April 11, 2019, our PPC-specialist Maria Dolgova held a webinar “Working with automated strategies in Google Ads” on the platform of the Academy of Internet Marketing WebPromoExperts.
We decided to deal with this issue in detail and transcribed video reports.
Read the transcription below. Or watch the record.
*video taken from official WebPromoExperts YouTube channel.
So, you have a business and a platform for sale – a website. The context advertising is running – advertising works, brings orders and sales, but you want to improve the effect. And you don’t know how to achieve this. For this, automated strategies are needed.
*According to Google, experts lose about 80% of their time on routine tasks.
Google’s automated strategies are designed to automate the work of PPC experts and move away from the routine.
When you have more than 1 account maintained, and there is critically insufficient time to set bids manually, process automation will help. This is particularly true for accounts with high activity. The flip side of automation – Google manages absolutely everything in your place.
Search algorithms are developed and improved. Information about users, in the form of millions of signals, is continuously analyzed, and this increases the effectiveness of advertising campaigns.
Requirements for Using Automated Smart Strategies:
You must understand what a conversion for your business is. The action should be as accurate as possible – you need to know what exactly you want to achieve (call, filling out the feedback form, etc.).
Today, PPC experts are trying to move away from "smart goals". What does it mean?
For the service business, the conversion is calls, sending feedback forms, clicking on a phone number, etc.
For e-commerce, this is a checkout, "Thank you" page, or a transaction with e-commerce configured. So we get the value of each conversion.
There should be minimal statistics collected.
Works simultaneously with manual bidding.
Works in conjunction with manual bidding.
The result is a conversion increase. Reducing each conversion cost with an advertising budget increase.
This strategy is used when we want to get maximum traffic within a given budget.
The result is a site traffic increase. Exceeding the daily budget is possible if the system makes sure that, the number of clicks has increased.
These 2 strategies are machine learning independent.
You need to set the desired number of impressions at position 1, at the top of the search results above the search results or on the first page to use the strategy.
Search campaigns only. Manual bidding possible.
The result is a quick increase in visitors traffic, showing up on the first page of the Search.
Opportunity to win at an ads auction of another domain. Available for search campaigns only.
The result – being at the top of the search results, increasing brand awareness.
Why are they called so?
Google Machine Learning is the basis of smart strategies. We give input data to the system and strategies work according to the given parameters.
Basis – an amount that we are willing to pay for a call to the office, one sale, etc.
It is necessary to set a specific CPA, which will be most effective for business.
The result – Google will increase the CPC and position of the ad with a high probability of conversion.
If the clicks are unpromising, they will receive a lower price, but the average will not change. As a result, the number of clicks will increase.
The strategy is applied when you know the necessary or effective RoAB.
The result is the minimum possible statistics will be collected in 2-3 months. Then you can increase investment in the advertising budget while maintaining profitability.
This strategy is useful when you know exactly what you need right now (order, calls, sales).
The result is conversions increase with a decrease in their cost and a constant advertising budget. Setting up goal tracking can help determine if a strategy is appropriate.
Google bots process 70 million signals in 100 ms. Thus, the system quickly analyzes all potential customers and sets their profiling.
The signals we can recognize:
The algorithm will quickly determine your bid after processing these parameters.
In our practice, we most often use 4 strategies:
Maximize clicks:
Why?
Imagine you set up and run manual bidding ads. In the future, you want to receive more applications, orders, and sales. It’s already worth learning automatic strategies.
Remember the necessary requirements for smart automation:
You should use the "Maximize clicks" strategy for quickly collect statistics: the more traffic you get in the first period, the faster the information will be collected about users who did the conversion.
Maximize conversions:
Why?
Why use advertising? To get KPI (maximum calls, applications, sales, etc.).
Understand if there is an ad result and how effective the ads are.
Target CPA:
Why?
Now you can determine how much one conversion should cost so that the advertisement is fully rewarded and doesn’t be in the red.
The peculiarity of the strategy is the need to really evaluate your conversion cost.
How to calculate this figure? You’ll see how much one conversion costs after collecting the initial statistics. Then you drop the price a little and indicate in the strategy.
Target ROAS.
Why?
You need close contact with the customer to apply this strategy. The client must have information about advertising costs, cost of goods, etc. Then we can evaluate the real ROI and indicate it in setting up the strategy.
The strategy is unlikely to be effective if you intentionally increased the ROI.
There are 2 ways to choose an automated strategy:
1. Having a clear and explicit purpose for the advertising campaign.
You either take as a basis a position in the search or maximize conversions.
Choose one main parameter, because it will be easier to choose an automated strategy.
2. Follow Google guidelines.
The system, based on the data from Analytics, will itself tell you which strategy is available to you and which will bring the greatest effect. The bot also predicts the number of conversions and the approximate cost of the number of clicks that you can achieve during the next period if you apply the chosen strategy.
Next, compare the proposed recommendation for your own purposes and, if coincidental, apply it.
The most common strategy is "Maximize conversions".
It is used when there is no conversion data for profitability.
At the initial stage, you only need conversion tracking settings.
An additional requirement is the collected minimum statistics. The strategy will start working faster (at least 15 conversions in the last 30 days, but the more there are, the better).
Target CPA.
Google recommends taking an average CPA (average cost over 3 months) and apply it in a strategy.
Target ROAS.
Suitable for advertising campaigns with transaction data with different values (for example, for online stores with a configured e-commerce block).
Try running the Experiment if you haven’t worked with automated strategies yet or worked unsuccessfully.
It will help not to spend all the budget and traffic on an automated strategy.
Process:
What do you need to remember after creating and running the Experiment?
