Survata is the only accurate way to measure the effectiveness of online brand advertisement.
Direct response ads (e.g. AdWords) are easy to measure: either people click or they don't. But have you ever clicked a banner ad about a new salad dressing, or a video ad about how your local utility is
more than just a power company? Neither have we, and neither has anyone else.
Companies spend billions of dollars on brand advertising without any way of knowing whether the advertising is working. Survata is the first tool that can tell brand advertisers whether their money is well-spent.
We're able to measure advertising impact by integrating with a broad network of websites and mobile apps who allow visitors to answer questions to unlock premium content (like an extra life in a mobile game) instead of paying money. By integrating with the advertising ecosystem, we're able to serve targeted surveys to visitors who we known have seen specific ad campaigns in the past. By comparing these
exposed responses to
control responses from visitors we know haven't been exposed, we're able to directly quantify the impact of the advertising.
Measuring ad impact at scale involves substantial challenges: we have to track and process billions of ad exposures each month. We have to use careful experimental design and rigorous statistics to ensure that we correctly identify the actual causal impact of the ads, and predict the impact of in-flight optimizations to ongoing campaigns. And we have to visualize the results of our measurement through a dashboard that is powerful, professional, and intuitive enough to allow marketers to understand and trust our recommendations about how to shift hundreds of millions of dollars of marketing spend each year.
Survata is big enough to have a solid team, ample funding, and a long list of happy customers. But we're small enough that there's room for you to own a substantial portion of our architecture, to tackle new and interesting technical challenges as we scale, and to have a clear, major impact on the trajectory of our technical systems and on the company's overall success.
Data Engineer San Francisco, CA, United States or Remote
Front-end Architect San Francisco, CA, United States
Front-end Engineer San Francisco, CA, United States
Full Stack Engineer San Francisco, CA, United States
Senior Front-end Engineer San Francisco, CA, United States
Data is our business. At many companies, data and statistics are used to optimize other lines of business. At Survata, our data and analysis is literally what we sell.
If you join us as a data engineer, you'll be working with high-scale systems that process 200+ billion records per month, and with statistical systems that using cutting-edge causal inference methods. If you join us as a product engineer (backend or frontend), you'll be working on professional software that helps our highly-motivated clients understand, visualize, and consume data that'll directly drive the reallocations of hundreds of millions of dollars of annualized marketing spending.
As a young company, we already work with Fortune 100 enterprise clients like, Clorox, Bayer, Verizon, and Unilever. Some of our contracts sizes are in the seven-figures (and might soon hit eight-figures).
Work with a strong and diverse engineering team. Our team includes a phD statistician from a top ten program, two core contributors to Django, and is approximately 40% women and/or under represented minorities.
We're a firm believer in
people over process and our CTO's motto is
we prioritize, not plan.
We use a customized version of Agile, but in the true agile spirit, we don't believe in rigid engineering frameworks and instead we focus on getting to know our engineers as people and creating a work environment that supports and encourages you to thrive and allows your individual talents to develop and flourish.
We develop products through cross-functional squads that include a product manager and as many engineers of different specialities as the product under development requires. Engineers of similar specialities report to common managers who are responsible for facilitating bottoms-up agreement on coding styles and standards and for providing adequate resources and opportunities for training and learning.
We've got two major technical challenges:
1) Keeping up with the massive quantity and speed of advertising data we need to track. We record billions of impressions served for ad campaigns by major advertisers like Clorox, and we need to make that data reliable and immediately available so we can quality survey respondents in real time.
2) Performing meaningful and believable causal inference on non-randomized survey data. Randomization is expensive, meaning that our survey data comes to us with substantial confounding baked in. By using careful experimental design and modern causal inference methods, we're able to statistically recover the pure effect of the ads, and predict how changing ongoing campaigns would change their ultimate efficacy.
3) Our measurement is so comprehensive that we produce more high-signal results than clients can easily consume. So we need a powerful-but-intuitive dashboard to allow clients to easily understand, explore, and synthesize the state of their campaigns and the most promising opportunities for optimization.
We built a Cassandra cluster that records over 30,000 advertising events per second, and can tell us in <20ms whether a given candidate survey respondent has seen an advertisement in the past.
We built a sophistical statistical pipeline to perform valid causal inference on survey data so that we can identify the actual effect of advertisement on consumer opinion (without confounding from the pre-existing differences in sentiment from how the ads are targeted).
We built a pipeline on AWS Batch to express our causal inference process as a Dockerized task-dependency graph (which allows us to run our causal inference process on surveys of any scale).
We are building a React/Redux-based dashboard that uses Plottable and D3.js to allow clients to quickly explore, visualize, and understand their data.
We're a bunch of nerds who place a high premium on curiosity and are very serious about using the best methodology and working with the highest quality data. We're big believers in collaboration and we maintain a friendly, high-trust culture where no jerks are tolerated and you're appreciated for who you are and are rewarded for helping make your whole team more successful.
We also strive to create a diverse and inclusive environment where people of any background are welcome and safe.
Survata offers unlimited vacation
We have catered lunch every Wednesday, and an endless supply of free snacks and drinks in our kitchen
We provide full medical and dental coverage
We have a company sponsored (but optional!)
booze class every Friday where you can learn about a new drink while relaxing with colleagues. We also have a variety of regular offsites which have recently included visits to a board game cafe, a bowling alley, cruises on the bay, and an escape room
We provide provide commuter benefits
Six week paid maternal/paternal leave
Interested in this company?
Skip straight to final-round interviews by applying through Triplebyte.