TRUST NO ONE: WHY DECENTRALISED DATABASES ARE THE FUTURE
Published on November 26, 2020
It’s often said that trust is the glue that bonds a society together. Think of all the things that can’t take place without it: business agreements, political treaties, personal relationships, and simple transactions. Without trust, the world would simply crumble.
Today, the impact of trust equally extends to the digital economy’s hottest commodity—you guessed it, data. In a single second, a person creates 1.7MB of data, providing marketers and advertisers with invaluable insights as to how to best target and engage them based on their preferences and shopping habits.
This process, otherwise known as a value exchange, used to be far more fair and equitable. Consumers would willingly share their personal information with brands and publishers in exchange for a personalised shopping experience, distinguished for its convenience and its uncanny ability to anticipate their wants and needs. That is, provided that their data was stored securely and used appropriately.
Unfortunately, consumer-brand trust has eroded significantly over time. Despite the conveniences it can offer, technology has become increasingly intrusive in people’s daily lives, so much so that a 2019 Edelman survey revealed that only 34% of consumers trust most of the brands they choose to buy from. When compounded with high-profile cybersecurity scandals, be it across social platforms or e-commerce websites, it’s no surprise that consumers have grown wary.
However, the past year has seen us become more dependent on the internet with the ongoing coronavirus pandemic prompting significant spikes in online activity with reported surges of up to 70 percent as lockdowns came into effect. Malicious parties have taken note, taking advantage of the many golden opportunities in recent months to exploit loopholes in data management systems.
As concerns around trust and privacy are left unrepaired, this will ultimately diminish brand loyalty. After all, consumers are twice as likely to stay loyal to a brand they trust—even when presented with trendier or more innovative options. To stay relevant in this hyper-competitive economy, consumer trust is simply far too precious to ignore. A good starting point would be to re-evaluate existing data infrastructures and data management practices to rebuild customer confidence, assuring them that their data is safe in our hands.
At Aqilliz, we’ve recognised the value of collaborative, decentralised databases. Confused? We’ll break it down for you.
Centralised vs. Decentralised: What should I know?
In order to fully appreciate decentralisation in the context of data sharing and management, let’s first re-examine traditional approaches.
A centralised database management system stores, analyses, and maintains data in one single location. This approach is commonly used for customer relationship management tools (CRMs) and data management platforms (DMPs), where a data repository is stored on an organisation’s central server or a mainframe computer.
Head over here to check out how we break down CRMs, DMPs, and CDPs for you.
However, with the swathes of data being ingested every day, relying on a single centralised system simply isn’t sustainable. Not only is it costly as storage requirements grow, but it also exposes organisations to significant security risks. This past March, a hospitality giant was fined £18.4m by the UK’s Information Commissioner’s Office for General Data Protection Regulation (GDPR) violations following a high-profile data breach, where personal records of up to 339 million guests were compromised following a series of cyber-attacks which exposed sensitive information such as names, contact information, and even passport details.
Continuing to store all data on a centralised location also means there will be an increasing need for capacity expansion. Simultaneously, the likelihood of a data breach increases during processing or transferring activities.
Beyond the regulatory requirements stipulated and the potential of amassing costly fines, the severe consequences of losing consumer trust are priceless. Instead of opting for centralised data repositories, the industry needs an interoperable, third-party platform where information can be securely shared and processed transparently. Switching to a decentralised approach to data management eliminates our reliance on a central administrator, and with it, the risk of losing all consumer data should an attack happen.
On this decentralised platform, first-party data obtained directly from consumers can be injected by participating brands and publishers, then shared across the network. This infrastructure can ensure utmost security, as its decentralised nature, powered by the distributed ledger technology (DLT), means that all data is shared securely across all participants. Even with one node on the network under attack, the rest of the network will not be compromised since access will be limited to that particular data node.
Furthermore, DLT ensures that all records of processes and transactions associated with a given data point benefit from immutability. This means that all records are tamper-proof and cannot be retroactively modified unless approved by the data owner and all authorised stakeholders. With that, brands, publishers, and platforms can easily comply with stringent documentation requirements stipulated in privacy laws, such as Article 30 of the GDPR on Records of Processing Activities (ROPA). This requires companies to maintain records of all data processing activities including the purpose of processing, data categorisation, and list of data recipients.
With the GDPR as the gold standard, it should come as no surprise that a similar clause was included in the recently-passed California Privacy Rights Act, which is set to replace its predecessor—the California Consumer Privacy Act—by 2023, pointing to the growing need for infrastructures that enable data provenance.
Regulatory compliance aside, data provenance allows brands to re-earn trust by being far more transparent and accountable. In effect, control is given back to consumers, providing them with end-to-end visibility on how their personal data is stored and used.
With the issue of security addressed, some may question how consumers’ rights to privacy can still be maintained if their information is shared with other brands, publishers, and vendors on a decentralised platform.
Right here, we’ve pioneered the use of federated learning to ensure this. With federated learning, all data analysis takes place at local storage—machine learning models are deployed to directly analyse data on-device and only relevant, aggregated summaries are sent back to a server. This also ensures that data analysis takes place in a decentralised manner.
To further enhance consumer anonymity, cryptographic algorithms such as differential privacy inserts “statistical noise” into a data set before being processed and aggregated for analysis. This allows for the further masking of individual identities as they are averaged with one another and ensures that the data cannot be reverse-engineered to be linked back to a single individual.
Less is Sometimes More
Today’s abundance of information has equipped marketers with powerful tools to accurately target and engage their audience. However, there is such a thing as “too much” when it comes to data and as we look towards a far more compliant ecosystem, marketers would do well to move away from an ethos of data maximalism. With cost-efficiency and security in mind, it’s clear that collaborative data-sharing can help mitigate the associated costs of centralised storage while addressing the risk of potential data breaches.
Through the adoption of decentralised technology, cross-organisational exchange of information is no longer impossible. Switching to a federated identity management platform allows marketers and advertisers to foster a more collaborative dynamic: one where meaningful insights can be shared and one where ethically-obtained, first-party data can be securely leveraged while upholding the principles of transparency and privacy.
In our view, decentralisation will remain the best way for marketers to maintain or win back trust from consumers for the foreseeable future. However, to fully yield results, more players need to hop on this bandwagon of decentralised collaboration for a more efficient, ethical, and unified marketing environment for all. So, what are you waiting for?