Newspapers everywhere are in agony. Big time. My local daily just announced huge layoffs. Even our good friend the Publicity Hound Joan Stewart, a former newspaper editor herself, has long since pushed her fellow hounds to get ahead of the changing trends and go seeking publicity not just in the (declining) newspaper and print media world, but in booming social media. Such is the trend.
And some of us thought AdWords might be able to rescue newspapers and print ads. Newspapers, for one, sure did. Google had well over 800 news outlets subscribed to their print ad service.
Problem is, advertisers just haven’t been interested.
So Google is killing print ads. Looks like advertisers want to target bigger audiences, growing audiences – and aren’t willing to gamble anymore on advertising media whose results are notoriously hard to measure … in comparison to, say, straight-on Google search ads.
Nobody’s ever asked me about doing newspapers ads, in all of the AdWords coaching sessions and calls I’ve done.
Newspapers won’t ever go extinct, by the way. Video did not kill the radio star, and it didn’t kill movie theaters either. Newspapers will just find a new voice and a new niche crowd to serve. It will happen, mark my words.
Used Conversion Optimizer Yet?
This guy believes that Google’s Conversion Optimizer is the best-kept secret in all of pay-per-click advertising right now.
Did you know there was such a thing? There is. It’s one of those “holy grail of pay-per-click” kinds of tools – at least in the eyes of some – because it provides free in-house automated bid management, infusing boatloads of private Google user intelligence that no 3rd-party service, and no individual, could ever have access to.
The idea: You set a cost per action or target cost per conversion, and multiple times a day Google updates your bid prices and even the frequency and geographic placement of your ads in order to ensure that you get maximum conversions at or below the target price you set.
How do you set it up? In your campaign settings under “Networks and bidding” click on “Change bidding strategy” and you’ll see it listed.
Oh – and you can’t use it on campaigns that have had fewer than 30 conversions in the last 30 days. Volume required.
Fans swear it’s the greatest thing to happen to PPC since automated split testing. And that it’s significantly increasing profitable conversions.
Others have other opinions.
Such as, “I want Google to know and be keyed into as little of my back-end sales funnel as possible.” So it’s a double-edged sword.
The fact that it’s automatically monitoring conversions, rather than just clicks and budget, keeps the fox away from the henhouse.
Test it. Post a comment below and tell us if it’s worked for you. Kick it on in one of your less-critical campaigns and see how it compares to the conversions you’re already getting.
Broad Match: Getting Nasty?
Here’s another item for you to post a comment about: Is broad match getting worse? Tell us.
We’ve seen it. Our associates have seen it. Our customers have seen it, and complained loudly. Costs per conversion on broad-match terms are going up, quality traffic and conversions are going down.
This guy has seen it and swears it’s an actual bug in Google’s system.
Let’s say you sell wagons. You’ve got a red wagon ad group with read wagon ads and red wagon all-exact-match keywords. And in another group, a “wagon” ad with broad-match “wagon” keywords. The glitch? The bug? A user searches verbatim for “red wagons” … and go figure: your generic “wagon” ad shows up.
Despite the fact that you’ve already got a high-quality fine-tuned red wagon ad group with exact-match red wagon keywords!
So what do you do? First, post and tell us about it. Next, go with a fine-tooth comb through your top 20% most productive ad groups and run a careful inventory of what the broad matches there are doing. Need to pause them? Do peel-and-stick with them? Cut them out entirely? Then do so.
But not without double-checking that you’ve already got a good, solid, near-exhaustive list of phrase- and exact-match keywords to hold their place for the time being.
And wait for Google to fix the bug. Or the economy to get better so that they don’t have to “adjust” their algorithms to meet revenue projections. (Ahem.) Whichever comes first.
Google Broad Match: A Change for the Worse:
Entering the “Conversation.” Literally.
Here are two different tools for “entering the conversation inside your customer’s head,” which is, as we have long been telling you, is the cardinal rule of sound marketing and killer Google ads.
First, www.SamePoint.com, which you can use to search on any topic and see what the current blog posts and wiki articles have to say about it. “What’s the latest buzz?” in other words.
Next, want to learn some slang? Try this tool. Great one to use to figure out the casual lingo that’s getting tossed around by people in your market. And learn what it means. And use it.
I searched on “snowboard” in the lexicon, and apart from the obvious terms found these too:
shred
steeze
skate
steez
jib
park
rail
gaper
slide
jibbing
pizza slice
snowplow
burton
tow board
toe-board
gnar
steezy
wakeboarding
postbang
snowplow bomber
prebang
Tells me I’ve got a thing or two to learn before I start selling to snowboarders.
