5 Ways to Outsmart Smart Bidding

December 5, 2019

As digital marketers, we know that Google has implemented a lot of machine learning features into its bidding platform in recent years. We’re at a point where you can create campaigns in Google Ads and the search giant will pretty much optimize everything for you.

Smart Bidding, in particular, optimizes your bids to maximize conversions, tapping into Google’s vast tank of signals—including user search query, browser, and language settings; location; and a user’s historical behavior to predict the likelihood of a conversion. From there, Smart Bidding increases your bids when a conversion appears more likely.

But which tools can you leverage to gain an even bigger advantage with Smart Bidding?

Get the basics right

When thinking of ways to be more strategic with Smart Bidding, the first thing to consider is your data integration. Marketing needs to be data-driven to be effective. There’s a lot of data you can generate in today’s world, and the data you’re pulling will be the core of a successful marketing strategy.

If you know your target user’s behavior, goals, pain points, and challenges, you can develop marketing campaigns that cater to their specific needs. Some of the data you could be collecting includes:

  • Web analytics
  • Social media data
  • Email marketing
  • Blog and content creation
  • Lead generation
  • In-store sales

Marketing analytics and data are playing a larger and larger role in strategizing your business forward. The more insight you have, the better, more strategic decisions you can make. Which begs the question: are you incorporating all the data you have available?

Once you have the right data integrated, it’s important to assess your KPIs, or choose what to measure. You should always measure quantifiable metrics that align with your organization’s goals.

We often see brands optimizing to a CPA. However, this approach is far from ideal if you have different margins for groups or categories of products, and doesn’t follow the logic that “not all orders are created equal.” It’s always good to occasionally review KPIs—you can often realize huge gains by adjusting to a more granular KPI.

In addition to the above, “not all customers are created equal,” either. At Marin, we believe that customer lifetime value (CLV) remains one of the best metrics to understand the overall impact of your performance advertising campaigns. Your customers aren’t just worth the amount of money they spend on your business today. They have future value if you’re able to retain them as customers.

Smart Bidding can only reach the next level of maturity when it works on conversion data that differentiates between the first transaction of a new customer and repeat purchases. Typically, only a CRM system can tell which of the two happened. And while connecting conversion tracking to CRMs presents a challenge to digital marketers, it’s possible with the right development resources.

Now that we’ve covered the basics, let’s talk about how you can improve performance with signals that matter to your business.

Align to your business needs with custom modifiers

Using Smart Bidding means relinquishing control across different market segments to what Google interprets as the most likely to convert. That’s fine in a vacuum, but an advertiser will commonly know more about their business than what’s reflected in historical online conversion metrics.

Still, there are many unique factors and trends that impact you as an individual advertiser that publishers aren’t aware of, such as seasonality, ratings, new product launches, coverage in the media, or social media buzz. Incorporating extraneous or contextual data into bid optimization can help you gain a competitive advantage and understand consumer intent—meaning more sales and revenue.

Weather and television-based contextual marketing are some of the more prominent examples of contextual marketing, but there’s no shortage of additional opportunities.

To give you an idea of how online KPIs can sometimes not match the real world, imagine a car rental company. If a particular location has very little inventory left, it might be worth considering bidding down, since a high percentage of traffic wouldn’t be able to rent.

Similarly, if a location has excess inventory, consider bidding up!

Modifiers can also be very useful to help handle promotional adjustments. Many brands have a calendar of sales or promotions, and while some of these happen at the same time every year, it’s likely that many are ad hoc, occur for different product lines, and happen for varying lengths of time each year.

Since all bidding engines leverage historical data, there will always be a delay from when a promotion begins to when bids are adjusted. Therefore, it’s important to modify your bid on the day the sale begins to capitalize, and likewise when a sale is over and conversions get back to ‘business as usual’—applying a negative modifier helps prevent overspending on bids while the bidding engine adjusts.

This is far more difficult to achieve with Smart Bidding, but easier by leveraging Google Ads scripts or a tech partner like Marin.

Full-funnel optimization

Once you’ve covered the basics and started to apply adjustments for your specific business data, another challenge that many brands face is optimizing to multiple stages of a conversion funnel.

There are many options for how to look to optimize in the most efficient manner. You could look to optimize to the point highest up the funnel—or the ‘lead.’ Or you can optimize to final sale.

However, if you have a long latency period from initial lead to final purchase, this limits how reactive bidding can be to current conditions, especially if you change targets. While it’s best practice to remove the days of latency from bidding decisions to make sure you’re using accurate data, it’s important to remember that it’s not current data.

