Customer Analytics, Privacy, & Consent: 5 tips to turn bafflegab to genuine CX

0
419 views

Share on LinkedIn

customer analytics
It’s common knowledge that deep customer analytics can unlock key insights into consumer behavior and improve marketing and selling efforts.  Yet worshiping at the altar of know thy customer while turning a blind eye to its ramifications misses the entire spirit of striving for customer-centricity.  To get there, brands need to view customer analytics and data through customers’ eyes.  But what we’ve seen building over the past 20 years is consumers realizing that isn’t happening, and as such taking matters into their own hands.  Increasingly, they’re using short-term defense mechanisms, such as ad and cookie blockers, while waging grassroots campaigns to lobby for systemic data privacy legislation aimed at curbing commercial surveillance.  

And while consumer data collection and customer analytics are not new, the ubiquity of the personalized web and mobile ads over the last two decades, along with major customer data breeches (44 so far in 2020; 122 in 2019; 107 in 2018 [i]) have helped call attention to the extent of the problem, which is:

  1. Consumers DON’T own their data (or insights gleaned from it).  It’s collected, modeled, and monetized by others – in many cases with entire industries built on it
  2. It’s often poorly secured, breached, ending up in the laps of bad actors
  3. Brands thrust sensitive insights onto consumers without forewarning and explicit permission
  4. Companies share and sell data & insights, often without consent

The fundamental quandary is there are two worlds at odds with each other:  one where businesspeople indiscriminately mine and refine customer data with the sole aim of maximizing profits; the other where consumers, as the strip mine and the exploitation target, are revolting.

To deliver better CX, companies must inculcate fundamental customer-centric values that permeate the culture – honoring customers’ rights to choose, to transparency, and to control over their data.  Every employee must be accountable for data, customer analytics, and applications.  The trouble is, in most cases, this isn’t happening and instead, profits are taking precedence over consumers’ rights.  

Unless customer analytics (and the resulting treatments) pass through a customer-centric lens, chances are nearly 100% that customers will feel abused.  As professionals, we must honor customers’ analytics & data rights while striking a careful balance to stay data-driven and focused on value-added services.

Here are 5 tips to help direct efforts toward data responsibility and innovative customer analytics capabilities that will keep customers’ best interests in mind and play the long game with CX.

Tip #1: Collect data transparently – with explicit permission

Presently, there are countless ways to hoover up customer data.  But just because it’s possible doesn’t make it right.  It’s time to review your customer data pipeline, going back to the source, whether mobile apps, anonymous web visits, registered web actions, emails, chatbots, or 3rd party appends, and inventory the means used to collect that data.  During this exercise, review company policy for informing consumers and obtaining their consent, and ask these questions:

  • Is there transparency with the granularity of collected data?
  • Is the plan clear for storing (and securing) the data?
  • Is explicit consent given to share it?
  • Are well-defined uses spelled out?

Granularity – In many cases, consumers have no clue of the breadth and depth of captured data.  For instance, mobile devices shed behavior like gestures and specific touch/tap events such as pinch, swipes, drag/drop, scrolling, and keystroke cadence.  And then there’s location, screen orientation, device details, battery levels, and more.  Companies can use these to predict very sensitive subjects, such as risky behavior and fraud.  It’s important to be upfront on what you plan to capture.

Storing and Securing – When consumers share their data, they have implicit expectations, but too often brands don’t meet them.  In response, regulations (such as GDPR and CCPA) have gained steam and give consumers explicit rights such as the right to be informed, obtain access, and demand erasure.  If you capture and store customer data, under these laws you might be considered a “data controller” and as such, are bound to secure and manage that data, accurately record consent, and honor it.

Sharing – This is the most sensitive and abused area.  Be prepared to address hard questions (with answers that are sensible to customers) like:

  • Why it’s necessary to share data?
  • With whom? 
  • What value will consumers get from this sharing? 

I’m amazed when I read modern-day personal information disclosures.  Many still flat out flaunt that consumers can’t limit sharing for joint marketing with other institutions.  That’s absurd.  Regardless of whether it’s legal or not, let the consumer decide – but at the same time, give them good reasons to consent to share in some situations and spell out how it helps them.

For years, opaque sharing has taken place behind the scenes, and again the situation increasingly has grown worse.  Take for example mobile ad tracking.   Many are unaware that Apple allowed app makers to embed Google and Facebook data sharing SDKs into their App Store apps without gathering any explicit consent for ad tracking.   This too, however, just reached an inflection point.  In June, Apple announced a major change in IDFA in iOS 14, meaning consumers will be prompted to give explicit consent to each app for ad tracking (previously the universal feature to limit this was buried under Settings->Privacy->Advertising)[ii].  And I mean explicit:

Data use – Brands can use data to determine if consumers are meeting goals, interested in something, showing a behavior trend or tendency, struggling with a widget (such as a shopping experience, an application process, or a calculator).  They can track search activities, and of course, from all this, try to predict future actions and personalize marketing.

With great power comes great responsibility.  Businesses collect, store, and distill massive amounts of data.  In some ways, it’s frightening, and certainly spooks consumers once they realize what’s happening.  Be frank with them about what’s collected and why.  Don’t bury it in fine print just to check legal boxes.  Instead, be transparent from the start.  Also, if you want to reduce risk and cost, you’ll need a more centralized strategy around your data and decision-making systems.