Experiment Example:
The screenshot shows the result of the "Maximize Conversions" strategy.
The strategy worked for 2 weeks and showed itself well, but it did not collect enough information for 100% veracity of the Experiment.
The discrepancy between the values is ok.
Finer points in working with automated strategies:
1. Bid adjustments are not available for all automated strategies, but only for the "Maximize clicks" strategy and manual bidding.
2. Any adjustment in plus will not concern the cost of a click or bid on a key, namely CPA.
For example, if you increase your bid by 50 for mobile devices, the system will set NOT a bid, but a CPA 50% higher.
3. In the strategy "Target ROAS", only 1 type of rate adjustment is available – a reduction of 100%.
4. The use of scripts that adjust bids in order to show up on a certain position will lead to incorrect work of the smart strategy.
Also, do not forget to turn off the scripts, otherwise the automation will place bids in its own correspondence, and the script will try to fix it. And this is an additional download for the bot. As a result, the work will fail.
Pitfalls in working with automated strategies:
1. Short term strategy.
Customers are always waiting for an instant result from contextual advertising; they are not ready to wait 2 weeks to see the beginning of effectiveness. And the system may not have time to learn during this time.
2. Unrealistic input (target CPA or ROI).
You should build on your latest real data. Start with what is now. And then increase bids and watch how advertising will show itself.
3. Indicators of low significance for business are used for conversions counting.
For example, "Smart goals", "View pages about the company", "Scrolling pages by 80%", etc.
4. Adjustment of advertising campaigns with an automated strategy.
Any data change in automatic strategies leads to rolling back before the algorithm starts learning.
5. Evaluation of metrics is in no way related to the main goal of the advertising campaign.
If the goal is to maximize conversions, you should not pay attention to the average position of ads in the search results.
6. Conversion window counting (Google recommends using the 3A strategy):
1. Automatedstrategy.
2. Attribution.
Try changing the attribution from the last click to another, because users do not do a conversion from the first time. It is necessary to take into account the period of interaction (for each click on the site, context, organic traffic or social networks, assign the weight of each finally completed conversion).
3. Audiences.
This is adding new audiences (remarketing lists, etc.).
7. Small budget
Then contact Google for help. For example, for the effective work of the tCPA strategy at a given value of the target CPA, the advertising budget should be at least 10*Target price.
The strategy will work more productively given all these aspects.
An automated strategy demonstrates itself very well in work, you can rely on it. But at the same time, it may work incorrectly.
1. How often can I change bids after learning?
If you use a smart strategy, bidding is set by the system.
When you have chosen the “Maximize clicks” strategy, the only thing you can change is the only bid limit.
If the "Maximize conversions" strategy is set, and traffic is low, then you can raise your bid to get more traffic. The learning period for this strategy is short. If your entire budget is not spent, then you can change it if necessary.
2. Can I run an automatic strategy immediately and do not wait until statistics are collected?
It is possible, but we don’t know how it will behave: what should the strategy rely on in the absence of rough data? It is likely that the desired results won’t be immediately achieved.
3. There is information that in addition to the minimum number of conversions for a campaign, there is a requirement for a minimum number per group for this campaign. Is it true?
Yes, there are strategies for groups. If you apply it to groups, then this requirement can be true.
4. What do you recommend: “CPC bidding” or “Maximize clicks”? What is more effective in your experience?
Manual or Maximize clicks? "Maximize clicks."
From experience: each advertising campaign of our customers ultimately automated runs.
At the start, it is better to select "Maximum clicks" (automated) to collect data and then switch to automated smart.
It should be borne in mind that if there are a lot of keys in the account, and some of it is near-targeted (and this needs to be monitored), then Google can get traffic for these keys. The Maximize clicks strategy does not focus on conversions, calls or purchases, its purpose is to bring maximum traffic. Moreover, it may not be of high quality, and this must be taken into account.
In case all requests are transactional and with reference to the city or brand, then you select "Maximize clicks". So it will be more effective.
5. What budget do you need to start? The goal is to create an online store of trending goods from China and pay off the investments in 3 months. What strategy do you recommend?
First of all, this is a question for businessmen: you need to calculate how much you will spend on all goods sold (prime cost and transportation to Ukraine): we allocate X budget for advertising. If you plan to configure an e-commerce tracking block for a site, then first it’s better to catch up traffic to the site, and only then try to run the ROAS strategy. At the same time, you need to take into account profitability and see how realistic it is.
Briefly, you have to catch up traffic, then run the Experiment, and when you get 50 conversions in the last 3 months, run ROAS.
But if e-commerce is not configured, and every purchase is taken into account, then you can try to set up a target CPA, taking into account the fact that you need to pay off the investments.
6. What to do if target CPA of the campaign dies, transfer it to CPC, make a clone and run?
This happens when an audience burns out. Here you can try to change the target CPA. Switching to manual bidding is the last option.
Perhaps right now there is some kind of deadlock in the strategy, but it will begin to work in a few days. You will lose time if you return to the manual strategy. Because it will begin to learn again. You can also try to slightly increase the CPA.
Google recommends updating target cost for CPA every 3 months. You can’t just set up a strategy and forget about it. It is necessary to monitor analytics, to see how the data has changed for the last 3 months.
7. How many conversions should I get for an automated strategy running?
From 15 for the strategy "Maximize conversions" and "Target CPA."
50 – for "Target ROAS".
You can enable it at the start, but the algorithm will argue for a very long time and not understand how to work. And this will affect efficiency. Therefore, it is better to wait until the minimum statistics are collected.
P.S.: when one automatic strategy burns out, try another.
