Five ways machine learning improves PPC advertising


30-second summary:

  • The machine learning-enhanced “Smart Bidding” capabilities available in Google Ads offer an easy way to see the campaign results you want while allowing you to take a more hands-off approach.
  • You can use Smart Bidding on the campaign level or, if you prefer, you can apply it to your entire ad portfolio, allowing you to optimize all of your campaigns using Google’s ML algorithms.
  • Marketing consultant Gabrielle Sadeh explains five different objectives that Smart Bidding can optimize for – and when to use each.

PPC advertising is a core part of many brands’ online marketing efforts. Unfortunately, getting the most out of your PPC bidding can be a tricky process, to say the least.

Some brands find themselves bidding too much for keywords, ranking highly for a brief period but blowing through their marketing budget. Others don’t bid enough, and never see themselves achieving the conversion rates they hope for.

The good news is that there is a smarter way to bid. Google’s Smart Bidding options, which are readily available via Google Ads, offer an easy way to see the campaign results you want while allowing you to take a more hands-off approach.

These advanced algorithms draw from several contextual signals to drive performance, and some of these signals are only available when using Smart Bidding. These include factors like local intent, remarketing lists, ad characteristics and seasonality. Bidding strategies are also available at a portfolio level, allowing you to apply machine learning to optimize all of your campaigns.

By letting Google’s machine learning algorithms take on a load of PPC bidding, you can focus more of your efforts on other marketing and conversion tactics, like optimizing your ad copy or landing page designs. Google Ads currently offers five Smart Bidding options — here is how each one can improve your PPC advertising.

1. Target Return on Ad Spend (ROAS)

Target ROAS, helps you maximize the value of your campaign based on how much revenue you’d like to get for each dollar you spend on PPC advertising. For example, if your goal is to achieve $7 in sales revenue for every $2 that you spend on clicks, that means you have a target ROAS of 350%.

After inputting your target ROAS, your Google Ads campaign will set bidding for cost per click campaigns in an effort to meet this goal. 

Target ROAS is strongly reliant on your historical data. Google requires that your campaign have at least 15 conversions over 30 days for when using Target ROAS to optimize search ads, and there are other minimums for other ad types. 

For best results, however, you should achieve 50 conversions over a 30-day period prior to implementing Target ROAS bidding. This way the algorithm has data to learn from.

According to Google’s own internal research, this bidding method can increase conversion value by an average of 35%.

2. Target cost per acquisition (CPA)

Target CPA allows advertisers to set a target goal for cost per acquisition (CPA) that they wish to achieve with PPC advertising campaigns. When you use this parameter as your optimization target, Google’s Smart Bidding algorithm aims to achieve conversions at an average cost that is equal to or less than your target CPA. 

This way, you can grow your sales without running the risk of overspending.

If you already have a Google Ads account, Google will draw on your historical data to recommend a Target CPA when you are setting up bidding. It’s often a good idea to follow this guideline, as setting a goal that is significantly lower than your current CPA could cause your conversion rate to crash.

PPC advertising - Smart bidding

Source: KlientBoost

In one recent case study published by KlientBoost, switching a client from manual bidding to Target CPA-based Smart Bidding resulted in a 107% increase in conversion rate, while also decreasing the cost per conversion by over 40%. By drawing from your historical data and contextual signals, Target CPA is an easy way to get more conversions for less.

3. Target impression share

Not every PPC campaign is going to be focused on direct sales and conversions. Many brands try to increase the number of search ad impressions they receive as a method of generating demand and awareness for a brand and its services. 

Quite often, these ads can prove effective in helping you gain new leads. Target impression share bidding will optimize your bids in order to maximize your ad impressions on the SERPs.

As part of this Smart Bidding strategy, you can select to have your ads shown at the “absolute top” of the search results page, near the top or anywhere on the page. Users can adjust these targets as they see fit — so your ads could appear at the top in some search results, and in a sidebar in others.

Device optimization can be especially important in this strategy. In a case study from Metric Theory, Target Impression Share bidding, the total cost per click increased with both mobile and desktop devices. However, the campaign increased mobile spend by an impressive 87%, as the bidding diverted to lower-cost mobile users.

PPC advertising - Impression share targeting

Source: Metric Theory

Because target impression share bidding is based on real-time data, it overrides the bid adjustments from your manual campaign. The one exception is the ability to turn off mobile bidding.

4. Maximize clicks

Similar to target impression share, choosing to maximize clicks as your Smart Bidding parameter is not necessarily focused on increasing your conversion rate. However, this can still be a viable bidding strategy for your PPC campaigns.

With Maximize Clicks, Google adjusts bidding to get as many clicks within your average daily budget parameters. Drastically increasing traffic to your website can serve as a valuable method of collecting data for future campaigns in preparation for transitioning to Target ROAS or Target CPA bidding. 

Getting more users to your website can strengthen your branding and help you build lists — an invaluable strategy for B2B companies that rely on more personalized sales calls. 

Indeed, a case study from PPC Hero found that while Maximize Clicks was universally successful in increasing the total number of clicks, and generally successful at reducing campaign costs per click, these campaigns had a tendency to deliver fewer conversions. 

Like Target Impression Share, Maximize Clicks campaigns should focus more on awareness and leads than actual sales.

This campaign option does allow for a few manual adjustments, such as the ability to schedule your ads to be shown (or not shown) on specific days or times.

5. Maximize Conversions

For conversion-focused digital marketers, Maximize Conversions is a highly appealing Smart Bidding option. As with Target CPA and Target ROAS, this bidding strategy draws from a mix of historical data and contextual cues to optimize ad placements.

The big difference with Maximize Conversions is that it is less focused on your acquisition costs and more on volume. This can be especially helpful when you’re working with a small budget. 

Increasing your sales volume in the early stages of your company will spur word of mouth growth in the future as more customers leave reviews for your products and share their impressions with others via word of mouth. This will make future campaigns more cost-effective thanks to the social proof you will have created for your brand.

Using machine learning for PPC advertising

Source: Google

However, this strategy doesn’t account for ROI — it just seeks to spend the total available budget to achieve the maximum number of conversions possible. This could increase your cost per acquisition or daily spend, which makes this option less than ideal if you are closely tracking those metrics.

A smarter way to bid

Trying to determine which bid will maximize your campaign goals without blowing through your budget can be a tough balancing act. By turning the heavy lifting over to machine learning, you can ensure stronger outcomes for your campaigns while also giving yourself more time to further optimize other key advertising elements.


Source link