Articles by: "Haley Adam, Product Marketing Manager—Lotame"

Marketing professionals, sales teams, and business owners are constantly chasing new audiences to target. This is only natural, as the more your grow your business and the more customers you reach, the greater your success. It is the core element of all business: the need to achieve stronger and greater consumer support.

But we all know that it’s not a “piece of cake” targeting new audiences, especially when they are beyond who you normally encounter or engage with. There are a lot of unknowns in increasing your audience targeting size. And marketers and publishers need to evolve with the technology around them.

While Lookalike modeling isn’t a term that is new to the marketing technology industry, it is still widely misunderstood. It is a tool that has been used by companies to expand their digital audiences while maintaining clear and relevant targeting practices. This piece will take a look into what Lookalike modeling is, how it works, and why it is so important for advertising campaigns.

What is Lookalike Modeling?

If you are looking to increase your targeting efforts with high performing audiences, the answers don’t lie in a mass chain of random messaging to consumers who have zero interest in your product. Instead, your focus should lie on your high performing audiences. What is it about your highest performing audiences that set them apart from the rest? Do they have common interests? Are they from the same geographic area? The best way to acquire new, high performing visitors is to focus on users who resemble your existing visitors, the users who have already shown an interest in your product or service.

Lookalike modeling is a process that utilizes machine learning to statistically analyze a given seed audience (already high-performing audience), identifying the demographics, characteristics, and different combinations of those and other data points. This creates new audiences composed of users that match these learned insights, which are continually updated.

For example, let’s say you’re are hoping to target people who are more likely to click on your ad or watch your video. Lookalike modeling uses machine learning to find more users who will take that specific action. This means your campaigns can scale to reach more people, with a high engagement rate.

How does Lookalike Modeling work?

Let’s say you are a clothing brand looking to boost online purchases for an upcoming sale. The first step would be to place a pixel – a small segment of tracking code – on your purchase confirmation page. This will allow you to track the behaviors of purchasers – during the current sale – as they move throughout the web.

The demographic and behavioral data points of anyone who completes a sale and makes it to the confirmation page can be ingested into a DMP platform and analyzed to identify which behaviors and patterns are most common among the audiences. Once those customer characteristics and data points are identified, you can use them create your new seed audience and ultimately engage with an even greater target audience.

It’s a truly incredible process that uncovers the hidden attributes that can optimize your performance for future campaigns. Icing on the cake? It’s all done in a centralized platform.

Why Use Lookalike Modeling?

Well, I guess the question would be, “Why not”? As an advertiser, when consumers within a specific audience converts, it’s a fulfilling feeling. And while you want to hold on to that high performing target audience, you still want to ensure you are growing your brand following.

By using a lookalike modeling tool, it helps you to identify a larger pool of possible customers. You can use the tool to seek audiences with behaviors that match up to your target audience, and so you have a greater chance to convert them. It’s essentially the same as building a robust profile of existing customers, only you are doing it for the audiences you have yet to reach out to and engage with.

In summary, Lookalike Modeling offers marketers and publishers a valuable approach to reaching new or current consumers in a cost-effective way, ultimately helping companies to grow the scope and reach of their businesses. Why not take advantage of your already high performing audiences and increase your brand awareness? As the industry continues to look for new and innovative ways to reach new consumers, lookalike modeling will help marketers stay true to their core and use what they already know to improve campaign success.

Perspectives

1st, 2nd, and 3rd-party data walk Into a bar…

January 30, 2017 Haley Adam, Product Marketing Manager—Lotame

Learning more about your audience is key to any successful data-driven advertising campaign, and using data, whether it be 1st, 2nd or 3rd party, is an important driver in making that happen. But not all data is created equal. When you are managing an advertising campaign, it is important to know the differences between each type of data and how you can leverage each to fit your business needs and goals.

First Party Data: The Ring Leader
Let’s begin with what is considered to be the most valuable form of data: first-party. First-party data is the data that you own and have gathered directly from your customers, and can include actions or interests demonstrated across your websites, social data, offline data, mobile data, and even TV data. First-Party data is the ring leader in your group of friends at the bar. Everyone wants to get a piece of him or her. People generally look to first-party data for guidance—it’s valuable, free, and trusted.

