Leave a comment
Get the GH Bookmarklet

AMAs

A graduate of The University of Washington in mathematics, Alexander has worked on numerous product teams over the last decade with the focus on providing deep dive analytic projects to shape the product roadmap. He joined the App Ads team at Facebook over a year ago where he provides data-based guidance to drive product strategy. His insights help develop new app ad products, increase the efficiency of existing ad products and create new optimization options for App Install.

Prior to Facebook, Alexander held various positions at Capital One and Amazon where he used data science to develop business insights to influence key product decisions.

Alex is passionate about technology, specifically the app landscape. He enjoys speaking to startup founders and helping developers/ marketers with their app growth strategy. His advice – Focus on metrics that go beyond the install and drive downstream value for you.

Feel free to reach out to him on LinkedIn

He will be live on Oct 27 starting at 930 AM PT for one and a half hours during which he will answer as many questions as possible.

  • ES

    Eleanna Sbokou

    3 months ago #

    Hi Alex and thanks for doing this AMA, I'm sure there's going to be many questions for you to answer.

    So here are mine :)

    1. Let's say you have a given budget for app install ads on Facebook. Would you rather spend it all within the first 2-3 days (right after a new app is launched for example) or would you distribute it evenly over a period of say 1-2 months to gain regular traction and not just an initial boost?

    2. What works better when it comes to driving app installs: Ads that show a demo video of the app presenting the in-app experience or image/carousel ads showing screenshots (e.g. onboarding screens and in-app screens)

    3. How many organic app installs would you say 1 paid app install can drive? There has been a few articles talking about how paid app user acquisition boosts organic app installs. I would be really interested to see if you have any data that shows a relationship between paid vs. organic app installs.

    • AB

      Alexander Barickman

      3 months ago #

      1. My strategy would be to understand what segments my ads are working best in and getting the objective I care about. Let’s take a game for instance the users I would be interested in are those that refer their friends to play my game as this helps build an organic growth for my app. The other segment I would be interested in are those that monetize. I would start out pretty broad and spending enough to get the data needed to make an informed directional decision on the best way on spending to get these two objectives. So in this instance it would probably be the later, collect the data first and then scale up over time once I have the data to make the right decisions.

      2. I think this varies app to app and by vertical. Video is incredibly powerful for gaming, whereas carousel is great for eCommerce showing multiple products. During the data collection phase above I would be doing these types of tests in order to best understand what type of format works best for my app.

      3. This one again is app to app dependent, I think a big driver of organic is utilizing recommendation and referral systems within apps to get your users to get their friends to install an app. Sometimes this turns into a form of advertising in itself if you are giving real money incentives to users it can be as effective as running app install ads.

      6 Share
  • TN

    Tri Nguyen

    3 months ago #

    If you had to choose, what are the top (up to 3) languages/tools that someone should learn to become a good data scientist?

    • AB

      Alexander Barickman

      3 months ago #

      Being able to take data from a database and transform it into a story is the most important thing. SQL is the standard for this and I would recommend that over anything. After that it's variable, we have amazing data scientists that work heavily in Python, R and even excel. As long as you can get to the story portion and be able to communicate it to influence the direction of the team the format and tools you use to get there are not as important.

  • HQ

    Hila Qu

    3 months ago #

    Hi Alex,

    Welcome to GH. So what are some secrets and tips you can share for marketers to increase the efficiency of Facebook ads? Can you eleborate a little more on "Focus on metrics that go beyond the install and drive downstream value for you" ?

    And a while ago, I saw a twitter discussion talking about the rate of different social media developing new ad options, Facebook is ranked #1. How do you iterate and develop new ad products inside Facebook?

    • AB

      Alexander Barickman

      3 months ago #

      Hey all, first off really glad to be here and answer your great questions. I will try to get to everything and answer the best I can.

      1. It’s really important to know what you ultimately care about when buying installs, every app is in a different life stage and goals will vary across developers. Some are in growth stage where growing active users is important, others want to monetize, tying back app ad spend to these true objectives is critical for ensuring app revenue is being spent the best way possible. This is why we launched products such as app event optimization so advertisers can optimize for things closer to their true objectives.

      2. We do a lot of inbound, our team is incredibly passionate about building the products our advertisers need for growing their businesses. I work heavily with our research team, marketing and client partner teams to constantly get a pulse on our advertiser base and listen to what they need. We make a lot of trips around the world every year to talk to developers and advertisers to make sure we are building the best products for the market. App event optimization is a great example of something we focused really hard on this year as it was one of the most requested new products.

