Best practices to track offline conversions in Google Ads

Best practices to track offline conversions in Google Ads

Marketers must be willing to take risks. There is no guarantee that any strategy will work.

I consider myself a calculated risk-taker. TSA PreCheck allows me to check my flight status remotely from home so I arrive at the airport an hour before my flight. I visit new restaurants, but I wait until I have read reviews and looked at the menu before I get in my car.

Similar is my approach to taking on risks at work. While I enjoy trying out the latest bidding strategies, my actions are designed to not disrupt performance.

Google Ads accounts that have experienced major changes in their settings can send the algorithm into a learning phase, which causes volatility and allows it to learn new things.

My patience is sometimes short, just like many marketers who are driven to get results today. To keep my clients happy, it is crucial to balance future-forward strategies with the possibility of upsetting KPIs.

We all know offline converter tracking (OCT), is an essential part of any account.

It can be difficult to swap out web-based campaign optimization objectives for CRM-based goals when setting it up for first time. These are five tips to reduce the risk.

1. Convert values with additional values

To maximize the potential of OCT, each action must be assigned a value.

Google offers a calculator to help you estimate if you are unable to use dynamic and actual values from your CRM.

For the algorithm to work, the values do not have to be exact.

Machine learning learns how much each conversion is worth by assigning values. The system can then bid according to value-based bidding strategies.

These values can be set from the beginning so that you can launch your campaign when you are ready.

2. Optimize for all stages of the funnel

Once you have all stages of your marketing funnel mapped into Google Ads, add their respective values to optimize campaigns.

This tactic will increase conversion data for your campaigns, and accelerate the learning phase.

An account risk-taker could change optimization across all accounts at once. You can take the bandage off.

• You can tolerate performance fluctuations.
• If your account is new.
• Or if conversion volume is already dismal.

To reduce volatility, I prefer to set up an experiment to test any major bidding changes.

The experiment should last at least two weeks before you determine a winner. However, this can be longer if you have a small budget or if the conversion lags are long.

As the main success metric, this experiment should see a rise in backend conversions.

3. Optimize to get the lowest possible conversion point

After the full-funnel approach has been successful, you can start optimizing for each conversion event. Each campaign should be treated as an individual and you can adjust campaigns to suit your needs.

An experiment should be set up to optimize to the lowest part of the funnel to reduce performance fluctuations. Google requires at least 15 conversions within the last 30 days to ensure this.

You can easily sort and visualize your data by creating custom columns in the Google Ads interface. Each campaign can be analyzed independently.

The top two campaigns in this example have sufficient MQL volume to optimize only for MQLs, rather than the whole funnel.

Be aware that not all keywords can be considered equal. Your top-of-funnel keywords are more likely to generate pipeline than your brand terms or non-brand terms.

The optimization goals of individual campaigns and bidding strategies must align with the funnel stage and what is working best.

OCT goals cannot be set in a single conversion event. Instead, they must be set for the entire conversion category. Custom goals are required when optimizing for a single point of the funnel.

If your OCT conversion is “sales,” then you will need to create a goal for that.

Google states “[…] using customized goals in campaigns may make your bid strategy less efficient,” however, if you have sufficient data and a well-established account, custom goals can work to your advantage.

4. Continue to drive lower-funnel results

You can push the boundaries and optimize for lower points of the funnel as data accumulates over time.

You can test any bidding strategy with this tactic. However, it is important that you only change one variable in the experiment.

The control campaign, for example, is optimized for MQLs and on target cost per-action (tCPA). The target cost-per-action (tCPA) should be maintained for the test campaign, but it should be optimized for the next stage of the funnel (Meeting Scheduled). Once you have determined the winner, conduct a separate experiment to test another variable.

It can take several months to find the right combination of optimization goals, bidding strategies depending on how large the account is. It is possible to see little or no improvement in performance after the first two weeks. However, it is worth running three cycles of experiments to find a winner.

If it takes two weeks for an MQL “Meeting Scheduled” to become an MQL, you should continue the experiment for at most six weeks. Check out <a href=”https://support.google.com/google-ads/answer/1722023?hl=en#:~:text=The%20%22Path%20metrics%22%20report%20shows,much%20those%20conversions%20are%20worth. To see how long each conversion takes, you can view Google’s path metrics report.

5. Try value-based bidding

Optimizing for lower-funnel conversion points is a good start. Maximize conversion values or tROAS bidding strategies are a better way to go.

Smart bidding strategies automatically adjust bids to predict potential conversions when users search for keywords that you are bidding on.

You can test these bidding strategies in an experiment, particularly if you are afraid of breaking the algorithm.

The benefits of volatility should outweigh any risks.

One of my clients experienced a 74% increase in the cost per MQL after optimizing for OCT backend conversions.

We optimized for MQLs when their brand campaign was generating more than 15 MQLs per month. This resulted in a 4x increase in conversion value.

Take calculated risks and get out there. When you feel apprehensive, be bold, smart and eager to try new things.

Search Engine Land published the post 5 best practices to track offline conversions in Google Ads.