Why the Migration to “Optimize for Conversions” is a Good Thing

A few months back Google announced to change the ad rotation setting for certain campaigns to optimize for conversions. This move has sparked some controversy, but I believe it’s actually quite beautiful.


It’s important to note that only campaigns which meet certain criteria will have their ad rotation changed. These criteria are:

  1. The campaign must be using Conversion Optimizer and/or Enhanced CPC.
  2. The campaign must be using the ad rotation setting optimize for clicks.

There is a conflict between those two: By using Conversion Optimizer or Enhanced CPC an advertiser indicates that conversions are more important than traffic. But the ad rotation setting optimize for clicks indicates the opposite.

However, the argument could be made that in most cases the advertiser simply forgot to change the default setting. Activating Conversion Optimizer or Enhanced CPC on the other require a conscious choice. But that’s not why this is a beautiful move.

The beauty of this move is that everybody wins. The reason is in the way Conversion Optimizer and Enhanced CPC work. Since Conversion Optimizer is more predictable, I’ll use this one to illustrate my point.

Without Conversion Optimizer and Enhanced CPC

If neither Conversion Optimizer nor Enhanced CPC are active, max CPC bids are fairly static. But it’s not really a secret that the CPC bid is not all that matters in terms of revenue for a PPC provider. The other factor is click-through rate. Both of them put together make up the actual bid: the bid per impression:

Bid per impression = Max CPC bid * CTR

This is why Google has a vested interest in optimizing for clicks. If they increase CTR, they increase an advertiser’s bid per impression. Essentially they make more clicks and therefore more money. So far, so good.

Enter Conversion Optimizer

Conversion Optimizer predicts a conversion rate whenever an ad enters the auction (more accurate: it calculates a conversion probability). Then it submits a bid according to a formula that probably everyone who’s ever done manual bid management knows:

Max CPC bid = CPA goal * conversion rate

If we combine this equation with the one from above we get this one:

Bid per impression = CPA goal * conversion rate * CTR

The CPA goal is a constant set by the advertiser, while conversion rate and CTR fluctuate and influence the resulting bid per impression. Now let’s look at the ad rotation settings.

The Ad Rotation’s Side of the Story

The ad rotation setting optimize for conversions is often mistaken for maximizing conversion rate. That’s actually not it’s main objective. As the AdWords Help pages put it:

Ads expected to provide more conversions are delivered more often into the ad auction

So the main objective is to maximize the number of conversions. Raising conversion rate means making more clicks into conversions, which will help to achieve this goal. But getting more clicks in the first place helps just as much.

Clicks come from impressions, conversions come from clicks – it’s obvious that Google has to look at both CTR and conversion rates simultaneously in order to maximize the number of conversions. The product of CTR * conversion rate is what Google actually has to maximize.

Putting it all together

With Conversion Optimizer active, an advertiser’s bid per impression comes down to

Bid per impression = CPA goal * conversion rate * CTR

If the advertiser uses Conversion optimizer in combination with optimize for clicks, Google is only able to optimize for the last factor of the equation: CTR. But if the advertiser uses optimize for conversions, Google controls both CTR and conversion rate, effectively optimizing the advertiser’s bid per impression.

Now the real beauty is that this is good for both Google and the advertiser. Google of course profits from higher bids per impression. But the advertiser profits, too. If optimize for conversions leads to another ad showing than previously with optimize for clicks, then only because the higher conversion rate outweighs the lower CTR.

It’s important to realize that the advertiser won’t lose traffice because of this. Yes, a lower CTR means a lower quality score and leads to a lower ad rank. But the higher expected conversion rate automatically leads to Conversion Optimizer submitting a higher bid into the auction. Since the increased converion rate outweighs the lower CTR the higher bid will outweigh the lower quality score as well. With a constant CPA this means more conversions for the advertiser.

Am I the only one who thinks this is a beautiful move?

About the Author

Martin Roettgerding is the head of SEM at SEO/SEM agency Bloofusion Germany. On Twitter he goes by the name @bloomarty, and you can find him regularly on #ppcchat.

  • http://www.sem-deutschland.de/adwords-agentur/ olaf

    Hi Martin,

    in theory it sounds logic. What about in real life. Did you tested it? I had bad experiences by using the Bid Management of Google in increasing CPOs.

  • http://www.videoschnittlauch.de David

    Hey there Martin,

    we use the Conversion Optimizer as well and achieved good results yet. Interestingly Google itself advised us to use it (it makes sense now because they make profit too…). I think the only tricky part is that you need to give Google some access to you conversion data…

    All the best,


  • Martin Roettgerding

    Hey Olaf,
    I haven’t done any testing with ‘optimize for clicks’ and Conversion Optimizer vs ‘optimize for conversions’ and Conversion Optimizer. I don’t think there is much that could be tested because the impact of the change is limited, especially if a campaign is already properly maintained.

    Basically the biggest impact of change the ad rotation setting is that ad B shows instead of ad A. If A is the former champion it means that A used to have a higher CTR than B. In order for B to replace A as a champion,
    - B needs to be active. Chances are you’ve already paused the losing ad.
    - B needs to have a higher CTR * CR than A. Since A already has the higher CTR there is a good chance that A wins in CTR * CR as well.
    - There has to be enough conversion data for Google to assess conversion rates. Otherwise ‘optimize for conversions’ is basically ‘optimize for clicks’.

    So the cases in which ‘optimize for conversions’ leads to a new champion ad are limited. Additionally, such changes are rarely measureable for us advertisers. We only get to see the end result.

    From a theoretical perspective I am convinced that this change is a good one, but I see no way to test this empirically – at least not without any drastic measures.

  • Martin Roettgerding


    Yes, in most cases we’re pleased with the results as well. I don’t think that Conversion Optimizer is primarily aimed at increasing Google’s revenue directly. I think it aims at increasing the value of AdWords for advertisers.

    By the way, I actually don’t have a problem with making conversion data available to Google. In fact, I always recommend it to clients, even when there is already another tracking system in place. Having this data available in AdWords not only enables the use of Conversion Optimizer and gives you access to Google’s search funnel data. It also gives you the flexibility to use other bid management systems as they (can) all use this data.

  • http://www.sticht.se Alexander Sticht

    Hi Martin experience has shown that “optimize for clicks” actually resulted in more conversions than “optimize for conversions” for us, however it has to be mentioned that in the majority of the cases, there was already a high conversion rate in the campaigns, together with a high CTR