> Anybody willing to share what it's like to work on a project like that?
I used to work at Business Insider and we definitely had similar problems. It basically boils down to marketing & advertising being more important to publishers than technology or user experience.
It's organizationally difficult to resist every marketer who wants to add "just one line of JavaScript" for their pet project. Then there's the fact that you're beholden to ad networks and can't do much more than whack-a-mole when they serve junky ads. Probably once a day we'd notice a bad ad and have to block it from being served.
At one point, we got management buy-in for me to work on performance issues. We managed to improve performance by ~90%, but the bulk of the project was just exhaustively listing every script on the page and axing those which didn't have a good justification. Amazingly, for a majority of the scripts, nobody even knew why there were there still (they were often added to please an old advertiser who wanted a specific measurement).
It's hard to place blame. Developers don't have the political clout to prevent the bloat and are often under-resourced as it is. Marketers want to gather metrics, but with the small development teams at most media companies it's easier to rely on third-party vendors over performant in-house solutions. Management recognizes the problems, but without marketing and advertising there's no business model.
Personally, I think ad networks are the ones who should be taking the lead here: they're big enough that they can afford to block bad advertisers without endangering their profitability. They should be investing much more in automated solutions to detect and block bad/heavy ads. Unfortunately, there's not really an economic incentive for them to do so: users blame publishers for slow pages, not networks. And ad blocking is much more detrimental to publishers than the networks.
Someone needs to give the marketing team a 'budget' just like how they have a financial budget they must stay within.
300ms average load time added by the marketing team, or 200kb JavaScript, or whatever. Split it all up however you like, and let them fight amongst themselves for what goes live on the site.
That won't work for the developers who lack the authority to make this a rule, but that's what you go to management with along with any of the various studies showing how important speed is to users.
I think ad networks are the ones who
should be taking the lead here
I don't understand why one ad network doesn't take this seriously and have strict rules and enforcement against "bad/heavy ads", and then use that fact to differentiate itself from competitors. The resulting conversion of sites to their network (perhaps with a certification of compliance badge that sites can display) would provide that economic incentive... and perhaps steer the broader market in that direction over time.
I used to work at Business Insider and we definitely had similar problems. It basically boils down to marketing & advertising being more important to publishers than technology or user experience.
It's organizationally difficult to resist every marketer who wants to add "just one line of JavaScript" for their pet project. Then there's the fact that you're beholden to ad networks and can't do much more than whack-a-mole when they serve junky ads. Probably once a day we'd notice a bad ad and have to block it from being served.
At one point, we got management buy-in for me to work on performance issues. We managed to improve performance by ~90%, but the bulk of the project was just exhaustively listing every script on the page and axing those which didn't have a good justification. Amazingly, for a majority of the scripts, nobody even knew why there were there still (they were often added to please an old advertiser who wanted a specific measurement).
It's hard to place blame. Developers don't have the political clout to prevent the bloat and are often under-resourced as it is. Marketers want to gather metrics, but with the small development teams at most media companies it's easier to rely on third-party vendors over performant in-house solutions. Management recognizes the problems, but without marketing and advertising there's no business model.
Personally, I think ad networks are the ones who should be taking the lead here: they're big enough that they can afford to block bad advertisers without endangering their profitability. They should be investing much more in automated solutions to detect and block bad/heavy ads. Unfortunately, there's not really an economic incentive for them to do so: users blame publishers for slow pages, not networks. And ad blocking is much more detrimental to publishers than the networks.