Data Science,

Agency 2.0: How Data Science Is Transforming Digital Marketing Agencies

Zlati

Reading time:

Agency 2.0 is a model of a future digital marketing agency that we all need to start thinking about.

Data has become the new marketing gold. As you embark on a digital transformation journey, you need to adapt your business to leverage first-party data and start bidding towards conversion value.

Let’s walk through some basics.

The Classic Marketing Agency Model

For several years now, the core pillar of the classic digital marketing agency has been the optimization of paid and organic channels.

Simply put, agencies have been zeroing in on optimizing for SEO, creating unique and relevant content, and offering expertise for ad platforms such as Google and social media.

That said, all industries have been experiencing the effects of digital transformation, including marketing agencies. With the accelerated pace of changes today, skilled and adaptable businesses are beginning to realize that it’s time to change.

To surge ahead of the competition, marketing agencies need to transform and hone in on applying innovative digital technologies to traditional concepts.

In broadest terms, this means that – along with optimization – the new Agency 2.0 must be able to leverage data science and continuously work on an extensive creative assets portfolio.

Why You Need to Reconsider Using Target CPA

Most of the digital advertising today, especially within the SaaS industry, is based on cost per acquisition (CPA).

This is telling Google to bid the same – and to value every lead the same way – regardless of whether that’s someone purchasing a basic plan and churning after 3 months or someone looking to get an enterprise plan and stick around for years.

Although target CPA can be profitable, segmenting leads so that they have different values is a more effective strategy for businesses with ranging price points and tiered, license- or usage-based pricing models.

Arguably one of the best ways to do that is by switching over to target ROAS (or target return on ad spend).

If you’ve been relying on CPA, figuring out how to measure ROAS can seem daunting. Once get the hang of it, however, you’ll be able to assign unique values to different conversion types, essentially allowing Google to bid towards conversion value and give you a better ROI.

Utilizing First-Party Data to Deliver Value to Clients

Acquiring new customers can be costly, so it’s essential that your business identifies and nurtures the most valuable customers that interact with you.

To do that, you’ll need to take first-party data as well as data from CRM, marketing automation, payment platforms, and the product itself. Based on all these different data points and signals, you can build a machine learning algorithm that can predict the lifetime value and the future revenue stream of a user at the time of acquisition.

What’s even more – with third-party cookie-based data phasing out – learning how to use first-party data is becoming even more pivotal for the success of digital agencies and their valued customers.

All this means that agencies are now expected to not only be platform experts but also help clients utilize their data. For that, they need an entirely new pillar of expertise – data analytics and data science.

A good place to start is by working out how to build a data strategy and follow three main steps:

  • Identify what first-party data is available
  • Unify that data ‘into a warehouse’
  • Create a machine learning algorithm and tweak, update, and test it on an ongoing basis

Let’s take a closer look at how to keep up with the increasing importance of data in the marketing field.

Focus on PLTV (Predicted Lifetime Value)

PLTV, or predicted lifetime value, is an important metric to measure at any digital marketing agency. By measuring PLTV, you can make predictions on how much money customers will spend on your product/service over their entire lifetime based on their past behavior.

You can think of PLTV as anything outside of e-commerce. That can be any type of lead generation, such as a free trial or freemium acquisition, where you have a future stream of revenue that is relatively unknown at the time of acquisition.

Knowing your typical customer behavior and identifying early milestones that separate users with high potential and users with low potential can be immensely useful during an acquisition.

With all that in mind, PLTV is the key to getting to the next level within ad platforms.

Pair it with first-party data and you’ll be able to predict the lifetime value at the time of acquisition and eliminate the dramatic difference in LTV between customers who churn out quickly and those who are in for the long haul.

Ultimately, digital marketing agencies that can do PLTV well will likely be in the top 0.1% in the agency world.

Take Advantage of Google P-Max

Google Performance Max, or P-Max, is an automated campaign type that leverages AI to show consumers relevant ads with an optimal bid to maximize the campaign performance.

P-Max campaigns are driven by your goals and the automation utilizes available data to create ads that make the most sense for those goals.

Unlike other campaign types – that typically rely on keywords – P-Max automates ad creation based on the creative assets that you have provided.

Performance Max campaigns are likely to outperform many other campaign types, including classic paid search campaigns and YouTube video campaigns.

After all, P-Max is specifically designed to unify all of Google Ads properties (e.g. Search, Display, YouTube) while leveraging the power of AI and your creative assets. All this enables Google to decide how to best allocate and optimize your budget across all those properties.

Why Agency 2.0 Needs Performance Creative Assets

To keep up the pace, Agency 2.0 will need creative, as well.

Many agencies are already well aware that creativity in marketing is one of the most powerful tools that you can take advantage of for your digital strategies.

Just consider this: For an effective P-Max campaign, for instance, you’ll need to feed in enough creative assets. Because the algorithm will be testing rapidly, you have to provide tons of images and videos, but also plenty of ad copy ( titles, headlines, site links, descriptions).

The more creative assets you include, the higher the chance that the algorithm will quickly figure out all the right combinations and all the right allocation of budgets across Google’s properties and deliver the best overall result.

Keeping up With the Digital Transformation

The future Agency 2.0 needs to continue doing great at optimizing platforms, but it also needs to add performance creative, and data science.

Together, these three functions need to be tightly integrated and digital performance agencies need to find effective ways to share information in a rapid, iterative way.

Ready to embrace the change? Here’s how to get started with data science for marketing.

Ready to scale your marketing-sourced revenue?