Effects of geotargeting
January 15, 2010 10:30 am geotargeting, new featuresHere we’ll be looking at how geotargeting will be affecting our members.
A lot of the effects are pretty obvious: advertisers will be happy with new and better ways to target their advertising! For publishers, though, the effects are a bit more subtle. Sure, each publisher now has geotargeted ad boxes on their sites, but what does that really mean?
For that, we’ll go back to some basic economics.
As you may have realized, advertising functions in a supply and demand system. You (as a publisher) have the supply, and advertisers represent demand. If you’ve got one advertising spot and two advertisers, then they compete for the space, and that drives the price of advertising on your site up! In contrast, if you’ve got twenty advertising spots and two advertisers, then you’ve flooded the market, and your advertising will probably go for really cheap.
When you apply to become a publisher, we check out your site, and when you’re approved we let you place up to five different advertising areas on your site. One of the reasons we have this ceiling is to help guide new publishers! Early on we had publishers who got over-enthusiastic and put way too many ads all over the place. This is like flooding the market with supply – plus, you’ve diluted the value of advertising on your page even more by putting so many ads on it. Any advertiser has to now also compete with every other ad on the page.
With geotargeting, each ad box is now actually four separate auctions: one for Canadian traffic, one for American traffic, one for European traffic and one for the rest of the world. Isn’t the flooding the market with supply? The answer is no, not really. There’s not actually four times the ads appearing on your site, because only one is showing at a time to a given viewer. Plus, each regional auction is a different SORT of product, with a different value. Someone who wants to advertise to Canadians isn’t like to be interested in the European market as someone in Italy is.
Imagine an ad box that get 100,000 hits/day and gets bids of about $100/day. And let’s assume, because it makes the math easy, that through fantastic chance, each region gets exactly the same amount of traffic. Here’s what you’d expect to happen!
America gets 25,000 hits at $25/day,
Canada gets 25,000 hits at $25/day,
Europe gets 25,000 hits at $25/day,
and traffic from everywhere gets 25,000 hits at $25/day.
For publishers, this is great, because they’re earning just as much as they did as before – and this is assuming that not a single advertiser is willing to pay a premium for regional traffic, which seems unlikely. There’s an important mental change happening for advertisers here too: they can get a day of regional traffic on your site for only $25, which is a much smaller number than the $100 they were facing before! Of course, if they wanted to buy ALL regions, it would still cost them $100 – but this smaller number is psychologically important. It seems cheaper, and it lets advertisers test the water without having to place a (maybe scary?) $100 bid.
So that’s good too! That’s two factors that work to make regional traffic more valuable than merged traffic.
There is, however, one scenario where geotargeting could hurt certain publishers, and we wanted to make sure it was clear and obvious, so that if you find yourself in this scenario, you can preemptively take care of it. If you’re a publisher with extremely low bids on your ad box, you could end up in a scenario where, before geotargeting, an advertiser was willing to pay 1 cent a day to be on your ad box, but after geotargeting, they’re less willing. Why? Because one cent is the lowest an advertiser can bid, and with your traffic now split across four regions, advertisers might not think it’s worth at least a penny per region to bid on your site.
Are you pooched?
Well, not really. We made a simplifying assumption above that regional traffic is split evenly: in real life, it’s really unlikely that this would be the case. It’s actually way more likely that one or two regions would make up the bulk of the traffic – which means that the region we’re getting our traffic from will maintain demand and stay at that 1 cent a day!
So that’s good – that helps us! In the vast majority of cases, these ad boxes that are already earning a cent a day will continue to do so. And in our remaining regions, where bidding is low, we can supplement with custom “Your ad here” images. These images can either either encourage bidding, or link to other places you’ve chosen: maybe to your shop, for example. You can choose a different “your ad here” image and link for each region, which is handy!
And if you’re a publisher who still thinks there might be problems, you can always try consolidating your ad boxes. By dropping the number of locations available on your site and reduce the number of spots available in each ad box, you’re increasing the value of your page: each change reduces the supply, which means that advertisers who want this advertising will pay more. You can also put your ad box on more pages of your web site, and the increased exposure will help increase its intrinsic value.
So, this is the one boundary case that you may want to be aware of. The vast majority of our publishers should have no issue, and the added value geotargeting gives advertisers will translate to higher bids on their sites. If you find yourself in the worst-case scenario we described above, consolidating your ad boxes now will increase the value of advertising on your site, and help you out. It’s worth doing, as too much supply means that your ad boxes may not perform as they should!
January 17th, 2010 at 3:03 am
I’m extremely surprised and concerned that you’ve separated Canada and America; the markets are virtually the same as far as both advertisers and publishers are concerned, and the campaigns they run are usually interchangeable due to releases and prices being virtually the same.
Canadians represent 0.4% of the world’s population. According to a recent study, they also spend far less online than the average American. Why on Earth, other than being a Canadian company (which is something we both have in common), have you selected this minute population over such markets as the U.K. (biggest online spenders), Japan (almost 1/2 of America’s population), or China (1.4 bl residents)?
