5 ms·

From the article: > Of the 400,000 web addresses JPMorgan’s ads showed up on in a recent 30-day period, said Ms. Lemkau, only 12,000, or 3 percent, led to acti

From the article: > Of the 400,000 web addresses JPMorgan’s ads showed up on in a recent 30-day period, said Ms. Lemkau, only 12,000, or 3 percent, led to activity beyond an impression. An intern then manually clicked on each of those addresses to ensure that the websites were ones the company wanted to advertise on. About 7,000 of them were not, winnowing the group to 5,000. They're pretty plainly saying that 388,000 sites were trash. How else to interpret no "activity beyond an impression?" I could print their ads on toilet paper and have the same result. > Do you find it unbelievable that there might be 400,000 legitimate websites in the set of available placements for Chase's ads, without spam? Why? Yes! I do find it unbelievable! The customer did! Everyone should! There just isn't that much valuable content on the Internet. There really isn't. Not valuable for you, not for me, not for Chase, not for anyone with a brain!

stvswn 10y ago
1-5% click through rate on a display ad is pretty normal. If had impressions on 400k sites, the distribution of impressions per site had a long tail, and then you removed all of those that had no clicks, you'd expect to have a small set of placements with clicks. It doesn't mean they're spam sites, it's just math.
- makomk 10y ago
  Yeah. It feels like the New York Times is lying with statistics and fearmongering about "toxic sites" to encourage companies to only advertise with big sites - something that, as basically the biggest name in news in the US, they're positioned to benefit from financially.
  - roganp 10y ago
    If 98% of the sites generate what is rounding error on your sales, then you measure that benefit against the risk that your brand gets associated with a hate speech site (maybe in the nytimes no less). It's math alright, but it means that the brokers are not adding a whole lot of value.
- ssharp 10y ago
  1-5% click through rate on display ads? You're missing a decimal point!
geocar 10y ago
> They're pretty plainly saying that 388,000 sites were trash. > How else to interpret no "activity beyond an impression?" I'd argue the reverse. Sites that generate a consistently high click rate are optimizing for people to leave. Sometimes the click target is obscured, so you'll end up with a high bounce rate. At the end of the day: If you don't prove the relationship between clicks and acquisitions you're going to just fund ad fraud. Meanwhile, if the ad is in-view, and your marketing team isn't an idiot and managed to get 30 or so impressions on someone in a month (and enough someones in your market) then you'll be able to detect (by surveying) an increased brand awareness and preference. This could be valuable because I suspect a lot of people who are considering another credit card will call in on the deal they remember, after talking about it with their spouse, than will click on an ad. Of course, Chase doesn't ask people why they are signing up, so the only attribution data they have is clicks and post-click signups. > I could print their ads on toilet paper and have the same result. Yes you could, but that says more about the tools they're using to measure the result than the usefulness of toilet paper. In this case, someone looked at their sources breakdown in mediaocean, it said the pay-per-click stuff was a winner. Well no kidding. And when I say probably, I should say: I've worked with the ZO guys on the Chase account in my past life. This is probably what happened.
biot 10y ago
When a major sports apparel company sponsors a famous athlete to wear their gear on televised events, they are getting no activity beyond an impression. Doesn't mean the sporting event is trash. It's up to the advertiser to determine whether or not those impressions have value.
- roganp 10y ago
  Isn't that the thrust of the article though? They narrowed their ad spend to whitelisted sites and have not seen a drop in sales. If the major sports apparel company stopped paying for endorsements and saw no drop in sales, you might be tempted to conclude that those endorsements are not good investments.
  - mikelward 10y ago
    I think the analogy is: A company moved endorsement money from lots of lesser-known teams to a few better-known teams and their sales stayed the same.
  - panzagl 10y ago
    Seems like it would be a pretty long term effect though, so the drop off might not be apparent yet.
- Mikushi 10y ago
  The ever elusive brand exposure. Another scam of the ad industry.
  - seppin 10y ago
    it makes sense for a select few, airlines and perhaps an entry country and their tourism industry. that's about it
cm2187 10y ago
Even if these websites were legit, their traffic might be trash. If it's some sort of white supremacist website (the reason for restricting the # of websites), I am not sure that the sort of inbred white trash living in a swamp in Louisiana and consulting this website is a valuable customer base for Chase. Readers of wsj.com probably more so.
- selimthegrim 10y ago
  Apropos of nothing probably, but I've definitely seen Chase branches in those parts of Louisiana.
erichocean 10y ago
> An intern then manually clicked on each of those addresses Even if the intern was paid, that's probably illegal. Interns generally can't perform useful work for a business—they're actually supposed to cost the business money, time, etc. and cannot work on a project the business would normally pay someone to do.
- maxerickson 10y ago
  In the US an intern paid minimum wage or more is just an employee, the special requirements are for unpaid interns. https://www.dol.gov/whd/regs/compliance/whdfs71.htm https://www.dol.gov/whd/regs/compliance/whdfs71.htm

