Leave a comment
Get the GH Bookmarklet

AMAs

Hello there! I'm Ivan Kirigin. I helped Dropbox grow 12X in 2 years. My startup YesGraph is focused on helping companies grow by boosting the performance of viral flows. I've run countless growth experiments and I know way too much about crafting and optimizing referral programs. Check out this talk about growth I gave at Airbnb: http://j.mp/1ALSWXn And this one is all about referral programs for 500.co: http://j.mp/1y0Kjw7 Follow me on Twitter here: @ikirigin and @yesgraph

  • BP

    bas prass

    over 3 years ago #

    What's your favorite analytics tool and why?

    • IK

      Ivan Kirigin

      over 3 years ago #

      No hesitation here: https://segment.com/

      1) It helps you abstract your instrumentation correctly. It's just good software design.

      2) Adding new analytics tools is easier, which makes answering questions easier. Answering questions is why analytics exists. Fancy graph porn, no matter how much I enjoy it, isn't the point.

      3) It makes it easy to keep all your data with their webhook. Keeping all your own data means you can answer questions when you need to. Answering questions more easily is the point of analytics. I'm repeating myself because many people don't understand this.
      https://segment.com/docs/integrations/webhooks/

      4) The price is cheaper than the engineering cost of supporting multiple direct installations.

      5) Their live event debugging tools make onboarding easier than most analytics tools.

  • SH

    Samuel Hulick

    over 3 years ago #

    Why did you have the bloody-faced lion avatar for a while, and why did you stop having it?

    • IK

      Ivan Kirigin

      over 3 years ago #

      I like things like nukes and big violent animals like cats. You'd probably have to plumb deep into my psychosis to understand that.

      I asked Twitter if I should get rid of it, and someone said yes. Real faces are better than bloody lions I guess? This wasn't a data driven decision. IIRC you can blame @msg

      I look forward to resurrecting some big carnivore in office art. I like topographical maps too.

  • GW

    Gordon Wintrob

    over 3 years ago #

    Hey Ivan. Thanks for hosting! Really like your series of "Free Ideas" (http://blog.kirigin.com/free-ideas). Could you share more about 1) how you generate these ideas (sources of inspiration, techniques like writing down an idea daily, etc.) and 2) why YesGraph was the company you wanted to focus on 100%?

    • IK

      Ivan Kirigin

      over 3 years ago #

      I'm glad you like them!

      1) I can't not have ideas. I talk about the future constantly with everyone around me. I see the behavior of, say, my kids with games, and I think "damn this art focused memory app should exist". I don't know SF neighborhoods, so I think a neighborhood map app should exist. I don't think Zapier & IFTTT are complete, so I think, there should be a way for a new google spreadsheet row to trigger a callback.

      These are all real ideas that I just haven't written blog posts about yet.

      I should say that my most recent post about the future of work is really compelling. http://blog.kirigin.com/digital-work

      I *really* want someone to make that.

      2) YesGraph is the result of hard earned knowledge about something missing in the world. This was true while it was a recruiting product, and now with its growth focus. I think the future is really bright and want to go and build it. Each of the other ideas I've posted are in their own way harder. It is obviously a judgement call, not something ruled by hard numbers.

      From a 50,000 foot view, I'm optimizing risk & reward.

      • GW

        Gordon Wintrob

        over 3 years ago #

        Amazing answers, Ivan. I also suffer from the "idea stream" and it can be hard to keep them organized. I love that you've been fleshing them out and I'm looking forward to future posts!

  • MB

    Morgan Brown

    over 3 years ago #

    Hey Ivan,

    Thanks so much for doing this AMA. You gave a great talk at 500 Startups Distribution conference on tips for optimizing your referral programs. You had 27 of them I think.

    What's the number one problem most people have with referral programs? Is it expectations, instrumentation, top of funnel traffic, something else?

    Thanks for your insight!

    • IK

      Ivan Kirigin

      over 3 years ago #

      Understanding the funnel is really important, and it is the reason for mismatched expectations.

      A quick funnel might be:
      - % user that use referrals
      - # referrals sent per active referrer
      - % invites that get accepted

      If only 1% of your users participate, you aren't going to be happy with the results. These three metrics also mean there is a lot to optimize. Each part is pretty different.

      If you follow or straight copy best practices from others, you can get pretty far. But even then you should understand that optimization and good design of the program are hard work.

      You should judge every potential growth experiment by cost, benefit, and risk. Knowing a reasonable guess for the cost & benefit of a referral program can help you prioritize the feature.

  • SE

    Sean Ellis

    over 3 years ago #

    What led you to make the jump from growth guy for other companies to wanting to start your own company?

