Advertising with privacy: How SoMin employs AI to build brands and preserve anonymity online

Ads are everywhere. From expensive billboards to car door stickers, ads occupy every visible surface wherever we go. At the turn of the century, advertisers began to shift their focus online.

In the span of twenty years, we established a thriving ecosystem where brands can reach their targeted consumers in increasingly creative ways across the web.

Whether it is Facebook ads, Spotify banners, or pop-up notifications, consumers today are bombarded by ads nearly every second of every day.

Whilst receiving relevant ads may value-add to our busy lives, it does get a little annoying when you see that same pair of sneakers you searched for just once start appearing across all the web applications that you use.

In the end, cheap advertising that simply shoves products and services in consumers’ faces stand a higher chance of backfiring on their advertisers instead.

The rise of advanced adtech

Riding on our good intentions to deliver relevant ads to consumers across the online space, privacy issues have been sidestepped and sacrificed by more and more advertisers. Cross-site tracking, made possible by the sharing of individual user data across global companies, have led to an erosion of trust in consumers.

According to a study by Pew Research Center, 72 per cent of people feel that almost all of what they do online is being tracked by advertisers, technology firms or other companies, and 81 per cent say that the potential risks they face because of data collection outweigh the benefits.

This means that digital advertising will continue to lose its value if advertisers do not address the growing concerns that people have about their lack of privacy online.

Thankfully, the advent of sophisticated adtech tools has made it significantly better for both advertisers and consumers. Advertisers no longer have to rely only on hard-sell tactics online that have been seeing declining ROI (return of investment) per marketing dollar.

Consumers will also benefit from the increased privacy and absence of in-your-face advertisements that clog up news feeds. All in all, sophisticated adtech tools create a win-win situation for both businesses and consumers alike.

However, building sophisticated performance tools requires skill and a wealth of industry knowledge. Fortunately, a team of experts at BLOCK71 Singapore have been working to build the future of ad optimisation since 2017.

This month, I sat down with Hendrik Schwartz and Aleks Farseev, good friends and co-founders of adtech startup SoMin, to dissect the intricate workings behind their AI-driven marketing performance system.

Read on to discover the game-changing powers of ad optimisation for both businesses and consumers.

In a nutshell, what does SoMin do?

The goal of SoMin is to help businesses innovate their marketing approaches through AI and automation. Our main product is a marketing performance tool that uses machine learning to help brands increase their ad cost-efficiency. Through different models, it aims to define and understand audiences through public data while preserving their anonymity.

From there it figures out different ways of approaching generalised audience clusters then uses that knowledge to automate the deployment of campaigns.

SoMin Marketing Performance Platform | Source:

Adtech solutions that provide customer insights for businesses are a dime in a dozen. How does AI make it better?

There are many ad tech solutions out there, but the problem with many of them is that they take an isolated approach to solve industry challenges.

Ad performance solutions are a good example of this. What most of them do is take the data found within the bidding ecosystems and use this to try and improve performance. Whilst there is a wealth of data involved in this, there are also a lot of other important elements that are not taken into consideration.

Factors such as initial audience definition, competitor movement, creative evaluations, and more are all important to the success of a campaign but their definitions are usually outside the realm of the bidding platforms themselves.

This is where the power of AI helps. Through a machine’s capability of understanding unstructured data, factors outside of a bidding ecosystem can be quantified and plugged into the entire process of ad deployment.

And when it does this it opens up different strategies that were not previously possible. Coupled with automation, we are able to come up with a much more holistic solution to what would make a successful marketing campaign.

Marketing strategies differ across industries. Are there any industries that SoMin’s solution may not work as well in? Why is that?

With the current product, the strength of the solution caters to B2C (business-to-consumer) mass audience targeting. Our system works well with many of those industries.  That being said we do not provide personalised advertising so for strategies that require those we have yet to make the decision to venture into that space.

This is mainly because personal advertising touches more into privacy concerns and we believe as a company that the wealth of public data is already enough to create effective marketing.  The trick is being able to use what we have effectively which is why we build machine learning modules just to do that.

Influencer marketing has started trending amongst many B2C firms in recent years. How can SoMin lead the change to revolutionise advertising in this new space?

We are also venturing into the influencer space. At current, our solution allows for brands to apply machine learning in their search for suitable brand influencers.

One of the biggest problems in influencer marketing is the curation of applicable influencers. The work is very tedious and drains many man hours, which then hinders the scalability of the activations. Through machine learning, we’re able to profile influencers better and match it to brands so that we can find better fits and ease this process.

We’re also further developing this feature and we now have a platform called Sopop that directly ties with the SoMin feature. This solution is catered towards influencers and addresses their most pressing needs.

Our plans involve helping them with data knowledge, financial tracking and engagements, brand matching, and collaboration tools. Put these in conjunction with the brand side solution, we believe we can really help scale the industry overall.

What do you see as the future of performance marketing?

If you ask me, the biggest hurdle in performance marketing is how brands – with their human-generated marketing campaigns – interact with platforms that are primarily driven by artificial intelligence. So much could be done, but our ability to harness the data that is available to us is so small that it becomes an area of missed opportunities.

This is also tied with the data issues we face today.  Because we lack the ability to harness all the data that we already have, we crave for more data to the point where we step on the grey line of privacy.  But so much data is already available and we just have to digest it properly so that we don’t have to come close to stepping on people’s rights.

Now the trend with AI is democratisation and I believe this would truly help performance marketing. In the future, by making AI systems more interactive, understandable, and available, it would allow brands to make full use of the data that’s available to them. This would in turn create different kinds of strategies or ways of working that will improve communication between brands and customers.

With SoMin’s expertise and forward thinking, we believe we are well placed to harness the full potential of performance marketing.