      5 Share
      • HQ

        Hila Qu

        3 months ago #

        Very cool, I haven't really used it but App event optimization makes a lot of sense.

  • MM

    martín medina

    3 months ago #

    Hey Alex!

    Thanks for coming on here and doing this AMA.

    Facebook is famous for running all kinds of different crazy tests and experiments what are some unique things you have seen while working on app ads?

    What are you most excited about for the future of mobile and what are you most excited about for the future of Facebook?

    Thanks!

    • AB

      Alexander Barickman

      3 months ago #

      1. Last year we did a test trying to improve the app ad experience for lower end phones and connections on our iOS app. We noticed that the latency for the store sheet that Apple provides for in app had a high latency for this segment. We ran a test to see the difference in performance for our ads if we kicked people out to the native app store instead of loading the sheet. We did this for all segments because we wanted to baseline the performance against high end phones and connections, we ended up seeing a pretty significant increase in performance for all segments. This was surprising to us as we expected the ones where the sheet was loading quickly to not see much difference in performance, invalidating our assumptions through tests has been a historical thing for our team and resulted in our team moto of “No sacred cats”.

      2. I am really excited for the future of chat bots on mobile, I think there is a large opportunity there and we see a lot of great applications already existing in Asia.

      2 Share
      • AA

        Anuj Adhiya

        3 months ago #

        Re: chat bot applications in Asia - is there any one (or more) that specifically stuck out to you?

      • AB

        Alexander Barickman

        3 months ago #

        My favorite I have seen is the QR scanning to get in line and you get a message notification when it's your turn. This saves the time it takes to transfer that information to a person and get it setup.

  • LS

    Logan Stoneman

    3 months ago #

    Hey Alex! I work heavily in the ecommerce sector and am curious to when you think retail/ecommerce websites should consider building an app, rather than optimizing their mobile experience. Are there signs you look for when making that decision?

    • AB

      Alexander Barickman

      3 months ago #

      There are a few key things I would consider, 1) What would the added value of the app be compared to a mobile site? Examples would be increasing stickiness and usage, but there is also what features can be implemented that are difficult for mobile web. 2) What is the expected behavior for users versus mobile web, if you have a heavy amount of frequent users it can make sense to build a dedicated app. From the user perspective I would expect better convenience from an app. 3) I also think feature parity is paramount, it is incredibly frustrating getting an app for a product I use on desktop and mobile and not being able to do everything that I was previously able to do. This causes me to usually abandon the app and go back to the previous experience.

      This is a brief few things I would consider and there is much more depth on this area but I think understanding the users of your product is the first place to start, getting inbound feedback from research can tell you a lot of things that your data can not.

      3 Share
  • AA

    Anuj Adhiya

    3 months ago #

    Hey Alex - so cool to have you on!

    I have so many questions for you but I'll limit myself to 2 (for now)

    a. What would you say are the top 3 things that are common between app ads that convert the highest (in general or with respect to a specific, popular persona if that makes it easier to answer)?

    b. What are the most common "traps" people fall into with constructing app ads that could potentially lead to budgets being wasted upfront?

    • AB

      Alexander Barickman

      3 months ago #

      1. I would say is understanding the true objective of the ad and optimizing the ads to get people that deliver that objective. There isn’t a general best practices I can think of that would apply to all ads as it is incredibly variable depending on the business.

      2. I think the biggest trap is assuming with all the targeting options we often what the best will be and not truly exploring with broader targeting. I have run many ads and been surprised which segments turn out to be high value and often contradict with what my assumptions were. An example of this was I was running ads for my friends horseback riding business and I found a lot of older people were converting really highly on them, it turns out they were buying the lessons as gifts for children/grand-children.

  • SA

    Shaker A

    3 months ago #

    Hi Alex,

    Thanks for doing this AMA!

    1)How do you increase your chances of determining that a feature that you are thinking of building is of genuine value to the user vs just leading down the path to a more bloated product? Do you have a process for vetting features? Additionally as your product grows how do you make sure your product doesn't become bloated?

    2)what are the most important lessons you've learned over your career?

    3)Can you mention some of the resources you use to stay up to date on business? (Books, courses, podcasts, blogs, etcs)

    Thanks

    • AB

      Alexander Barickman

      3 months ago #

      1. We do a lot of inbound feedback in the app ads space for new features we build, we never build anything that we don’t think there is a lot of demand for. Sometimes we launch features that turn out to not be heavily used, we track the usage of all the things we build for the product. We will sunset features that don’t get heavily used or go back to the drawing board and figure out what improvements need to be made so that it gets the desired outcomes for our advertisers.