    • IK

      Ivan Kirigin

      over 3 years ago #

      I started my first company before I learned a lot about growth. It was Tipjoy in YC's W08 batch. That was a payments startup and I left a job in robotics to start it. Why I did that is probably a better question!

      But the answer turns out to be the same. I want to make a big impact in the world. That can mean starting my own company, or joining a rocket. I learned a lot at both Facebook & Dropbox, but I really love building something from scratch.

  • MB

    Morgan Brown

    over 3 years ago #

    Thanks @ikirgin, what are the things you learned about referrals and friend networks that led you to build YesGraph? What are some of the insights/nuances that most people miss but turn out to be super important?

    • IK

      Ivan Kirigin

      over 3 years ago #

      It's a numbers game. So if you push hard and follow up strong, your results can be better. YesGraph (both the old recruiting product and the new growth focus) take these bits into account.

      There are a bunch of tiny things we're optimizing that is hard to boil down into something here. That is really the lesson here for YesGraph: we do the hard part of analysis so you don't have to.

      For people doing friend networks in real life, the most important lesson is that weak connections matter a lot. It isn't your closest 10 friends but your moderate next 100 friends that will help you, for example, get your next job.

  • AS

    Adam Szabo

    over 3 years ago #

    What's your opinion on creating giveaways for growth?
    Do you believe a viral giveaway tool like KingSumo Giveaways can be effective for a small company?
    Thanks!

    • IK

      Ivan Kirigin

      over 3 years ago #

      You need to judge what to work on by cost, benefit, and risk. A giveaway can be really motivating for some audiences. If a service makes it easy and cheap to run the experiment, that lowers the estimated cost.

  • DL

    Dylan La Com

    over 3 years ago #

    @ivankirigin thanks for doing this AMA!

    What is the biggest win you've gotten from a growth experiment?

    • IK

      Ivan Kirigin

      over 3 years ago #

      You need to normalize wins by how much effort they've taken, right?

      For example, Dropbox launched a camera upload feature that dramatically increased the amount of photos uploaded. That was a major effort that included app and web redesigns. The results were excellent! But it took months of work.

      I once ran an AB test on transactional emails to include a "PS Get Extra Space Free" at the end. It bumped participation in the referral program dramatically, for every user, forever. It took an hour.

      You can hear a bit more about that in this talk: https://www.youtube.com/watch?v=KMGnOU3lwQg

  • DL

    Dylan La Com

    over 3 years ago #

    @ivankirigin what was your day-to-day like working at Dropbox and seeing the product take off? Any critical lessons for other early stage growth teams you'd like to share?

    • IK

      Ivan Kirigin

      over 3 years ago #

      There is no silver bullet. All the growth was incremental. All the experiments were incremental.

      Dropbox is really perfectionist, which is amazing but can mean they move more slowly than I'd recommend a growth team moving.

      I joined around 20-ish people. I left at ~160. They now have at least 5X that. Headcount growth is pretty shocking to experience first hand.

      A lot of the analytics team's focus early on was fighting scaling fires. Many of the tasks weren't done by other companies. For example, the data per user at Dropbox is incredibly high because of all the files being edited and managed.

      Probably the best lesson to take away for other teams is that great engineers are really important. Many engineers at Dropbox had a really sharp product mind. This means individuals can be empowered to make changes that make the user experience much better.

      Keith Rabois said it well: hire more barrels not more ammunition: http://startupclass.samaltman.com/courses/lec14/

      • DL

        Dylan La Com

        over 3 years ago #

        Super interesting. Was this product-intuition something they were actively screening for in their hiring process?

        • IK

          Ivan Kirigin

          over 3 years ago #

          Actually most of the engineering assessment was for high end technical knowledge. I wonder if there is a relationship there?

          I've definitely worked with some very capable engineers that don't know much about user experience, so I don't think so. Maybe it was a cultural filter that wasn't explicitly represented in the assessment.

  • LS

    Lai Saelee

    over 3 years ago #

    Thanks for doing the AMA, Ivan. For someone who is non-technical who wants to get into growth marketing, what technical skills should they learn right away to be most useful and helpful?

    • IK

      Ivan Kirigin

      over 3 years ago #

      Learn tools that don't require coding really well. This includes ad systems, excel, AB testing tools, analytics tools, BI tools. There is a lot more than just coding.

      You can also get better at product, psychology, and copywriting. There is a lot here!

      If you do start coding, being able to manipulate data is more valuable than building product features.

      • LS

        Lai Saelee

        over 3 years ago #

        Thanks for the tips! Looking forward to hunkering down and learning some skillz.