stvswn 10y ago

1-5% click through rate on a display ad is pretty normal. If had impressions on 400k sites, the distribution of impressions per site had a long tail, and then you removed all of those that had no clicks, you'd expect to have a small set of placements with clicks. It doesn't mean they're spam sites, it's just math.

makomk 10y ago
Yeah. It feels like the New York Times is lying with statistics and fearmongering about "toxic sites" to encourage companies to only advertise with big sites - something that, as basically the biggest name in news in the US, they're positioned to benefit from financially.
- roganp 10y ago
  If 98% of the sites generate what is rounding error on your sales, then you measure that benefit against the risk that your brand gets associated with a hate speech site (maybe in the nytimes no less). It's math alright, but it means that the brokers are not adding a whole lot of value.
ssharp 10y ago
1-5% click through rate on display ads? You're missing a decimal point!

geocar 10y ago

> They're pretty plainly saying that 388,000 sites were trash. > How else to interpret no "activity beyond an impression?" I'd argue the reverse. Sites that generate a consistently high click rate are optimizing for people to leave. Sometimes the click target is obscured, so you'll end up with a high bounce rate. At the end of the day: If you don't prove the relationship between clicks and acquisitions you're going to just fund ad fraud. Meanwhile, if the ad is in-view, and your marketing team isn't an idiot and managed to get 30 or so impressions on someone in a month (and enough someones in your market) then you'll be able to detect (by surveying) an increased brand awareness and preference. This could be valuable because I suspect a lot of people who are considering another credit card will call in on the deal they remember, after talking about it with their spouse, than will click on an ad. Of course, Chase doesn't ask people why they are signing up, so the only attribution data they have is clicks and post-click signups. > I could print their ads on toilet paper and have the same result. Yes you could, but that says more about the tools they're using to measure the result than the usefulness of toilet paper. In this case, someone looked at their sources breakdown in mediaocean, it said the pay-per-click stuff was a winner. Well no kidding. And when I say probably, I should say: I've worked with the ZO guys on the Chase account in my past life. This is probably what happened.

biot 10y ago

When a major sports apparel company sponsors a famous athlete to wear their gear on televised events, they are getting no activity beyond an impression. Doesn't mean the sporting event is trash. It's up to the advertiser to determine whether or not those impressions have value.

roganp 10y ago
Isn't that the thrust of the article though? They narrowed their ad spend to whitelisted sites and have not seen a drop in sales. If the major sports apparel company stopped paying for endorsements and saw no drop in sales, you might be tempted to conclude that those endorsements are not good investments.
- mikelward 10y ago
  I think the analogy is: A company moved endorsement money from lots of lesser-known teams to a few better-known teams and their sales stayed the same.
- panzagl 10y ago
  Seems like it would be a pretty long term effect though, so the drop off might not be apparent yet.
Mikushi 10y ago
The ever elusive brand exposure. Another scam of the ad industry.
- seppin 10y ago
  it makes sense for a select few, airlines and perhaps an entry country and their tourism industry. that's about it

cm2187 10y ago

Even if these websites were legit, their traffic might be trash. If it's some sort of white supremacist website (the reason for restricting the # of websites), I am not sure that the sort of inbred white trash living in a swamp in Louisiana and consulting this website is a valuable customer base for Chase. Readers of wsj.com probably more so.

selimthegrim 10y ago
Apropos of nothing probably, but I've definitely seen Chase branches in those parts of Louisiana.

erichocean 10y ago

> An intern then manually clicked on each of those addresses Even if the intern was paid, that's probably illegal. Interns generally can't perform useful work for a business—they're actually supposed to cost the business money, time, etc. and cannot work on a project the business would normally pay someone to do.

maxerickson 10y ago
In the US an intern paid minimum wage or more is just an employee, the special requirements are for unpaid interns. https://www.dol.gov/whd/regs/compliance/whdfs71.htm https://www.dol.gov/whd/regs/compliance/whdfs71.htm