Our Terms of Use & Cookie Policy

StackAdapt uses cookies to improve your online experience. Cookies are used to help us better understand where our visitors are coming from, recognize you when you sign in to the platform, or better personalize pages you visit. Cookies are placed on your computer automatically but if you chose to opt out from any tracking, you can change your cookie settings through your browser.

Hide

Data Science in Advertising: How Thompson Sampling is Revolutionizing Performance

As a Data Scientist at StackAdapt, my role is targeted primarily at two things: improving the user experience and improving the performance of our ad campaigns. My team’s objective is to reach users who want to engage with the ad content (creating an non-disruptive viewing experience) and to ensure our clients hit their targeting metrics such as desired click-through rates (CTRs), low effective cost per click (eCPC), placement on premium domains, high user engagement rates, and more. We meet these objectives through a variety of innovative techniques, but one, in particular, stands out from the pack. Here is how Thompson Sampling is revolutionizing campaign performance.

Predicting CTRs

Before we get into any specific techniques, it is important to note that the key to achieving the two objectives listed above is the ability to predict CTRs. One of the most difficult things in data science is determining the “features” or pieces of information required to make this prediction. The information can be broken down into several sources: information about the advertisement itself, information about the publisher or website where the advertisement is displayed, and information about the interests of the user.

One of the most difficult things in data science is determining the “features” or pieces of information required to make this prediction.

For example, information about the advertisement itself might include the words in the title of the advertisement, or information about the image in the advertisement. Information about the publisher or website may include a categorization of the website’s content using StackAdapt’s custom content categorization engine, and information about the user may include their past behaviour when engaging with StackAdapt’s advertisements.

Thompson Sampling

In addition to predicting click-through rates, StackAdapt’s data science team regularly creates new methods of executing campaigns, such as alternate ways of bidding and placing advertisements for an advertiser. Bidding for advertisements is closely tied to click-through rate prediction because, intuitively, one wants to bid higher on user-publisher-ad combinations that produce high click-through rates.

Bidding for advertisements is closely tied to click-through rate prediction because, intuitively, one wants to bid higher on user-publisher-ad combinations that produce high click-through rates.

Recently, StackAdapt’s data science team has applied some novel implementations of Thompson Sampling, a method of automatically selecting bidding strategies, to improve the performance of campaigns on StackAdapt’s platform.

For example, suppose that there are two ways of executing a campaign, method A and method B. Ideally, we would want to know which method produces better metrics for a specific campaign and use only that method for that campaign. However, it takes some time to learn which method has better performance.

Thompson Sampling is a methodology for acquiring that knowledge. Specifically, it employs method A for some time, then method B, then switches between the methods probabilistically, based on their performance. This method is highly optimized and even though it is learning as it goes, it is able to achieve performance that is similar to having apriori knowledge of the better performing method.

This method is highly optimized and even though it is learning as it goes, it is able to achieve performance that is similar to having apriori knowledge of the better performing method.

Soon all campaigns on StackAdapt’s platform will have access to multiple bidding strategies and be employing Thompson Sampling to automatically optimize their performance. This will allow StackAdapt to deliver better click-through rates, lower effective costs per click, and higher user engagement rates than in the past.

The central feature of this optimization is that it is on a per-campaign basis. In other words, we are able to automatically do what is best for each campaign based on that specific campaign’s goals and performance.

Using click-through rate prediction and Thompson Sampling opens the doors to creating more custom bidding strategies and leveraging our vast amounts of historical data to automatically optimize performance.

Using click-through rate prediction and Thompson Sampling opens the doors to creating more custom bidding strategies and leveraging our vast amounts of historical data to automatically optimize performance.

In the future, we have plans to completely reinvent our core bidding methodology on a per-campaign basis, so that each campaign bid is based on the campaign’s goals while at the same time producing engaging, unintrusive advertisements that a user wants to engage with.