      2. The most important lesson for being an analyst is owning up to mistakes in analyses, no one likes to be wrong and we are all human and make mistakes. I do hundreds of data analysis things a year and I will never 100% do all of them correctly. It’s important when you produce bad analysis or data to scream from the roof tops the error and communicate to everyone influenced by the data. This builds a heavy amount of trust with the team in my experience.

      3. Our research and marketing teams at Facebook are amazing and we utilize the FB platform and groups to share information with each other in an open way across the entire team. Everything that someone comes across or researches or gets inbound on goes to our teams’ core group where we can discuss it and drive action from the findings. This leads to everyone on the team from engineering to design to data science informed about every topic of our business.

      2 Share
  • ES

    Edward Stephens

    3 months ago #

    Hi Alex,

    Can't wait to have you on here. A couple of questions that I had:

    1) What do you think the untapped potential for Facebook advertising is, e.g. how much more can be done to optimise how ads are served up and will this be worked into the FB bot strategy e.g. a personlised bot that fetches interesting ads to show you?

    2) Have you identified and or are you working on an Oculus Rift ad-serve strategy and what do you imagine this might look like?

    3) Having formerly worked at Amazon what do you think their potential is for the Echo?

    • AB

      Alexander Barickman

      3 months ago #

      1. I feel we have barely scratched the surface on advertising attribution is still mostly being done last click basis and I think this will change rapidly over the next few years. I also think that optimizing for advertisers true objective is something we are constantly working towards and a lot of improvement will come from this area. I am not sure how bots will incorporate ads, in theory it shouldn’t alter the ads you see compared to ones already showing up on display, they both would utilize the same back end models for determining the best ads to show to the person.

      2. Haven’t even scraped the surface in this area.

      3. My partner actually works on the Echo team and we have a few at home, I love interacting (i.e. Yelling) with it, I think right now for myself it’s great for music and I think the opportunities are endless but I will let the Amazon team figure all that out.

  • FS

    Felipe Spina

    3 months ago #

    Hi Alex,

    Sounds good your AMA

    One question, what is your "black hats" on Facebook Ads? Such as Bid at $0,01? CRM with web personalization?

    In general, what's your hack in Ads platform?

    I'm looking forward to hear you

    Cheers
    @fspina

  • MW

    Marvin Wey

    3 months ago #

    Hey Alex,

    If you had a budget of $20k and wanted to release a reading app (where you can read free novels from aspiring authors). What would you do to rank as sustainably high as possible in the Apple App Store? Thank you very much!

    • AB

      Alexander Barickman

      3 months ago #

      I have tried to unravel the ranking algorithms of the app stores in the past and always led myself to dead ends. What I would focus on instead of trying to climb the app rankings in hopes that it would lead to a large amount of organic installs is focus on what you actually want your users to be doing in your app. I would then optimize my ads to go after those users and build a larger base of the desired user.

  • DH

    Dani Hart

    3 months ago #

    Hi Alexander,

    It's so great to have you here today! Thanks for joining.

    1. What tools and processes do you use to turn your deep dive analytics into an action item on the product roadmap?

    2. Can you give us an example of how one of your deep dive projects has impacted the app ad products?

    3. I imagine there are multiple Ad teams at Facebook. Is this correct? How do the teams stay in sync?

    Looking forward to hearing what you have to say!

    Cheers,
    Dani

    • AB

      Alexander Barickman

      3 months ago #

      1. I see a lot of great analysis from data scientists that ends up not being used and it’s often times because it is not communicated properly or advocated for. Whenever I find out something that I think is critically important for the team to know I communicate it out in multiple channels, I post in our working groups, I send the big tag line findings through messenger and email and I talk to people personally about what we should do about the findings. Communication is one of the most critical skills data scientists need to be great at as it turns findings into outcomes.

      2. A big project that I started through my analysis was the view through install reporting we pass to our mobile measurement partners that we launched this year. I did an analysis proving our assumption that they were overlapping with advertisers other ads on other networks was not accounting for all of the installs. We ran a ton of lift tests to validate that when holding out a segment on FB there was still unexplained incremental installs that click attribution was not accounting for. I also worked with our research team to understand why we were getting view installs and it turns out that there was different methods of getting installs that were not being captured by our click attribution.

      3. This is a great question and something that I have seen work poorly at other companies. Because we have groups and an open culture anything you are interested in for a product area you can get a constant stream of information on. We also heavily partner and meet with our partner teams in ranking, interfaces, and growth to ensure we know what each team is doing and how we can move forward in a complementary way. Our product manager and marketing teams are also always on top of what is going on internally and make sure the entire team is informed of everything we need to know about.

