RSS feed Get our RSS feed

News by Topic

BizReport : Social Marketing : November 10, 2020

Expert: How to take back data control

Since The Social Dilemma documentary came out, and since many brands took a step back from social platforms in the wake of social justice issues in the US, many have wondered what is next. How do brands, who have put such effort into social sites, regain control of their data? How do they engage with existing and new customers without the 'help' from social platforms? We asked a digital expert for advice.

by Kristina Knight

Kristina: The Social Dilemma documentary has a lot of people talking about how social media companies are capturing and using consumer data. What few are talking about is how these companies are using brand data - should brands be as concerned as consumers currently are? 

Kasper Skou, CEO & Co-Founder, Semasio: I think they should be! The old internet B2C adage goes: "If the product is free, you are the product." If we are to reinvent that for B2B, it might go a little something like this: "If the product is really convenient to use and only requires you to constantly upload your CRM database in order to deliver good results, then you are the product." Personally I see this in action when occasionally buying things over ads on walled gardens. No sooner have I made the purchase before I'm bombarded by offers from the same category. As a professional in the industry, it seems very plausible to me that the algorithms have honed in on me as a buyer of that category using data points from the actual brand I bought from.

Kristina: You say it isn't possible for companies like Facebook and Google to truly anonymize brand data collected - why is that? 

Kasper: A very popular way machine learning is deployed to help drive good campaign results is called supervised machine learning. It refers to the fact that a seed set of users, which have performed a desired action for a brand (such as buying one of its products), is fed to the algorithm(s). This allows the algorithm(s) to learn the specifics of the seed set and produce one or more models of what makes them stand out. The models are then applied to the entire population in order to score each user as to his or her propensity for buying the product in question.

Even if walled gardens anonymize certain attributes (e.g., blinding the fact that attribute A means that a consumer has bought a piece of high-tech carry-on luggage from brand X) this makes no difference to the supervised machine learning algorithms. They will detect with merciless and objective clarity that there is a high correlation between the presence of attribute A for a given user and this user's propensity to buy the product of brand Y - a competitor to brand A.

Hence to me, anonymization of attributes carries very little meaning in the context of supervised machine learning.

Kristina: What does this mean for businesses in the social space?

Kasper: Businesses generating a large proportion of their revenues via social media advertising need to start looking at other avenues to connect with consumers. Preferably, these should make you smarter from every interaction with the right consumers to better learn who to engage with, when and where and with which message. By taking control and ownership of the data assets generated from each interaction with relevant consumers, you take the driver's seat and start to build the foundation for true data-driven differentiation instead of using the same platforms and MOs as the competition.
Kristina: How can brands and marketers take back control of their data? 

Kasper: I think the first step is an attitudinal change toward control and ownership of the data assets that really drive engagement with the right consumers. Constantly giving a platform your CRM database in return for short-term campaign results is not sustainable and cannot form the basis for long-term data-driven differentiation. You're just doing what everybody else is doing.

Let's close with another old adage in the form of a question: "Are we selling them a fish every time they're hungry, or are we teaching them to fish?" We believe only the latter holds the promise of long-term, strategic customer relationships and that the former is a transient phenomenon in a young and still immature industry.

Tags: consumer data, control of data, data tips, data trends, Semasio, social marketing, social marketing data, social marketing trends, social network trends

Subscribe to BizReport



Copyright © 1999- BizReport. All rights reserved.
Republication or redistribution of BizReport content is expressly prohibited without the prior written consent.
BizReport shall not be liable for any errors in the content, or for any actions taken in reliance thereon.