant contact with demographically targeted respondents with low attrition rates and extremely low recruitment costs.

Donati D., Gars J., and Rao N. (working paper)

Professional Aspirations in Nigeria

What do urban young people in Nigeria want for their future and how do they think about opportunities in the public and private sectors?

Recruitment was carried out via an advertising campaign on Facebook targeted to residents of the 10 largest cities in Nigeria, with population above 1 million people (Abuja, Lagos, Kano, Ibadan, Benin City, Port-Harcourt, Jos, Ilorin, Kaduna, Enugu) aged between 13 and 19. To incentivize participation in the survey, the Ad banner displayed a picture of a brand-new Nigerian-made tablet (TECNO DROID PAD 8 II) and a sign saying that anyone who completed our survey would enter a raffle to win up to 3 tablets.

Overall, during one week, about 3000 people clicked on the Facebook Ad, about 890 people started the survey and almost 560 completed it. The survey collected information on 1. Beliefs about Public and Private Sectors; 2. Career choices; 3. Use of media.

Interested in running Virtual Lab on your own infrastructure? All code is open source and runs on any server cluster running Kubernetes + Helm. Also, we're hiring.

Jobs

Our mission is to make it easier and cheaper for researchers, governments, and NGO's to perform statistically valid inference about the attitudes and behaviors of populations. We leverage digital tools to reach individuals, ask them questions, and involve them in experiments.

As an engineer with Virtual Lab, you will join a small team dedicated to making smart choices and building a robust, reliable, and scalable platform that leverages the latest technologies. The platform is open source, we use open source tools, and you will be encouraged to contribute to existing libraries and/or abstract logic into reusable tools where appropriate.

Interested? jobs@vlab.digital

Senior Software Engineer

You should:

1. Be excited to work with a large set of services written in Go, Node.js, and Python.

2. Love evented systems, queues, immutability, idempotence, visibility, and reproducibility.

3. Be comfortable designing both transactional and analytics databases, as well as moving data in/out of them efficiently and reliably.

4. Love container orchestration. We currently use Kubernetes.

5. Love testing, including complex integration tests involving many containers and services.

6. Be ready to grow into a technical leadership role with architectural and design responsibilities.