      3 Share
  • SE

    Sean Ellis

    3 months ago #

    Hi Alex, thanks for doing this AMA with us. I'm curious how you guys balance the needs of facebook users and facebook advertisers. It seems the more you meet advertisers' needs, the more you sacrifice the user experience. How do you guys make these decisions?

    • AB

      Alexander Barickman

      3 months ago #

      This is incredibly complicated so I will do my best at a concise answer, engagement on Facebook is really important as it allows us to show ads and monetize. For every product we launch we look at these engagement metrics in combination with our monetization metrics, we have incredible measurement tools that allow us to look at all the critical business metrics and make sure we are maintaining a proper balance of these two objectives.

  • AK

    Addison Killebrew

    3 months ago #

    Hey Alex,

    There are always new channels and ad formats being created. As a marketer I love them but are they getting to complicated to use for average day customers? How does a brand manage using the innovative new formats compared to classic image and video formats?
    Canvas ads for example do require a degree of tech savy to navigate so is it important to target to early innovators just simply because its an new format because regardless of the product you cant get grandma sue to convert on that format?
    what point should you jump on the train of new ad formats? How do you balance ad innovation and market-fit?

    Its just a matter of time before canvas style ads are the future but is it effective (or even logical) to pioneer that movement? Do ad platforms like Facebook work to integrate their pricing to reward those marketing trailblazers who are shaping the way people understand and interact with ads?
    Do aggregate conversions numbers of those ads formats exist by different demographics?

    • AB

      Alexander Barickman

      3 months ago #

      I think you bring up an incredibly great point on the investment it takes to run ads. We have gotten a lot of feedback similar to this on the complexity of the product and comments such as “If we didn’t have to invest so much on managing our ad spend we would spend more!” It’s something our team is well aware of and there’s a lot of solutions we could do to make it easier for advertisers such as reducing complexity, automating the operational portion of ads, giving better insights and information about expected and actual performance. I think the future of ad buying will move towards this things to make sure that products are scalable for everyone regardless where they are in their lifecycle.

  • JF

    Javier Feldman

    3 months ago #

    Hi Alexander, thanks for being here today! Here's my question:

    Are there any examples of Facebook Canvas Ads that you've come across that you thought were particularly well done/converted very well? Also what is it about this ad format that you think people still don't understand/understand very well?

    Regards!

    • AB

      Alexander Barickman

      3 months ago #

      Canvas was originally created in the idea that you could create an immersive story for your product and customize it the best fit for your desired outcome. It started off in our Brand advertising and we brought it over to our direct response objectives due to demand. I think it's something that has a lot of potential, I think it requires a lot of investment into getting it to work and depending where you are at may not be worth the investment until in the future where we have better case studies and examples of success. It's a really new product so I think a lot of what you bring up is still being worked out.

  • RH

    Robert Huang

    3 months ago #

    Hey Alex. I have an app that can't be listed in the Play Store because of "adult" content. Can I still use Facebook to drive installs? For example, using website conversions but optimize for installs. Thanks!

  • SL

    Sugianto Lauw

    3 months ago #

    Hi Alex,

    Thank you for doing this.

    What is your take on productizing an AI product? What sort of business model in your opinion would be better to monetize AI products?

    Much appreciated.

    Cheers,

    Sugi

  • RB

    Ry B

    3 months ago #

    Alex,

    Thanks for doing this AMA. Excited to learn from you!

    1)How do you look at balance in your professional and personal life? When it comes to work, how do you decide what you have to work on today (I'm sure you have a lot of fires to put out every day)?

    2)How do you look at hiring? Can you talk about some of the mistakes you've made hiring (and also seen others make)? What have you learned about hiring A+ talents?

    • AB

      Alexander Barickman

      3 months ago #

      1. I love working at Facebook because each person is empowered to work on what they think is impactful. I am not required to do things because someone requested it. This allows me to easily get all my work done in the office and keep my work to a 40 hour schedule. If there’s impactful work that isn’t being done because I don’t have the time I will let me leadership team know and we will work out a way to get more resources to make sure that we can meet the work load.

      2. I haven’t really seen mistakes in hiring I think the interview process at all the companies I work at have set a really high bar that ensures talent can perform in the role. I specifically look for people that are passionate and excited about the data science area.

  • AA

    Anuj Adhiya

    3 months ago #

    Couldn't resist - one more for you:

    What's your Analytics stack?
    Can you talk about what role the individual tools in the stack play in helping you get whatever insights you need?

Join over 70,000 growth pros from companies like Uber, Pinterest & Twitter

Get Weekly Top Posts
High five! You’re in.
SHARE
61
61