Company Blog
Product Blog

Why We Build Our Own Stuff

Last night at NYBeta I had some folks asking why we do all of our own development. Why code an ad server when there are ad servers? Why track analytics in house when there are companies that do that?

It sparked an interesting discussion in Campfire this morning. This is why:

  1. We have a team who can.
  2. It’s faster, and we accrue less technical debt.
  3. Any company who has a long term focus would build their core products in house. If you aren’t pushing the boundaries of existing solutions, what is your company doing?
  4. It’s fun.
  5. We know how to fix it when it breaks.
  6. It breaks less.
  • http://www.paymo.biz Jan Lukacs

    We do the same to some extent, however building everything in house means doom in the end. Externalization works great in a lot of cases, it’s more affordable, less exepensive and breaks less :)

    • http://buysellads.com Todd Garland

      I guess it all depends on what you’re externalizing. We’ve always taken the stance of not externalizing anything that’s core to the business and user facing.

    • http://icondeposit.com Matt Gentile

      I have to disagree with that comment that Jan Lukacs made, look at Facebook and Twitter for example, it’s pretty much guaranteed that they will NOT be doomed in the end. BSA is sort of headed down the same path as them. Making everything “In House” makes your company more unique and makes you stand out from the rest of the trillions of sites on this planet, pretty much guaranteeing success.

      • http://icondeposit.com Matt Gentile

        Put it this way, would you want someone else in control of YOUR site?

  • http://silodrome.com James McBride

    We love BSA’s in-house approach but we have had some issues with the BSA analytics calculating about 20% less pageviews than Google Analytics or WP Jetpack. Are there any tips on how to ensure the numbers are more accurate?

    • http://buysellads.com Todd Garland

      @James – I have a couple things that might help us better understand the 20%:

      1. There are some pages on your site where GA is installed but BSA is not. Yeah, these pages are relatively insignificant, but perhaps it adds up – pages like about, advertise, etc. This probably doesn’t get us to 20%, but it might get us to 15% or 16%.

      2. The ad you have on the article detail pages at the top – do it’s numbers roughly match to GA and Jetpack? This might give us a good benchmark for #3.

      3. It might be worth doing a test on the site to see how many visitors are using adblock: http://tutorialzine.com/2011/12/how-to-block-adblock/ fire a GA tracker when you detect that someone is using adblock to get a sense for the numbers.

      4. When I watch the waterfall on your site in firebug/net tab the bsa.js is loaded before the ga.js, but then there’s a facebook, quantserve, fbcdn, and apis.google.com that gets loaded before we complete. So, there might be something there that also helps explain this. We’re poking around with this now to see if there is anything we can do.