Impact’s VP of Global Sales and Strategic Partnerships, Regina Gray, sat down with Noa Ashkenazi, Chief Operating Officer of Bright Storm, to discuss how Bright Storm fights ad fraud with Forensiq, the role AI plays at her organization, trends she’s seeing in the industry and more. Read on…
Regina: Thank you so much for sitting down with me today. Can you please tell me a little bit about Bright Storm and your role there?
Noa: Of course, thanks for having me. Bright Storm is a programmatic DSP for app developers and brands focused on ROAS. After many years in the Ad Tech industry and working with 200+ media platforms, we saw that automation in current systems could only get you to a certain point and from there, the human marketers’ hand needs to complete the analyzing and optimizing process. This was limiting the performance, scale and the actual value that advertisers were getting. So, with that in mind, we took upon ourselves the challenge of creating a technology that would answer all the needs for both advertisers and marketers’ goals.
As I see it, the numbers will always tell you the true story. Every story is unique, you just need to know how to read, analyze and make decisions in real time where to navigate the plot.
Our technology knows how to translate all the data we accumulate across the funnel into automatic decision-making AI to create a customized algorithm for each campaign.
Bright Storm was established to answer the ever-growing need for brand-safe, top quality media for mobile advertisers while generating real value to exceed ROAS goals.
My job basically is to oversee the company’s global business operations, driving the company’s product capabilities to surpass customer satisfaction and retention as well as spearheading strategies to steer the company’s future in a positive direction.
Regina: That sounds like really interesting technology. What are some of the benefits for brands and media partners who work with Bright Storm?
Noa: I would have to say the main benefit that distinguishes us is first, the capability of our AI to create a customized algorithm per campaign as opposed to implementing the same algorithm to all campaigns across the board. Our system analyzes the user’s reaction to the specific product, while considering the targeting, creatives and KPI goals. The AI makes decisions in real-time based on campaign data of the entire user flow from pre-bid to post-install revenue events in order to create incremental growth.
Another value we provide our advertisers is that our service is fully managed, the advertiser gets a login and can see everything we see but without having the hassle of allocating internal resources towards buying media on a new unfamiliar self-serve platform.
Regina: AI clearly plays an important role for Bright Storm much like it does for our Forensiq product to detect fraud. Can I ask how you first learned about Forensiq?
Noa: In a world where fraud and bots are polluting our advertising environment, we’ve set it as our top mission to eliminate all fraudulent activity.
As part of our vision, we wanted to ensure we provide 100% quality and brand-safe traffic to our advertisers. After testing many fraud tools, Forensiq proved that they have a strong technology that integrates with our technology in a great synergy. It’s rare to find partners that share the same vision and philosophy about the Ad Tech industry and we’re very happy with the extra safety net and transparency they provide to our partners and ourselves.
Regina: Thank you! So how do you plan to use Forensiq exactly?
Noa: We are currently working with Forensiq on our DSP with the following solutions: post-bid analysis through FQ JS tags as well as post-install verification through sending S2S post backs. We’re also very keen on testing out the pre-bid solution via API to block suspicious activity on the bid level before the impression.
So far, the post-bid solution has proven to be very valuable and useful to our BI and data management teams as well as help optimizing the campaign’s performance. Since we started working side by side with Forensiq’s dedicated team, we’ve managed to analyze large volumes of data while having the confidence that our traffic is brand safe. According to the data we analyzed so far together with FQ, stats show that our traffic comprises of 99.5% real users. Now our main challenge is to make it 100%!
Regina: That’s pretty impressive! Can I ask what are your expectations of our team and Forensiq are?
Noa: We expect Forensiq to keep doing what they do best, which is always being a step ahead of the fraudsters. We realize the war on fraud is an ongoing struggle, so we do expect FQ to keep developing their technology to fight even the most advanced malicious activities while being open to hearing feedback and brainstorming together to find more features and technical solutions that can benefit Bright Storm as well as our partners.
Also, we always believe in having absolute full transparency with all our partners including FQ and we expect to keep having a strong partnership with open communications in the future.
Regina: We’re a firm advocate of transparency and building strong partnerships. Switching gears a bit, can you share some trends you’re seeing in the industry?
Noa: One of the strongest trends we currently see in the industry is that since the ecosystem is so polluted with fraud, advertisers are starting to realize that in order for them to get true value and engaged users with high monetization, they need to shift their focus from trying to lower CPIs as much as possible, towards focusing on CPM and on ITP (install to purchase) rate. Once the bidding is done on CPM, this shift in focus gives them more control to balance the entire conversion funnel and the overall performance.
In addition, advertisers are starting to ask for much more transparency from their media partners in order to have better control on where their ads are shown while getting more information on what their ideal target audience would be.
Regina: That’s insightful. Building on that, do you have any advice for brands or media partners in the space grappling with fraud?
Noa: One of the most important things in fighting fraud is to try to detect non-human behavior as well as the ability to block any malicious activity in real time before it creates any damage to your brand.
Monitoring and comparing the CTIT (click to install time) from the App Store and the CTOT (click to open time) from the attribution platform, can help detect and prevent fraudulent activities. Also, always look out for abnormal user behaviors in massive scales such as big retention drops from day to day, fake purchase events etc.
Always work with a small number of partners you really trust, ask for full transparency (including passing impressions, app names etc) and make sure you work with partners that can prove they allocate substantial resources towards eradicating all fraudulent activities.