  • DL

    Dylan La Com

    over 3 years ago #

    @ivankirigin I asked this about Dropbox, your answer was awesome, and, sorry, but now I'd love to ask you the same thing about Facebook :)

    Here's the question again:
    What was your day-to-day like working at Facebook and seeing the product take off? Any critical lessons for other early stage growth teams you’d like to share?

  • RO

    Raf Ore

    over 3 years ago #

    Hi Ivan,

    Could you share with us what your growthhacking toolbox looks like?

    I'm sure you've used many things out there - I'm interested to see what your top favorite tools are (no particular categories). You already mentioned https://segment.com/ as your perfered analytical tool.

    Thanks in advanced!

    Raf.

    • IK

      Ivan Kirigin

      over 3 years ago #

      Segment is great for moving data. I also love optimizely and mixpanel and python for actually analyzing it. It depends on the application of course. Sometimes BI tools are needed to run deeper analysis.

      For making decisions, I use google docs and spreadsheets. A spreadsheet is a good way to organize a list with a few extra dimensions, like cost, benefit, and risk. Triage is best done using a tool like this.

      I figure out these estimates with docs around each experiment. Knowing the engineering cost and organizing relevant stats requires more space than a row in a spreadsheet.

      When working on a project, a doc should be expanded to include motivation, product change description, new instrumentation/analytics, predictions, press plan, subtasks etc

      Then I move to a task management tool like Asana or Trello to actually get things done.

      These are all things any product manager might use. I tend to have a very product focused approach to growth. Social and viral flows tend to always be product features, not just a growth experiment.

      The same framework can be used to organize things like paid acquisition campaigns. Instead of product changes, you're changing copy, creative, targeting etc.

      You still need to triage your efforts there.

      • RO

        Raf Ore

        over 3 years ago #

        Thanks mate!

        PS: With BI tools are you referring to tools like Similarweb or tools like Tableau?

  • JS

    Jeffrey Shaw

    over 3 years ago #

    What's one PPC targeting criteria that worked really well that surprised you?

  • BR

    Brad Ruderman

    over 3 years ago #

    What tools have you used for outbound emailing? Did you follow any specific processes for Outbound sales and tracking that in Salesforce/CRM?

    • IK

      Ivan Kirigin

      over 3 years ago #

      I've used LeadGenius for an outsourced SDR. I've used streak and yesware for outbound emailing, but I know there are a lot more sophisticated systems out there. I especially miss something that connects more closely with marketing and product events.

      For lifecycle emails, I didn't have a good experience with Customer.io. Intercom was OK. I just hand-code these and send using the transactional services of SendWithUs.

      I've never actually used Salesforce, and people say I should consider myself lucky. I use RelateIQ as a CRM.

      • BR

        Brad Ruderman

        over 3 years ago #

        Thanks for the info. Yea I use all those, but the problem is I am having trouble with an OOB that can I can do simple things like track outbound email campaign, email variation, # sent, # open, # replied, # calls scheduled, # of conversions. I am pulling data from so many locations! I guess its an integration problem 101, for growth hacking.

  • AG

    Andrew Gaffney

    over 3 years ago #

    Ivan, would you be interested in speaking at our Demand Gen Summit next month in Scottsdale. We are doing a session on Growth Hacker Marketing.

  • DN

    Delexan Navaratnam

    over 3 years ago #

    I signed up only a few hours ago ( Literally!) - What would your approach be with growth hacking resources and where to apply it in the end goal being to be a growth hacker at a startup ?

    Thanks Ivan in advance!

  • AM

    anoop munshi

    over 3 years ago #

    Emailed you

  • AZ

    Allen Zhao

    over 3 years ago #

    Hey Ivan, great to meet you here. Glad that YesGraph is coming up and I have already given my email for the exciting news of release.
    I'm with ptengine.com, it's an easy to use web analytics and heat map software looking for aggressive growth, it's not very easy to do that given the competitive situation. Are there any suggestions on what to do to hack the growth?

  • IK

    Igor Kryltsov

    about 3 years ago #

    Missed http://www.producthunt.com/posts/yesgraph-2

    If you don't mind a question about Yesgraph - can you list some even anonymous data how suggested recommendation works against standard now ability most web products offer to share product via Twitter, Facebook, LinkedIn which basically pushes "I am using YesGraph - best way to ....." to user's timeline?

    Thanks

    • IK

      Ivan Kirigin

      about 3 years ago #

      I don't think I understand your question. Can you email me? ivan@yesgraph.com

      One difference is direct invites vs posts to social networks. YesGraph doesn't promote feed posts. We improve direct invite flows like emails and SMS.

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

Get Weekly Top Posts
High five! You’re in.
SHARE
58
58