4 min read

Performance Marketers Are Increasingly Relying on Intent-Based Targeting

In 2016 StackAdapt released proprietary technology we call Custom Segment. Named for the intent-based audience segments it generates through machine learning and natural language processing, these segments are becoming increasingly popular with our clients.

Made up of users who have actively shown intent to purchase on a rolling seven-day basis, relevant users are constantly dropping into a brand’s audience pool while old data is removed, ensuring marketers reach interested audiences in a timely manner. As the popularity of this targeting technique increases, it is clear that performance marketers are increasingly relying on intent-based targeting to reach the right person at the right time.

Popularity of Custom Segments by the Numbers

In the month of February 2018, the number of advertisers actively using StackAdapt’s Custom Segment technology to target unique audience pools in their campaigns rose by 35%. Advertisers who run conversion campaigns have the highest adoption rate of custom audience segments at 56%. Overall, 30% of all impressions being served on the platform are now reaching users through StackAdapt’s custom audience segments. What are the reasons behind the increasing popularity of intent-based targeting?

Advertisers who run conversion campaigns have the highest adoption rate of custom audience segments at 56%

Why Are Marketers Relying on Intent-Based Targeting?

More Relevant Audiences

From my perspective as a Business Intelligence Analyst, I believe intent-based targeting is becoming more popular among marketers and advertisers because it allows them to collect more relevant audiences than with third-party or contextual targeting. Often, clients cannot find their narrow audience with third-party segments, or the third-party segments available are too broad to reach bottom-of-the-funnel intent (something they need to capture in order to acquire more customers).

I believe intent-based targeting is becoming more popular among marketers and advertisers because it allows them to collect more relevant audiences than with third-party or contextual targeting.

An example of this would be an airline company who wants to target travellers going to Florida. The most relevant third-party segment available might be “Domestic Travel”, “Frequent Travel”, or “Vacation Travel” but these people could be going anywhere. With intent-based technology we can home in on people specifically reading about Florida vacations, browsing for hotels & lodging in Florida, and searching up flights to airports in Florida.

Transparency

Another reason for the increased popularity of our intent-based targeting is that we provide complete transparency into the topics or pages that the users are visiting. Marketers no longer have to guess where these users are coming from or why exactly they were targeted: it’s right there in the Custom Segment dashboard.

Increased Engagement & Conversions

Intent-based techniques like custom audience segments often have a lower than average click-through rate (CTR) when compared to campaigns using other targeting methods, but this hasn’t deterred performance marketers. This is because users who do click are more likely to be engaged, spending at least 15 seconds or more on the site, and ultimately, are more likely to convert. Just as JPMorgan Chase slashed its programmatic scale from 40,000 to 5,000 sites, many performance-driven marketers are choosing to bid smarter. In recent experiments by clients who have A/B tested custom audience segments against third-party segments in conversion campaigns, custom segments have generated 20-75% lower cost-per-conversion.

As intent-based targeting moves into the mainstream, I am interested to see how agencies and brands get creative in leveraging the increasingly intelligent technology at their fingertips.

4 min read

Women of StackAdapt Voice Challenges Still Faced by Women in Today’s Society

In celebration of International Women’s Day 2018, we are highlighting the amazing women at StackAdapt as they voice the challenges still faced by women in today’s society and how we can better support women both individually and as a collective.

 women-of-stackadapt-kristen

Kristen Singh
UI/UX Designer

A challenge women still face today is having the ability to speak with confidence and assertiveness without being labeled as “Bossy”.

Your Role

As a UI/UX Designer at StackAdapt, my goal is to create a world of better experiences through design thinking. I work with our amazing Product team to improve our product through testing, research, and iteration. Nothing is more satisfying than knowing that I’ve helped to change someone’s routine for the better.

A Challenge Women Still Face

A challenge women still face today is having the ability to speak with confidence and assertiveness without being labeled as “Bossy”.

How Can We Support Women?

