5 ms·
Leave the math to us: Advertisers increase profits with Conversion Optimizer
- pg 17y agoOdd to think it's taken till now for someone to build the whole loop. http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PALL&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.htm&r=1&f=G&l=50&s1=6,631,372.PN.&OS=PN/6,631,372&RS=PN/6,631,372 http://patft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sec...
- tptacek 17y agoA patent?! He's a witch! BURN HIM!
- wildwood 17y agoNot so odd, really. Building the whole loop requires a fair amount of integration and trust between the search engine and the advertiser, which is often easier said than done. Didn't Google have a CPA beta last year for AdSense? It's interesting that they're coming up with a new product there now.
- jimbokun 17y agoSo...does this mean Google is infringing your patent?
- pg 17y agoYou mean Yahoo's patent. I think it may have played some kind of role in some negotiation long ago between Google and Yahoo, but I doubt it has any effect now.
- dminor 17y agoTried this for about a month and ended up with a drop in conversions and slightly higher cost/conversion, so it doesn't work for everyone.
- calambrac 17y agoThis puts a lot of consultants out of business :/
- axod 17y agoThe adwords conversion optimizer has been around for at least a year, maybe 2 or 3. So I'm not quite sure what's changed now. I tried it with really good results on a website where I wanted to only pay a specific price for each new user.
- patio11 17y agoI started using it in September 2007, which was probably within a week or two of it launching. The major visible improvement since is that the system needs conversion data prior to you being able to apply CO. Back when it launched that was 300 conversions in the 30 days prior to hitting the switch, which limited its utility to small businesses. Over the years that number went down -- it was 200 for a while, and now it is 15.
- robk 17y agoThey're just trying to raise awareness of the Conversion Optimiser tool. Surprisingly, there are quite a few small to medium sized advertisers who don't know about it even though they are perfect fits for the product. In general, due to quality score and various keyword quality concerns, I've found many advertisers are terrified to change their bread-and-butter campaigns for fear something will go awry and they'll be penalized somehow. Makes sense really. Shame that Google can't do a better job helping people over this fear.
- encoderer 17y agoThere are a lot of CPA Networks out there already and I'm not sure Google brings anything new into the table with the offering. It obv works for advertisers wanting CPA and only willing to advertise with Google, but for the rest, it's just a me-too effort from Mountain View.
- URSpider94 17y agoI bet many marketing departments would be happy to switch to CPC, even if it increases costs, because they are paying for results, not traffic. From Google's perspective, this also gets them out of the business of auditing clicks. Giving them insight into conversions at the customer side also gives them another metric to evaluate the quality of content pages where they are placing ads, and potentially for refining their search algorithm.
- miracle 17y agoWhat happens if you only send 1/3 of the conversions to google? Do you have to pay a higher conversion price then?
- patio11 17y agoThe key thing to understand is that Conversion Optimizer is just a user-friendly wrapper around cost per click bidding. It will look to you like you're paying cost per conversion, but you're not, you're paying cost per click. So, if you start "shaving" conversions (failing to report them), here is what happens: 1) I tell Google I am willing to pay 25 cents per conversion. My website converts 25% of clicks, which for the sake of easy math we'll say is constant. (Conversion Optimizer is a good idea because it is not constant, in particular, because it tends to be different for every source of clicks in ways which are very time-intensive to track yourself.) 2) After Google's system has figured out the above two bits of information, it will figure out "OK, he can pay about 6.25 cents per click. In general, I should bid that much on his behalf in the auction -- avoiding clicks priced at 8 cents, and 'backing up the truck' on clicks priced at 4 cents.") 3) Then you start shaving conversions, so that while your site continues to convert at 25%, you only report a conversion rate of 20%. Google figures "Wow, sucks to be him, he can only afford clicks priced at 5 cents now and still make his 25 cent per conversion desired price. OK, I'll stop bidding on clicks between 5 and 6.25 cents, and continue bidding on clicks priced below 5 cents." Thus, lying to Google doesn't particularly hurt them, it only hurts you. This is in sharp contrast to real cost per conversion (cost per action = CPA) advertising like affiliate advertising, where you pay for every conversion you report, so shaving conversions means you get money from customers but get to keep it for yourself.
- brown9-2 17y agoIf you don't ship internationally, Conversion Optimizer will learn that clicks from international users don't convert and will eliminate traffic from outside the United States. I'd be curious to know how they would determine something like this (Googlebot-scraping the html of order pages and checking presence of form data?) - or if it's just marketing hyperbole.
- gcanyon 17y agoWouldn't it be a simple optimization? Google isn't learning that you don't ship internationally, it's just learning that people from other countries don't convert well for you. It's probably tracking all sorts of metrics on the users, and constantly trying to figure out what profiles work best for you. They don't have to know or care why.