There’s been lots of chatter lately about manual vs. automated bidding in Google Ads (hey, AdWords, we still remember you!). On the one hand, manual bidding gives you much more control over spending. On a flip side, it also
wastes requires lots of your valuable time that you have to spend on bid management (‘Below first-page bid’, huh). The Enhanced CPC feature helps with that but doesn’t entirely solve the problem.
That’s when I remembered that I had conducted interesting tests earlier this year. They were specifically aimed at comparing the performance of campaigns with manual bidding against the performance of the same campaigns but with automated bidding.
The Experiment Details
This was a large-scale test as it affected around 50 campaigns.
The following things have also influenced the experiment results:
- Some of the original campaigns had Enhanced CPC enabled
- For this account, each conversion is counted (it makes sense here as a user can search for 1 product but buy 2 different items instead).
A few words about the bidding types.
A bidding method that lets you set your own maximum cost-per-click (CPC) for your ads. This differs from automated bid strategies, which set bid amounts for you.
Enhanced CPC (ECPC)
Enhanced cost per click (ECPC) helps you get more conversions from manual bidding. ECPC works by automatically adjusting your manual bids for clicks that seem more or less likely to lead to a sale or conversion on your website. ECPC is constrained by your max CPC bids when optimizing for conversions.
An automated bid strategy that automatically sets your bids to help get as many clicks as possible within your budget
The main difference is that ECPC lets you set max CPC while with Maximize clicks bidding you set a budget and the system decides which CPC to use.
This optimization is based on your conversion history, so Google used to suggest to have at least 30 conversions for last 30 days to get maximum out of this automated strategy. Many Google reps though claim that even ~10 conversions are enough for the system to optimize your bids.
The aim of this post is to show how particular campaigns performed during particular time with different bidding strategies. Though this research gives some food for thought, it doesn’t conclude that one bidding strategy is better for every advertiser, it doesn’t encourage a one-size-fits-all approach.
Manual CPC vs Maximize Clicks: Overall Results
No more words, here are the numbers:
I will share the results by campaign below. But even with those numbers will tell you a lot.
At first sight, everything seems to be fine with the automated bidding: the impressions are more than 40% higher, the clicks have also increased, search impression share is up by 23%. Even the number of conversions has increased from 113 to 119.
But look at the last column – ROI. Wow! This difference is huge, the ROI is down by 58% and even a 5% increase in conversions won’t make u for that.
So technically, the objective of the experiment is achieved: the experiment campaigns had more impressions and clicks. But at the end of the day, that’s ROI what matters, and from this perspective, the experiment is far from the success.
Manual CPC vs Maximize Clicks: Results by Campaign
Ok, 100 campaigns are hard to digest, so here is a screenshot of 15 first campaigns. You can find a full list in Google Sheets:
I highlighted the most interesting results from my point of view. And here are some observations:
- CPC even for the branded campaign is much higher compared to the original campaign ($0.77 vs $0.10) while CTR is much lower (8.82% vs 17.33%). We got a conversion from it but at cont/conv. of $16.17 it’s not that valuable (for a branded campaign).
- Campaign 3: 8 conversions at cost/conv. of $10 for the original campaign vs 2 conversions at cost/conv. of $72.36 (!!!). This doesn’t even need any explanation. As well as 10 conversions at cost/conv. of $5.92 vs 5 conversions at cost/conv. of $36.80 (Campaign 5). You can also have a look at Campaigns 16, 22, 25, 36 and some others that follow this pattern.
- Interestingly, some of the campaigns were extremely successful with Maximize clicks bidding. Take Campaign 34 as an example: 1 conversion for $126.92 in the original campaign vs 4 conversions for $62.08 each in the experiment campaign.
- Overall CTR is higher (sometimes much higher) in the Manual CPC (original) campaigns which totally makes sense. Though some campaigns (like Dynamic Search Ads) don’t follow this pattern.
The results of this experiment are quite similar to those I had from another experiment: in most cases, Manual CPC worked better. I also tested Target CPA there:
Should You Use Automated Bidding?
Absolutely. This may sound counter-intuitive but automation saves tones of time. You just need to use it with caution. Google is really good at pushing your spend but if you learn how to monitor it and review your account manually from time to time, this will be a good balance between automation and strict control of CPC.
When to Use Maximize Clicks Bidding?
This can also be applied to other automated bidding strategies as they all rely on your account history.
- If you have enough conversion data. I would still stick to the rule of 30 conversions for last 30 days (per campaign, not the whole account!). Otherwise, the optimization will be nothing more than shots in the dark.
- If you want to increase clicks (who would have thought :), i.e. drive more people to your website.
- If you have a large account and want help from Google to control it
- If you want to test different bid strategies (just do it!).
A Few Tips about Conversion History
As stated above, some conversion history is required on a campaign level. So if you have multiple campaigns with a few conversions each, you can merge those campaigns into one and set up an experiment a month later.
But make sure that such a consolidation makes sense. I wouldn’t recommend combining unrelated campaigns or campaigns targeting different locations.
Another way to get more conversion history is to optimize for soft conversions, i.e.mid-of-the-funnel actions that are likely to lead to conversions. For example, instead of transactions, you can track adding products to cart.
To use or not to use: that is the question. Everyone should answer it for themselves since there is no perfect bidding type for every account. Think, test, analyze, implement, repeat. And remember: no matter how smart the system is, it’s the human who decides how to use it.