Whether it’s International Women’s Day or not, I always aim to uplift and support fellow women. Many of us live the same struggles and it’s important to build an alliance of love to help strengthen one another. Attending meetups and connecting on LinkedIn, going out for coffee, and helping mentor younger girls are all ways you can help, and it’s definitely how I help.

women-of-stackadapt-BDRs

Dorothea Kefer and Tessa Walsh
Business Development Reps

A challenge women continue to face, despite generational gaps, is double standards between men and women.

Your Role

As Business Development Representatives, our role at StackAdapt is to be the entry point of new business opportunities at StackAdapt! We’re growing the company one prospect at a time, and we’re proud of it.

A Challenge Women Still Face

A challenge women continue to face, despite generational gaps, is double standards between men and women.

How Can We Support Women?

To close the gap, we try and raise other women up by encouraging every member of the team to use their voice and support each other.

Stefanie Zaarur
Digital Marketing Manager

women-of-stackadapt-stefanie

Equal pay continues to be a big challenge. It’s important for women to be encouraged to ask for what they deserve.

Your Role

My role as Digital Marketing Manager at StackAdapt entails several responsibilities such as building and optimizing digital marketing initiatives, to increase quality and flow of leads from marketing to revenue.

A Challenge Women Still Face

Equal pay continues to be a big challenge – women across industries generally make less for the same role. It’s important for women to be encouraged to ask for what they deserve without the fear of not getting the role they want.

How Can We Support Women?

To change societal mentality I try to break gender stereotypes and encourage my female friends, family members and colleagues to do the same on a daily basis.”

Natalia Vassilieva
HR Manager

women-of-stackadapt-natalia

It’s important to cultivate an inclusive culture for women.

Your Role

I started at StackAdapt over 2 years ago as employee # 20 and have watched this team grow and blossom into what it is today– 70 (and counting) of some of the most talented, bright, driven people I have ever had the pleasure to work with. Importance of my role? To keep bringing on amazing people to StackAdapt and ensuring everyone loves coming to work every day! 🙂

A Challenge Women Still Face

Well, I think it’s still tough for women to make the decision to take time off to start a family.

How Can We Support Women?

We can help by cultivating an inclusive culture for women. Whether it’s from the moment they walk through our doors for their first interview, or the supportive Mat Leave policy we design that support them when they start their families.

Kaeshi Sri
Marketing Automation Manager

women-of-stackdapt-kaeshi

A challenge I find women still face is the notion that just because you’re an introvert and a female that you can’t achieve the same goals as men.

Your Role

As a Marketing Automation Manager, my goal is to strategize and execute on an integrated marketing mix of emails, social media, webinars, and other digital tactics for lead generation at StackAdapt.

A Challenge Women Still Face

A challenge I find women still face is the notion that just because you’re an introvert and a female that you can’t achieve the same goals as men. We need to overcome that notion, and we can all do this by highlighting our strengths and ideas as it encourages and brings out the best of those around us. We often fail to recognize the women that have made a difference in their field.

How Can We Support Women?

I try to contribute by sharing women of our time, such as Ada Lovelace, Florence Nightingale, Hellen Keller, and making sure to remember them for their contributions to our advancements.

Amika Evans
Account Executive

women-of-stackadapt-amika

As women, we need to stop selling ourselves short.

Your Role

As an Account Executive, I help drive revenue. Once the new business comes in, I work to close the said opportunity.

A Challenge Women Still Face

That being said, a challenge women in sales still go through is the idea that people still expect a guy to be on the other end of a call. #WomeninBusiness

How Can We Support Women?

As women, we need to stop selling ourselves short. I find some women tend to condemn or doubt their female counterparts. We are faced with so much pressure (professionally, our bodies, relationships), and frankly, misery loves company, so it’s just become easy to be mean. I will not let that come to fruition in my circle, and just continue to uplift my sisters. Sometimes we just need to hear the positive reinforcement from each other.

Maggie Clapperton
Content Writer (Freelance)

women-of-stackdapt-maggie

One-third of global businesses have no women in senior management roles, up 0% from 2011.

