I will start with current state of business Define how many ads are shown Am I PM on newsfeed looking only for product improvement or overall business including partner channels what is business objective

The mission of FB – customer experience SWOT Industry trends Privacy Ecosystem

3 options Increase – keep it the same or decrease.

Based on the above analysis provide the answer and follow up with A/B testing and small cohort launch in developing countries that don’t have too many privacy issues

General Structure:

1. Run through what our goal is (maximize ad revenue while not sacrificing user engagement / customer experience) and incorporate any other high level goals Facebook is trying to accomplish.
2. Walk through your top 3 A/B tests (The question mentions quantity — but I would ask the interviewer if I could potentially venture beyond that and include modifications to frequency, location, size, and ad medium.)
3. For each test, talk about top line metrics and what would make it a “success” or not.
4. Discuss relevant tradeoffs for each A/B test. It would be nice to call out which A/B test you hypothesize would give the best results.

*Note: You should walk through the A/B test, metrics, and tradeoffs for each individual test at the same time. Will be much easier to follow that way.

Important Ideas I would nail home:

• Our metrics should be very conscious of how ad changes could negatively affect the platform — like MAU, time spent on site, interactions per visit (as they will see less content and more ads), NPS, etc.
• More ads doesn’t necessarily mean more revenue. If we clutter their newsfeed with ads, it will be too much they won’t engage with the ads at all. We need to look at the number of ad clicks per user to accurately gauge the change in revenue.
• A financial model could be useful, but the “best result” might not play into Facebook’s strategy and/or account for long term affects (6 months +). For example, MAU could decline 10% over 2 months with increased ads, but clicks per MAU could go up 25%. This would likely increase ad revenue, but it might not play into Facebook’s long term user retention goals.

The opening of people spending a lot time on Facebook is way too long. Almost half time of the video.

The analysis of Facebook business model should be very brief. People know Facebook business model.

Maximize the ads display while balancing the user experience is the key. She can drive to this point much faster.

Segmentation of user group feels irrelevant to the question.

I would answer the question in a different way.

Ads revenue is the single most revenue stream for Facebook. So we can safely assume Facebook found out the answer for this question long time ago and is continuing to tweak it to make it better everyday. For this question, I think we’re pretending that we travel back to the time when we tried to answer the question for the first time.

Maximize the ads display while balancing the user experience is the key. With that in mind, I’ll do the following.

I would devise a test showing ads in 10%(1 ads every 10 posts) to 50%(1 ads every 2 posts) of the newfeeds, in incremental of 10%. I will be doing this test in a few user groups, alternating ads contents. What I’m trying to find out is when the ads CTR is plateaued. I will also measure the average session time for each group. I would assume too many ads won’t provide interesting and relevant information to the user. User will shorten their session time. Then I’ll see at what percentage the CTR plateaued. The I’ll try to correlate these two information together and try to figure out what’s the best percentage of newsfeed to display the ads.

Once we have a baseline information, I would push this to a much large simple group, by monitoring more variables like user age, which category ads(grocery, electronics, events) and which format(Text, photo, videos). Still paying attention to CTR to see any impact.

I just checked my facebook newsfeeds. I can see one sponsored ads every two or three posts. I guess 20~30% of newsfeed is ads.