How to balance scale and accuracy in customer profile enrichment

14 December 2020

Will personalization at scale driven by 1st party data ever be more than a pipe-dream?

When it comes to profile enrichment, the reality for marketers has been filled with broken promises:

‘Hundred-percent accurate look-alike models’ is the assurance, but in reality, identities snowball into false positives to create a mess of mistargeting and wasted media spend.

‘Real-time enrichment’ meansslowdowns in sending siloed customer datasets to multiple parties, by which point mistakes have been made, regulatory red-tape has been compromised, and profiles have exceeded their expiry date – quickly slipping into irrelevance.

Get it right, and you get growth. Get it wrong, and you get churn, extra costs, a GDPR nightmare, and an endless struggle towards ROI.

How can brands…

  1. Find more of their best customers, then
  2. Discover new and valuable things about them?

Before we approach these questions, the first step is to be smart about how you recognize customers, prospects, and suspects across channels. We’ve covered this topic in our guide to identifying more customers with your CDP through smart 1st party data onboarding.

Once you have your identity resolution in check, it’s time to start scaling your approach by unlocking undiscovered customer attributes across 3rd party audiences, attaching them to profiles, and segmenting in new, creative ways.

 

Finding more of your best customers

First, let’s tackle the accuracy issue: as a marketer you want to rely on facts, not assumptions when it comes to extending your reach… but you also don’t want to depend entirely on the walled gardens of media giants or multiple external parties, because you’ve recognized the need to keep control over your most valuable asset – customer data.

But it may also be the case that you’re missing data which sets you back in building addressable audiences. Perhaps you need to discover other potentially high-value customers who own both a car and a business. Maybe you need to find customer look-alikes who are both homeowners and live within a certain catchment area.

Extending your reach starts with supplementing probabilistic data with concrete 3rd party insights – deterministic data which has clear, traceable origins back to legitimate sources.

This means building reliable bridges with car dealerships, chambers of commerce and household censuses based on solid audience qualification criteria, leaving zero room for best-guesses: think log-in data and fixed attributes such as demographic and geo-targeting data, as a driver for changeable parameters like browsing history or behavioral characteristics. 

But building multiple data partnerships eats up resources, integrations take time, and your customers and campaigns aren’t waiting around. With this in mind, marketers need to start building smarter bridges with enrichment partners who own a wealth of data in all the right marketplaces – both geographically and vertically.

 

Getting to know your existing customers better

So you’ve found a way to scale your existing 1st party attributes into addressable audiences – based on finding a solution with the breadth of reach in the right markets.

Next up, let’s fill in the blanks to add new, valuable insights to your existing cohort of customers. This newly collected data can help you build your highly coveted single customer view, to understand the behavior of individuals across owned and paid channels in a way you’ve never been able to before. From this, you can refine and adapt your message depending on factors such as loyalty status and previous buying behavior, personalizing back down from one-to-many, to one-to-few or even one-to-one targeting.

Herein lies the balance between scale and accuracy: whilst scaling up your profile data into ready-to-convert audiences requires a big reach in the right places, adding more identifiers to existing profiles means you need to stay closer to your 1st party targeting.

That’s why at digitalAudience we decided to build a solution from customer data onboarding outwards to the right profile enrichment partnerships; because we see that building on what you already know about your customers works better than best guesses and assumptions.

 

Best of both worlds?

That’s not to say that there’s never any trade-off required between the scalability of audiences and the precision of 1:1 targeting. 3rd party profiles are by nature evolved from demographic characteristics rather than individual identifiers, so to start with, expect to cast a wider net over watertight quality.

On the flip side, discovering where your customers have been, and who they are through untapped online identifiers requires that brands understand exactly how they want to enrich customer profiles, and which kinds of insights they need – not to mention what they want to do with the data once they get there.

Many digital marketers are entrepreneurial by nature, and against a backdrop of exponential change in customer behavior, this means that intrepid exploration and agile experimentation are a must.

But personalization at scale takes patience and the right kinds of partnerships – in an approach we call disciplined audience extension.

Disciplined audience extension – the fine balancing act between precision and scale, is a distinct possibility when brands consider the following:

  • There ain’t no party like 1st party: targeting improves the most when you start with a strong solution for customer identity resolution
  • Since this is the core of your customer profile, make sure that 1st party onboarding and profile enrichment are sourced natively or by the same partner, so you can understand the origins of the data, as well as retaining control and autonomy over your most valuable asset
  • That said, look for a solution which specializes in the market segments you seek to reach, so you can guarantee the quality of your audience data based on your marketing strategy and broader business goals.

Keep these tips front-of-mind in your pursuit of strong profile enrichment, and personalization at scale will begin to take shape. It’s well within your reach. [1] 


Maybe add to contact dA or some other next step: ask the reader to inventory their 1PD, audit how they use 3PD today, or some other task to show the difference…