6 ms·

This article misses one key point: sales and product people. Specifically, sales and product people who neither understand nor care about user experience. Thes

This article misses one key point: sales and product people. Specifically, sales and product people who neither understand nor care about user experience. These are the kinds of people who say yes to every deal, since the only metric they are judged by is quarterly income. Typically these people are also (rightfully so) terrified of losing their jobs. So they close tons of ad deals, by which point it's too late for the developers to stop them. Back when I worked for Newsweek they were the biggest barrier to a quality user experience. The site developers wanted to built and the site had to was the main reason I left that industry behind.

michaelbuddy 11y ago
absolutely. calling the devs incompetent because of a bunch of ad network bullshit is not cool. Most of the JS probably comes in directly from the ad system that devs are unable to improve because they are proprietary. I mean just look at scripts like "Add This". The product people want social sharing easily done. Add This gets installed then suddenly the page hangs up an extra 15 seconds loading god-knows-what of 200 legacy social network scripts and trackers.
- pbh101 11y ago
  Thompson was not saying the devs are incompetent. The dichotomy he made was that either the devs are incompetent or there is something else going on, and then proceeds to delineate what that "something else" is. And while he doesn't explicitly state which he thinks is true, the premise of the article is that it is the latter.
smacktoward 11y ago
But this is a double-edged sword, is it not? I mean, "the site the developers want to build" will invariably be a site with no ads whatsoever. It loads so fast! Look at how clean the code is! Then everybody loses their job because there's no actual revenue. In theory, reconciling these opposing forces should be what top management does. In reality, top management is usually too terrified to do so, because they're subject to the same pressures the sales people are: they have corporate masters of their own, demanding a revenue chart that always drives up and to the right, who will cheerfully throw them overboard the moment they appear to not be 100% behind that goal. So it's ads ahoy and UX take the hindmost.
- notjustanymike 11y ago
  Agreed - and there have to be ads on the site. The problem is that user experience is always sacrificed for short term profit. The same people who sold a 6 month 300x600 scabies home page takeover (seriously...) wanted to know why our returning user rate was so low. Developers want to build and experience people will return to, while Sales and Product will sacrifice anything for short term gain.
- emodendroket 11y ago
  Well it'd be a pretty unusual graph if it did not go to the right.
  - Chinjut 11y ago
    Stop drawing revenue, and the graph _will_ eventually stop going to the right...
- morgante 11y ago
  > I mean, "the site the developers want to build" will invariably be a site with no ads whatsoever. I think that's shortchanging developers. They know where the money is coming from and if the company has good incentive structures (ie. equity) want to make sure revenue grows as well. Fortunately, developers are also aware of the cost which is associated with every third-party tool included on the page. That's why it's really important to include someone technical in any deal conversation as early as possible, and they need to have the ability to block a deal which would substantially degrade the user experience. They can also weigh different providers by their implementation cost to users, which is crucially important: a sales guy loves to hear "just one line of JS" but I'd much rather use a REST API. The key to getting this sort of control and buy-in is demonstrating to management the fact that page load time has a direct effect on traffic (and thus revenue).
hibikir 11y ago
This hits on the big barrier to human collaboration in firms: Role separation that has little to do with company goals. We see a small microcosm of this in software development. Big, lumbering software firms that spent millions of dollars getting very little done end up with teams that have separate DBAs, frontend programmers, backend progrmmers, testers, business analysts, project managers, software architects, build specialists... Lots of jobs, lots of people, and none of them really has value to the company in their mind. In comparison, a small company replaces that entire army with two people, and delivers cheaply and quickly. Only through shared goals we ever get anything complicated done. This is what agile, lean and devOps are really about. Most of the rest is overperscriptive marketing. Focus more on building the right product, than in building the product right.
lebek 11y ago
Where is the convincing evidence that user experience improvements (e.g. reduced load times) increase profit more than more ads? Maybe that would persuade sales teams to be more selective.
- greglindahl 11y ago
  The metric is called LTV, lifetime value. The problem is that it takes a while to measure it. It's easy to see that your short-term obnoxious ad campaign brings in more money in an A/B test. It's much harder to figure out that it lowers the LTV over a week or month.

michaelbuddy 11y ago

absolutely. calling the devs incompetent because of a bunch of ad network bullshit is not cool. Most of the JS probably comes in directly from the ad system that devs are unable to improve because they are proprietary. I mean just look at scripts like "Add This". The product people want social sharing easily done. Add This gets installed then suddenly the page hangs up an extra 15 seconds loading god-knows-what of 200 legacy social network scripts and trackers.

pbh101 11y ago
Thompson was not saying the devs are incompetent. The dichotomy he made was that either the devs are incompetent or there is something else going on, and then proceeds to delineate what that "something else" is. And while he doesn't explicitly state which he thinks is true, the premise of the article is that it is the latter.

smacktoward 11y ago

But this is a double-edged sword, is it not? I mean, "the site the developers want to build" will invariably be a site with no ads whatsoever. It loads so fast! Look at how clean the code is! Then everybody loses their job because there's no actual revenue. In theory, reconciling these opposing forces should be what top management does. In reality, top management is usually too terrified to do so, because they're subject to the same pressures the sales people are: they have corporate masters of their own, demanding a revenue chart that always drives up and to the right, who will cheerfully throw them overboard the moment they appear to not be 100% behind that goal. So it's ads ahoy and UX take the hindmost.

notjustanymike 11y ago
Agreed - and there have to be ads on the site. The problem is that user experience is always sacrificed for short term profit. The same people who sold a 6 month 300x600 scabies home page takeover (seriously...) wanted to know why our returning user rate was so low. Developers want to build and experience people will return to, while Sales and Product will sacrifice anything for short term gain.
emodendroket 11y ago
Well it'd be a pretty unusual graph if it did not go to the right.
- Chinjut 11y ago
  Stop drawing revenue, and the graph _will_ eventually stop going to the right...
morgante 11y ago
> I mean, "the site the developers want to build" will invariably be a site with no ads whatsoever. I think that's shortchanging developers. They know where the money is coming from and if the company has good incentive structures (ie. equity) want to make sure revenue grows as well. Fortunately, developers are also aware of the cost which is associated with every third-party tool included on the page. That's why it's really important to include someone technical in any deal conversation as early as possible, and they need to have the ability to block a deal which would substantially degrade the user experience. They can also weigh different providers by their implementation cost to users, which is crucially important: a sales guy loves to hear "just one line of JS" but I'd much rather use a REST API. The key to getting this sort of control and buy-in is demonstrating to management the fact that page load time has a direct effect on traffic (and thus revenue).

hibikir 11y ago

This hits on the big barrier to human collaboration in firms: Role separation that has little to do with company goals. We see a small microcosm of this in software development. Big, lumbering software firms that spent millions of dollars getting very little done end up with teams that have separate DBAs, frontend programmers, backend progrmmers, testers, business analysts, project managers, software architects, build specialists... Lots of jobs, lots of people, and none of them really has value to the company in their mind. In comparison, a small company replaces that entire army with two people, and delivers cheaply and quickly. Only through shared goals we ever get anything complicated done. This is what agile, lean and devOps are really about. Most of the rest is overperscriptive marketing. Focus more on building the right product, than in building the product right.

lebek 11y ago

Where is the convincing evidence that user experience improvements (e.g. reduced load times) increase profit more than more ads? Maybe that would persuade sales teams to be more selective.

greglindahl 11y ago
The metric is called LTV, lifetime value. The problem is that it takes a while to measure it. It's easy to see that your short-term obnoxious ad campaign brings in more money in an A/B test. It's much harder to figure out that it lowers the LTV over a week or month.