Your Role

As a content writer, my role is to help agencies and brands understand the value of programmatic native advertising and provide resources that help them succeed.

A Challenge Women Still Face

In business, women are still underrepresented in the c-suite. According to Catalyst.org, one-third of global businesses have no women in senior management roles, up 0% from 2011. Yep, 0%…

How Can We Support Women?

On IWD and every day, I provide support (and I am supported by) the extraordinary women in my life. This IWD I will be marching in the Barcelona Women’s March in solidarity with every woman who has ever been harassed, judged, assaulted, paid less than her male counterpart, or simply made to feel small and voiceless.

Michelle Hart
Senior Customer Success Manager

women-of-stackadapt-michelle

The confidence to speak up in public is something that a lot of women still struggle with. The idea that their opinion or point will not be valued or accepted should never hold them back!

Your Role

As a CSM, my main objective is to ensure clients are well equipped with the resources, training, and recommendations they need, allowing their campaigns to deliver on the goals they are looking to achieve, ensuring success on our platform!

A Challenge Women Still Face

The confidence to speak up in public is something that a lot of women still struggle with. The idea that their opinion or point will not be valued or accepted should never hold them back!

How Can We Support Women?

I continually try to motivate and encourage the women in my life. Females are strong, intelligent, and powerful, so it’s very important to me to remind and help ensure other women feel their strongest to achieve their goals and always be their best self!

For more “Women of StackAdapt” profiles, visit our Instagram @stackadapt_ 

4 min read

Travel Marketers Are Focused on Mobile, But Are They Considering the Entire Mobile Journey?

According to eMarketer, travel brands plan on investing in a variety of marketing tech in 2018, and the key focus is rightly on one thing: mobile. As illustrated in the ComScore chart below, our mobile phones have become our primary digital tool. Online Travel Agencies (OTAs), Hospitality, and Transportation all name mobile experience and analytics as their top marketing priority for the year. But analyzing eMarketer’s results, I can already detect glaring oversights in the way the industry is splitting its focus. Travel marketers are focused on mobile, but are they considering the entire mobile journey?

comscore-mobile-primary-tool

While some sections of the travel industry, notably OTAs, plan on leveraging emerging technologies like Artificial Intelligence (AI) and Machine Learning (ML) (19%) to boost their mobile presence, the majority are not:

eMarketer_Technology_Priorities_for_Their_Company_According_to_Travel_Marketers_in_North_America

An Unnecessary Divide

You’ll note that programmatic buying/retargeting, and AI/ML are far down the list of technology priorities for 2018. But these technologies are by and large what make mobile marketing sophisticated, frictionless, and non-disruptive. That’s because these technologies power mobile-first marketing techniques such as programmatic native advertising.

The practice of segregating mobile experience investments from investments in programmatic and AI technology demonstrates a lack of technological understanding common across the marketing landscape regardless of industry. The mobile experience doesn’t start on your web property, it begins well before that, during marketing & advertising distribution.

The practice of segregating mobile experience investments from investments in programmatic and AI technology demonstrates a lack of technological understanding common across the marketing landscape…

Mobile From First-Touch to Last

“To me, mobile experience means the interaction with consumers across the entire journey,” says StackAdapt COO Vitaly Pecherskiy. “From the first point of interaction on mobile, be that on a social feed or a web publishers’ property, marketers need to think ‘how can we speak to our target audience in a way that sparks interest in what we have to say.’ Building on-site mobile experiences is a great step — that’s where you need to start. But ultimately, it’s the distribution that makes or breaks marketing campaigns or initiatives. Most marketers fail before a user even arrives at their perfectly polished website because they lack the tools to reach in-market audiences across the modern web landscape.”

Most marketers fail before a user even arrives at their perfectly polished website because they lack the tools to reach in-market audiences across the modern web landscape.