AdSense Ad of the Week
Content network expert Shelley Ellis pointed this one out to me:
Not a bad suggestion, eh?
To your AdWords success,
Bryan Todd
Share This Post
3 Comments on “Google Kills the Newspapers, PPC's Best Kept Secret? and the Broad Match Curse”
Nice Ad Bryan!
Hi Brian,
great post. I have a full article on Conversion Optimizer at
http://www.davidnrothwell.com
I’m seeing good results with Conversion Optimizer – it’s delivering consistent conversions under our target CPA, so it’s really working.
You still have to work with it though, and keep monitoring and testing. And since you set a conversion target, it’s best to have ad delivery Optimised, not Rotated (not what we’re used to!)
Adding to the comments above, it seems best to have very granular ad groups sometimes with only a keyword or two in there.
Also, for Content, since this works at ad group level, each ad group must be very tightly themed with a small group of keywords. With CO in charge, if target CPA is not being met on Content, your entire ad group will be disqualified with all its keywords.
I have tested the conversion optimizer on a content network campaign and the results were disappointing: traffic dried up completely. ( Well I could have set a ridiculously high cost per conversion but that would not be useful of course…)
I have then switched back to another strategy, namely milking google’s reports with software for every piece of info which is there, in order to identify loosers and champions in terms of countries and websites. And then eliminate those loosers.
Plus do peel-and-stick with champions. On the content-network, this is more a 99/1 rule than 80/20.
Plus do some manual adjustments with bid-prices.
This worked quite well for me. ( cost per lead reduced by about factor 7 )
Plus I made an important observation about bid prices in the content network:
Contrary to what you may expect, the cost per conversion may go up if you reduce bid prices. I guess the major reason is that you may disappear from winner websites and only continue to show up on looser websites. ( “winner” and “looser” meant with respect to cost per conversion with advertising on them, not with respect to what those websites do. )
So, properly managing bid prices on the content network is *really difficult* and may get you surprizing, counterintuitive results.
Probably the way to go is to kind of split-test different price points. (Real random-rotation split-test is impossible for bid-prices of course but there are ways to approximate that. I didn’t yet try this out but I probably should.)
***
As to the conversion optimizer: It seems to only adjust bid-prices for clicks on an adgroup level and seems to apply an assumption that cost per conversion and the click-bid-price have a simple relationship in the form that lowering the bid per click will lower cost per conversion and increasing the bid per click will increase cost per conversion.
But this assumption is not valid on the content network, so the algorithm fails.
But it should do a completely different thing:
It should optimize cost-per-conversion *and* optimize the selection of the right splittesting-ad in each “cell” where each “cell” has at least the following coordinates:
– Country ( or subterritories in each Country )
– Website ( and possibly specific URLs on a website )
– day of week
– time of day
Because within one cell, the assumption stated above is probably true.
Of course if traffic within one cell is not large enough, then do the same optimization for groups of cells where a group can e.g. be a set of cells where some, but not all coordinates have the same value, or grouping similar values such as grouping all coutries in one continent etc.
***
Maybe my observations are not correct. If anybody has different observations please post a comment… I am very interested in that.
Especially if there are reasons to believe that Google’s conversion optimizer actually does optimizations on a cell-level, provided that there is enough traffic.
Not enough traffic may have been one problem with my campaigns, but that can only be a part of the picture because I got enough traffic for getting statistically significant group-of-cell data which I used for pretty crude optimizations in terms of complete elimination of group of cells which did not perform sufficently well.
So, I guess that Google’s conversion optimizer algorithms actually have some room for improvement… and that’s an understatement. :-)
But there may be a reason for that: Maybe Google actually does sophisticated cell-level optimization, but optimization of their income per impression. And it is not possible to optimize two functions at the same time. There’s got to be a conflict of interest which cannot be resolved on an algorithmic level.
And since Google essentially does not have a serious competitor (yet), there is not much point in optimizing value for their customers beyond the point necessary for maintaining their (almost)-monopoly.
***
If you advertise on search, the whole picture may look completely different.
Search has much fewer variables than Content, so conversion optimizer may actually work well there even without sophisticated cell-level optimizations. But you should watch your numbers closely…
-Chris
P.S: I have written most of this post before Perry has posted his video. And I think my observations fit well with the major picture which Perry has described…