Google Conversion Optimizer: The Controversy

I’ve been using Google Conversion optimizer for quite some time and the main feature is that it saves lots of time.

Conversion Optimizer is like printing money!

I was an early user of Conversion Optimizer and over the last two years, Google has done a tremendous job of making it better.

I can also attest that it works and works really well.

Since my background is affiliate marketing, I started using the Content Network and the Conversion Optimizer to literally print money on demand.

I bid on a CPA basis and keep the difference between what I make per conversion from the publisher and what I actually pay Google for each conversion.

After all the Google banning and hoopla surrounding affiliate marketing on Google, I have since moved into performing this service for clients, teaching and training this technique, and using it for my own products.

A couple of keys to making Conversion Optimizer work is:

1. Focusing on increasing the conversion rate of your landing page. Over time this will allow you to increase your ROI because the better you can get at converting the traffic, the less it will cost to get the conversion.

2. Knowing how much a lead is worth. Most people don’t take time to determine the true value of a lead, which means they can’t make the proper bidding decisions on a CPA basis.

Perry, you are absolutely correct about how this will be the future of how we will buy and sell traffic online.

I’m sure you already know that many of the media companies are already ahead of Google on this and you can buy their traffic on a CPA basis.

I was thoroughly impressed with how well you were able to explain how this process works.

Great stuff as always!

Tommie

Perry,

Great video but i think something should be clarified to those who may not understand Conversion Optimizer fully…

You don’t “just” pay per action with the conversion optimizer.

In other words, you can send 4 million people to your page and pay for only the 10 leads you gathered because your process sucked.

You still pay per click technically.. if your sales/leads process sucks, and you don’t get leads, you’re still paying for the traffic.

It’s not a full CPA model per sae.

It will auto adjust bids for you on many different levels, even down to operating system however, you still pay for clicks.

Another thing I would caution people about is to make sure your google conversion counter is working properly. I’ve seen instances where google only counts as little as 60% of the conversions, and thus Conversion Optimizer totally drops your CPA bid (due to false data) and traffic tanks.

Anyway, great video Perry

Peter

Perry,

While I don’t have a problem in principle with Google’s stuff, we’ve found that conversion optimizer does more harm than good.

We tend to see alot of lost impressions that drive qualified clicks with Conversion optimizer. We’ve got case studies where just turning it off increased conv. rates dramatically.

As far as “automatically managing CPAs”, we’re transitioning on to a bid management tool with alerts to handle that. (Although manually checking every once in a while seems to do fine once you’ve got the campaign running steady. For some reason, Google just doesn’t seem to be up to par.

Hi Perry,

I wrote the article you referenced.

Thanks for the taking the time to read it and make a whole video in response! A healthy debate is always welcome.

First, I agree with you on your primary point: We only buy paid search advertising to for conversions/sales. We all need to be optimized for maximum ROI.

I’m guessing from the emphasis of your video, that your concerned with my viewpoint on Google Conversion Optimizer (GCO).

As Peter pointed out, you do not in fact pay per action when using GCO. You pay per click and give Google control to determine what the bids for individual keywords will be based on their historical and predicted conversion rate.

I don’t have a problem with Google offering GCO. For a small business with limited time and expertise to manage PPC, it may be a worthwhile option. Also, it’s optional. So, if people don’t like it, they don’t have to use it.

Tommy had a good experience, David did not. I’d like more folks to share their non-Google PR stories.

That said, I don’t believe GCO is a suitable option for larger advertisers with higher budgets. I believe optimizing based on last click attribution, bidding at the campaign level and targeting CPA vs. ROI (especially for retailers) leaves profit on the table. I plan to tackle those topics in my next column.

My problem with GCO is that Google is creating a separate and unequal world of data. They’re using information like the actual search query, browser, etc as seen BEFORE the click to modify bids with GCO, but not sharing access to that information to allow advertisers to make their own bid decisions. (Just to make a fine point on this: we can see those data after the click, but there are limits to how we can use it to adjust bidding before the click).

GCO also requires you to tag your site with Google conversion tags. Some businesses are not comfortable sharing that level of information with the same company that sells them advertising. I do believe there is an indirect risk with giving them detailed information.

You called my article a rant, but I didn’t intend it that way. I’m glad Google is innovating, offering new tools and creating new ad formats.

The goal of my article is to show how Google benefits from what they’re doing. I want people to consider what they get out of these tools.

As an advertiser community, I think it’s important for us to demand equal access to the data they use in their tools through the interface and the API. That way, we have choice. We can let them do it, or we can do it ourselves. That’s my goal.

Cheers,
-Alex

Great post… not sure I agree with the “risks” above. If you set it up properly, it’s the least risky way to buy traffic. Only on REAL risk I’ve seen is if you’re CPA is based on opt-in’s and Google finds you opt-ins that don’t convert. Which happens.

Lucas

Many people think Google does some kind of magic, but there is nothing magical at all – run the campaign, get at least few thousand clicks/few thousand conversion, test most possible keywords and ads (well, test everything) and then turn it on if you feel that GOogle has enough data with YOUR keywords. It works on simple stats it gathers over time, but you must do the initial work yourself, drive as many conversions as you can and get as many clicks as you can, otherwise you won’t like what you will get in the beginning.

My own results show that if you have a hungry market not satisfied by the first page of search results, then a low or second page position resulting from deflated bids by GCO is not at all a problem. And we’re doing around 200 lead conversions a day in the hungry and competitive “work from home” marketplace in the UK (700k impr /day or so on Search)

If your prospects either don’t read down through the ads on the search results to find yours, or don’t do so in enough numbers, then a low-bid position putting you on page 2 or lower can damage your impression count, and consequent click throughs, sales and leads.

Having said that, it’s only simply the symptom rather than the underlying problem. If your competitors are on page 1 for higher bids and higher cpa costs, their sales process is working more cost-effectively than yours, their site conversion is better, or they can subsidise it enough to win the business up front and succeed on the back end. (Unlimited Traffic Technique, hey Perry?).

So your site conversion is critical, as is your knowledge of profitable cpa costs. Quality Score and click costs are obviously also a factor, so that’s implicating your ctr and ad creatives.

The last-click attribution is a very valid point also, I first ran into this back in 2006 as I write about here:

http://www.davidnrothwell.com/adwords-search-funnels-why-i-needed-them-in-2006-987/

The new Search Funnels reporting will help determine where keywords with clicks and no attributed conversions continue to have value, and should be kept and nurtured as assistants to the actual “salesman” keywords.

My Google rep told me when I asked a while ago that Search Funnels attribution was not yet built in to Conversion Optimizer, but that it would come in time, which is a consistent, logical and required move by Google, given that the intelligence and machinery exists now and just needs integrating (just as they usually do…).

Interesting to hear both sides about Conversion Optimiser. I’ve found it to be a practical and useful tool at increasing conversions. It’s also likely to push your budget, leading you to increase spend – but as long as you’re getting the conversions – everyone’s happy.

I don’t believe in a fully automated AdWords Optimisation approach, but tools such as Conversion optimiser can be an asset.

Thanks,
Louise