In other words, investing in a mobile experience goes beyond investing in a mobile-friendly site. It starts with the way you spread the word. If you’re using traditional, desktop-centric marketing tactics to reach a mobile audience, you’ve already lost. There is only so much room on a phone screen and users hate to be interrupted from their intended browsing experience. Enter programmatic native advertising.

Fitting In

Ask any tech-savvy marketer or digital agency director, and they’ll tell you programmatic, specifically, programmatic native advertising is a true mobile-first approach. Business Insider reports that native ads are poised to take over mobile by 2020. The reason, according to the Mobile Marketing Association (MMA), has to do with the way native blends in with the consumer’s intended browsing behaviour: “When ads fit with the environment where they are served, people notice. This is the inherent benefit of mobile native advertising, which is a form of paid media where the ad experience follows the natural form and function of the user experience in which it is placed.”

In a word, the ads ‘fit’. The in-feed nature of the ad is adapted to the scan and scroll behaviour of users across mobile apps and browsers, leading to increased click-through rates, longer time on site, and even increased conversion rates. For example, Porter Airlines decreased their cost per customer acquisition by 35% using in-feed native advertising and the corresponding intent-based targeting.

Cross-Device Targeting & Attribution

Moving beyond the form and function of programmatic native, Machine Learning (ML) and other aspects of Artificial Intelligence (AI) allow for sophisticated, cross-device targeting and attribution. In a 2017 survey, StackAdapt uncovered that 47% of consumers still aren’t buying on mobile and 82% of consumers who do buy on mobile say that small screens and poor landing page experiences often deter them from converting directly on their smartphones.

Moving beyond the form and function of programmatic native, Machine Learning (ML) and other aspects of Artificial Intelligence (AI) allow for sophisticated, cross-device targeting and attribution.

This means that (much to marketers chagrin) transactions become muddy and attribution unclear. In a video titled “Your Attribution Model is Missing a Third of Your Customers” Pecherskiy points out that “38% of people discover a product for the first time on mobile and then they choose to use desktop to actually complete the purchase. So, oftentimes marketers overvalue direct or organic traffic from desktop, when in reality more than a third of consumers discover your product for the very first time on mobile.”

38% of people discover a product for the first time on mobile and then they choose to use desktop to actually complete the purchase.

By integrating AI into our platform, StackAdapt has built intent-based targeting technology that targets users with in-feed native ads across devices and locations, and with the implementation of conversion tracking, provides insights into the time the user first clicked on your ad, the website that ad was placed on, the time to last click before they converted, and even the devices that these users were on.

Performance Optimization

As you can see, in-feed native advertising drives in-market audiences to mobile properties and has the potential to track the conversion journey from first-touch to last. The third point in its favor is that the same programmatic and AI technology that gathers these insights also allows for ongoing optimization of campaign performance by capturing and storing, then learning from, and finally, utilizing the learnings of the data in a closed loop of ever-increasing intelligence.

AI technology allows for ongoing optimization of campaign performance by capturing and storing, then learning from, and finally, utilizing the learnings of the data in a closed loop of ever-increasing intelligence.

If travel marketers, indeed all marketers, plan on in investing in mobile experiences, it is integral that they consider the entire mobile journey, from first-touch to last; from the way you distribute your content, to the way a user consumes your advertisement, to the way they convert on your web property. It makes no sense to create innovative mobile experiences without reaching audiences through mobile-first tactics. Programmatic, retargeting, and AI technology are no longer a “nice to have” and are no longer limited to the big name elite. These technologies are a key function of mobile first marketing, specifically the front-runner: native advertising.

4 min read

User Experience in Ad Tech Sucks and No One Seems to Care

The internet is littered with articles berating ad tech for the complexity of the system. And it’s true, the process of an automated ad impression reaching a specific user the moment he or she opens a webpage is complex, some would argue unnecessarily so. But a fundamental and vastly overlooked aspect of this complexity is rooted in the software that powers the process. Continue reading “User Experience in Ad Tech Sucks and No One Seems to Care”

4 min read