Perhaps you should group the markets into North America, Europe, Asia, and “rest of the world.” At least that way it makes more than an ounce of sense whilst remaining extremely basic.
January 17th, 2010 at 6:13 am
Hi E!
I appreciate the sentiment! But we didn’t pull these markets out of nowhere. We chose these initial regions by looking at two things: where our traffic is coming from – which breaks down into where our publishers are already reaching – and where our advertisers are targeting.
US and Canada make up a large amount of our traffic. It would make sense to group them together, except for the fact that for many advertisers, it’s still an important distinction. Most companies operating in both countries have separate advertising budgets for them, and the Canadian division is not allowed to advertise in America, and vice versa. We also found some other legal restrictions, where for example if you’re running a contest in the States, it may actually be illegal for you to show the contest advertisement to non-Americans, who can’t take part in the promotion.
Scenarios like that make it sensible to separate out American traffic from the rest, which helps to leave Canada as a separate entity. As we move forward and get more advertisers and publishers in the countries you mentioned, it will be relatively simple to add them as new regions, separated out from the “everywhere else”. However, at the moment, we don’t have the reach into China or Japan – the two examples you mentioned – to make separate regions make sense. The traffic just isn’t there yet, nor are the advertisers – despite their population on the planet. To put it another way, the deciding factor wasn’t population on the planet, it was population on the planet that’s using Project Wonderful. Before we’d announced this change, we’d received quite a few requests for US/Canada geotargeting – outweighing the requests we received for geotargeting in every other country on the planet.
I hope this gives a little justification of where we came from! And as I said earlier, it’s always possible to add more regions in the future, just as we’re adding these now.
January 18th, 2010 at 2:34 am
You’re not serving your own audience; you’re serving others’. A glance at my site, for instance, put U.S. traffic above and beyond Canadian traffic (which is actually rather far down the list). There are many other publishers who will have this common metric, regardless of what you’re tracking internally. For us, our Canadian ad slot will be worth a tiny fraction of what the others are, and even if publishers are receiving strictly North American traffic, the Canadian group will still be worth, at most, one tenth – and that’s before you consider the fact that Canada is still many years behind the US in terms of online spending.
If you’re going to divide it, at least do it according to the publisher’s traffic. If the traffic receives the majority of their traffic from North America, then divide it into the US, Mexico, Canada, etc, and Other. Consider the publisher’s largest regions and let them capitalise on it.
Your new setup will create a lot of redundancy, and one of the following will occur:
1. The publisher will just switch geotrageting off.
2. The publisher will grin and bear it.
3. The publisher will find an alternative solution.
One reason Project Wonderful itself doesn’t have many “outside” and large advertisers is because they’re distanced by your system as it currently stands (US dollars, PayPal only, near-to-no advanced statistics). By segregating it and clustering everything other than the U.S. and Canada into an “other” junk bin, you’re further alienating advertisers from those regions and limiting your chance of growth.
To return to my initial point; Canada and the US are virtually interchangeable when it comes to marketing. I say this with a lot of experience and as a veteran of the industry. Your average Project Wonderful advertiser doesn’t have an office in Canada as well as the US, since the markets are so close and so similar (in fact, I’d wager that the average Project Wonderful advertiser doesn’t even have an office at all). And since they’re advertising on Project Wonderful already, they’ve proved that they clearly don’t care about reaching a Canadian or two when serving ads to the US, for reasons mentioned in my previous comment.
Your advertisers are going to be far more concerned about reaching our heritage country (my other example which you seemed not to notice), since not only do far more people live there, but they spend more time and money online, their product prices/releases are completely different, their marketing approach is extremely different (less in your face, more subtle), and they’re extremely accessible. Your average Project Wonderful advertiser who works on a professional level (i.e. they’re a business with employees and an interest in reaching more than a single region, which is what this update is all about), likely has operations there in some shape or form.
Referring to your “we’d received quite a few requests for US/Canada” stance: it’s great that quite a few members suggested it, but your members aren’t always right. As Herbert Bayard Swope said, “I can’t give you a sure-fire formula for success, but I can give you a formula for failure: try to please everybody all the time.”
I’m not the only person complaining about your choice of countries/regions and the way you’ve grouped them. I have friends who work in marketing (some higher up than others, and somebody who I’d consider to be one of your largest and most professional advertisers), and we all talk.
I’m sure you’ve done your research, but I think you need to take this back to the drawing board, and then test it privately with a small focus group to discover its strengths and weaknesses, rather than suddenly mindlessly implementing it as soon as it’s ready.
I’m not looking for further justification. Please; either do it right, or don’t do it at all.
January 18th, 2010 at 4:22 am
Hi E!
Well, if you’re not looking for further justification, it looks like I’m not going to be able to convince you. All I can say is give it a try when it launches, and I’m pretty certain you’ll be (pleasantly) surprised.
Thanks!
January 25th, 2010 at 9:02 am
looks promising