So...what is the best bid strategy to optimize to downstream revenue in the bid calculation?

In the above example, we have a number of users who see an ad for a performance marketer. Some of the users will click the ad and decide it’s not a good fit for them, others will register, fewer others will continue to pay for a trial after being contacted, and finally even fewer will decide to buy the full product/service.

At Marin, we recommend trying to find the ‘best of both worlds’ to account for every step of the funnel, which requires making multiple bid runs. First, run bid calculations to the highest point of the funnel, or the ‘lead’, and then additional bid runs to calculate for the other points in the funnel or final revenue.

In this example, one keyword creates lots of leads, and therefore an accurate bid optimizing to leads equates to $5.

The second keyword creates far less leads, so a bid calculation for this keyword equates to $1.50.

However, the first keyword results in far less revenue from all those leads than the other keyword, and therefore a second bid run would need to calculate the adjustment required.

The first keyword is adjusted -50% to create a final bid of $2.50. The second keyword, which creates more revenue overall and far more revenue per lead, is adjusted by +500% to give a final bid of $7.50.

Marketers can leverage Marin’s full-funnel bid strategy to stay reactive and make sure the top of funnel is optimized.

Budget allocation

These days, a data-driven marketer has to answer many budget allocation questions. Where should I invest my next dollar? Am I going to hit my revenue goals? Will I spend my entire budget? Plus, today’s online customer journey isn’t as simple as it used to be.

Consumers switch between channels (and devices) as they move down the purchase funnel, from inspiration and product discovery to more focused product research and comparison, and finally to conversion.

The million dollar question is always how and where to spend your budget. While we’re not going to touch so much on attribution, it’s important to note that leveraging predictive models to correctly allocate budgets across different campaigns and channels can help drive overall uplift.

With tools like Marin’s Budget Optimizer, marketers can forecast, monitor, and automatically reallocate budgets across their most efficient campaigns and marketing channels with ease.

Take this WorldFirst customer example.

Using historical performance, Marin’s algorithm paced the budget across search and social to achieve the best results per the customer’s business objectives. As a result, WorldFirst has seen conversions increase by 107%, and has a much better understanding of which campaigns drove more sales and where budgets were best allocated.

This kind of financial modeling can be applied on top of your bid strategies to have a continuous view of investment opportunities and high-potential returns.

Leverage technology for better automation

Finally, let’s finish with thoughts of technology.

To recap the above: while Smart Bidding does offer a quick answer to some common PPC challenges, Google’s broad data doesn’t necessarily reflect your target audiences, which are unique from the average user.

More importantly, Smart Bidding gives you no visibility or control over the data being used, which means an advertiser has no access to the complete bid history of any keyword or any individual auction. For advertisers looking to analyze granular performance, the possibilities with Google are limited.

To truly maximize your campaign performance, you need to take control of your own bids and the data being used to optimize them, and do so in an automated fashion.

Leveraging a third-party advertising technology tool like Marin can help you coordinate with Smart Bidding. Marin works as a performance layer for those who want a powerful bidding solution to automatically scale their accounts while maintaining a certain degree of control.

eCPC is a smart bidding strategy that allows you to continue to place the base bid outside of Google, while still allowing Google to make the ‘Auction Side Adjustments’ based on the data they hold. This is almost the “best of both worlds” and allows you to capitalize on some of the strategies we’ve discussed, without giving up Smart Bidding as your automated bidding solution.

By using this hybrid approach, Marin continues to help brands outsmart Smart Bidding, or more accurately, work hand in hand with Google to get an edge over their competitors.


We’ve talked about several areas you can focus on to drive uplift:

  • Getting the basics right with accurate data, correct KPIs, and data driven targets
  • Aligning your online KPIs to real business needs by leveraging Custom Modifiers
  • Considering a full-funnel bid strategy if you have a multi-conversion funnel to optimize to what matters most, while still remaining reactive to auction or real-world conditions
  • Knowing the true value of all your digital marketing activity across different campaigns and channels to maximize your return and drive incremental lift, with better budget allocation
  • And finally, leveraging technology that can help you easily implement and automate these strategies

Want to learn more? Join our webinar, How to Outsmart Smart Bidding, on Wednesday, December 11th at 10am PT / 1pm ET.

Farah Shalwani

Marin Software
By submitting this form, I am agreeing to Marin’s privacy policy.

See why brands have relied on Marin to manage over $48 billion in spend