Tip #2: Centralize key decision-making data and analytics

A seasoned data scientist knows that finding, preparing, refining, and understanding data is key to building accurate models and surfacing useful insights. That’s why they’re called data scientists.  And because firms increasingly scatter customer data across systems, their job is getting harder.  Very few companies have the luxury (like Netflix or TurboTax online) of sourcing, munging, and storing all the data needed for their customer analytics in one cloud – start to finish.   

Still, fast access to 1st party behavior data is paramount.  With much of it trapped in on-premise, legacy storage (data marts, data lakes, and warehouses), the challenge isn’t just gaining one-time access to it, but creating a working pipe where it can be continuously joined up with other data to build a well-rounded view of the customer.

Despite the long-term promise of cloud-based systems, on-premise data is still prevalent.  And because it’s not practical to ship all data to the cloud due to considerations such as privacy, cost, latency, and technical difficulty, for now, get the aggregate data that matters into the cloud (data that is known to drive key decision-making policies and strategies) – for example the classic recency, frequency, and monetary data in areas such as purchase history and shopping behavior.  And obfuscate it as needed, to protect personally identifiable information.

To do this, get a cloud platform that will act as the organizational brain.  Feed it only the data and customer analytics it will need to make CX decisions.  Besides on-premise systems, data needed may come from other clouds, streaming sources such as IoT, web, and mobile analytics – but when it does do not cut corners on collection and governance standards.  Now more than ever, having this data in one agile environment will prove invaluable for making nimble adjustments to rules, policies, strategies, and actions to drive consistent, timely, and relevant treatments across channels.

Tip #3: Put data governance practices and controls in place

Having data organized and centralized helps, but that alone won’t ensure data safety and conscientious use.  So much can happen with data, from the time it’s first rounded up, to when it’s accessed for analytical purposes, shared, and finally used to drive personalization and decision making.

Data governance practices and controls are necessary to protect (and enhance) these valuable assets (and the company itself).  But what is data governance?  It’s the discipline that enacts universal rules for data use, procedures to meet compliance requirements, and processes to oversee data collection and management.

When putting a data governance framework in place, be sure it covers principles such as:

  • Auditability – ability to trace and report on data activities
  • Ownership – appointing people/roles who are ultimately accountable for data
  • Data quality – enacting standards which ensure processing integrity and efficiency
  • Change management – procedures including testing (such as a bias checking) and approvals

Tip #4: Monitor and measure data value through customers’ eyes

Most agree that data and customer analytics add value, yet measuring their exact worth is difficult.  As a proxy, many rely on convenient measures, like access statistics.  But just because analysts access the data frequently, does that alone supply an accurate enough measure of value?  

Ultimately, firms should measure the value of data and customer insights in terms of customers’ response behavior – an indicator of value through their eyes.  For instance, if data is included in modeling intent to purchase, and the resulting insight is used to present a product recommendation, the value of that insight (and the data that drove it) is best represented by customers’ actual response to the offer, and whether they transacted.

Perhaps companies should measure data and customer analytics value in terms of outcomes, and not simply data access frequency.  Doing this would require working back from the value of conversion/response events and attributing worth to data and analytics used to drive recommendations.   With this attribution approach, the intent model gets credit and, in turn, the variables driving the model get credit.

This method does call for a more sophisticated data accounting system, requiring an extra attribute for each field and model score, but it’s worth the effort.   And the place to start is with data and models that play a role in real-time decision making.   These items occupy high-rent space in real-time interaction management systems and consequently must justify their costs.

Tip #5: Add value by finding valuable insights without being creepy

“Strive not to be a success, but rather to be of value.” – Albert Einstein

“The most intriguing people you will encounter in this life are the people who had insights about you, that you didn’t know about yourself.” – Shannon L. Alder

Valuable insights are not aggregations, compilations, or statistics.   Calculating the average order value for people between the age of 18-25 in July is not a valuable insight.  To be valuable, insights need to come from careful investigation of struggles or problems.  If last year the average order value for young adults was 50% higher than this year, there may be a problem.  Now investigate it further by asking: why is this happening?

Maybe a certain popular product was available last year that is no longer available.  Or perhaps it’s because many of these buyers are now 26, but your promotional targeting criteria are static, and they are no longer seeing certain ads.  Whatever the case, do the leg work to get to the bottom of it.

And when using the insight, remember, there’s a fine line between being valuable and responsible, and being creepy.  To learn more about this topic, check out:

Finally, any proposed insight must pass the “so what” test.   Meaning, customers must see that using it provides a clear advantage over the status quo (e.g., has measurable value), that it’s compatible with current needs, simple to digest and understand, and actionable.

Conclusion

Delivering great customer experience involves a delicate balancing act: collecting data – and doing so with integrity, transparency, responsibility, and explicit permission; being good stewards of the data by protecting it and treating it as if it were your data; finding valuable insights and unveiling them without being presumptuous or crossing customer creepy lines.  

The tug-o-war between consumer privacy and commercial realities will continue.  Be on the right side of history by staying on the side of your customer.


[i] Identityforce.com, https://www.identityforce.com/blog/2020-data-breaches, July 2020

[ii] Forbes, https://www.forbes.com/sites/johnkoetsier/2020/06/24/apple-just-made-idfa-opt-in-sending-an-80-billion-industry-into-upheaval/, June 2020

LEAVE A REPLY

Please enter your comment!
Please enter your name here