Navigating Privacy - how to win in a challenging ecomm landscape

Fuelled by a series of high stakes data leaks and a global shift to living online during the pandemic, cybersecurity sits at the forefront of social, political and economic discussions. The importance for businesses to protect their customers' personal data has become paramount and the subject of new laws and regulations. Having your databases’ information exploited due to insufficient security is a situation that could jeopardise any business, but especially start-ups seeking to establish credibility and attract new customers.

Navigating this is difficult, particularly due to the changing regulations and the impending death of third-party cookies. In order to find out what this means for scale-ups, I have enlisted the help of James Taylor, the CEO of Particular Audience. Particular Audience is an API software that allows retailers to offer personalization without compromising personal data.

Launched in 2019, the company saw $100 million in gross merchandise value through its products in 2021. This demonstrates that the demand for collecting secure data is most definitely there, and I asked James why.

“There’s been a huge amount of bad media surrounding data leaks, and consumers are now much wiser to the fact that they can have their privacy and still get access to a lot of things,” James tells me.

A number of surveys show that internet users do care about their data security but will still use nefarious sites due to the perceived benefits (cough, Facebook). Now however, internet users across the globe are putting their foot down and demanding better protection. Regulatory change is on the horizon particularly in Europe and parts of America.

“Australia is undoubtedly going to have to follow the example that is being set by overseas regulatory bodies,” James says.

I wanted to know who James considered was already doing this well, and who start-ups could look to for inspiration.

“The best examples of companies doing privacy well are the companies that have no use for personally identifiable information whatsoever… platforms like Spotify and Netflix don't need to know anything about your demographic information to show you content that you like, or content that you're most likely to want to engage with.”

These platforms sit on one end of the advertising spectrum and offer what is called “contextual advertising”. These do not require third party cookies to be effective but are still likely relevant to the user. For example, on a recipe website, the advertisements displayed may be the latest kitchen appliance or a celebrity chef’s cookbook.

“This is the sort of advertising that is going to persist in a world with no third party cookies. On the other end of the spectrum, you have the ads that follow you around the internet, showing you the shoes you clicked on once and making you feel as though your privacy has been infringed.”

James argues that collecting personal data from your customers can also be largely ineffective.  

“Trying to make a prediction based on demographic or age is not a great indicator of someone’s interests. Things like suggesting content to watch after you finish a series you rate highly or new songs after you listen to a playlist are more likely to be effective. These are predictions based on context rather than anything related to the individual. The analogy is ‘Prince Charles and Ozzy Osbourne have the exact same demographic makeup, but their interests are vastly different'.” 

James argues that a tenuous link between a person and a product is not likely to produce sales on e-commerce websites. For example, suggesting umbrellas to people if it is raining in their location is not really a suitable gauge of interest. Although this is technically a way of using personalisation, it is probably not a super effective use of data.

I wanted to know how start-ups can win in this challenging e-commerce space, and what the best method of advertising is.

“There is wisdom in the crowd-based algorithm, determined by behavioural data. However, with a site catalogue that’s growing, you have unseen products and minimal customer input to drive the algorithm. Using alternate approaches like natural language processing to find out which items have similar attributes to other items, or computer vision to see which are visually similar are ways to automatically filter websites.”

These examples also require no data from users, but still provide a more relevant experience for each individual that comes to the site.

Are you in the early stages of a startup and wonder how best to put the right products in front of your customers? Or perhaps you are wondering how to navigate a world without third party cookies? James gives some great insights into how you can overcome these challenges and what he believes the future of e-commerce advertising holds.