So why is first-party data so amazing? First of all, first-party data is yours, and it will always belong to your brand. Your company owns that data set (unless you choose to sell it, which demonstrates its value). Second, because you gathered it straight from the source, it’s accurate, and relevant; there’s no question of quality. And third, it is unique. Your first-party data is different than anyone else’s, no matter how alike you are in brand and awareness strategies. Lastly, it is less regulated than the other forms of data. No rules, no expectations. Nobody can tell you what to do with your data, because it’s yours.

So what are the limitations of first-party data? It’s finite. It is only as large or broad as your brand footprint. For example, if your site produces fitness news, your first-party data will only consists of the fitness behaviors that occur on your page. If you are looking to target more people with different interests, you may not have those other interests in your first-party data. There is no such thing as extending your first-party data, especially for free. And that will come at a cost.

Third-Party Data: The Party Animal
Third-party data is the friend who may not be the most trustworthy, but they sure know how to gain a following. They are friends with everyone, and no one can match their reach.

Third-party data refers to the customer data that is ingested from an outside (or “third”) party, at a cost. It is generated on other platforms and often aggregated from many data providers, while the original source is left unknown to the data buyer. For example, a third-party data provider will choose to pay certain publishers to let it collect information about their visitors (essentially, the publisher’s first-party data). The third-party data provider will then aggregate all those data points and organize them into a broad behavioral hierarchy, ranging from automobiles to recipes, and beyond.

Typically, third-party data falls into large, industry-standard categories based on attributes ranging from behaviors and interests to age and gender. This data is housed in dozens of marketing and ad tech platforms and can be sold to advertisers who are looking for help in targeting their advertisements. As readily accessible as it may be, third-party data is sometimes considered less valuable than other types because it is a) very readily available in many places – including to your competitors, and b) you can’t be sure where it came from.

Let’s not throw third-party data under the bus though. It is great for demographic, behavioral, and contextual targeting; and because of its scale, it plays a critical role in solutions like audience targeting and reach extension. When you don’t have enough first-party data to fulfill a campaign, many people turn to third-party data to save the day.

Second-Party Data: The Peacekeeper
This three musketeer group at the bar wouldn’t be the same, however, without second-party data. Being the middleman of the group, and at most times the peacekeeper, second-party data is all about relationships.

In essence, second-party data is another company’s first-party data they sell directly to you, rather than through a data exchange. Through this direct, transparent relationship, you are able to define exactly which segments are being bought or sold, the price of the data, and any other important terms you may require, allowing you to be fully confident in its quality.

For niche brands, second-party data can be a powerful option. For example, a 5-star hotel might want to partner with a travel website to gain location information about certain buyers. The hotel can serve specific ads to consumers who have searched flights in a specific location. In turn, if a consumer books a hotel first, the travel site can send ads for good deals on flights. Through a second-party relationship, each company involved can access a new type of audiences, providing both with scale and precision.

Second-party data is the new kid on the block, and not all publishers have embraced it. Although many believe it is higher quality than third-party data (as it’s coming directly from the source, and not available on the open market), it’s also harder to get your hands on. Often this is because a buyer doesn’t know where to go if they don’t already have personal relationships with other publishers or brands.

Rather than managing this exchange, a shared cookie environment (such as a DMP) can help foster your data and the privacy requirements that go along with it. A DMP serves as the “middleman” organization that both parties can trust to help with the syndication of their data, especially in a privacy-friendly manner.

Obviously not all data is equal. Just like the three uniquely different friends at the bar, first-, second-, and third-party data comes in various forms. Data-driven marketing campaigns wouldn’t exist without the data to run them, but thankfully there is an almost unlimited amount of it available for marketing campaigns. Truth be told, there is a place for each type of data in the digital marketing/advertising world. The key is to know the ins and outs, strengths and weaknesses of each type, and use them accordingly.

Haley Adam is fascinated with all things sitting at the intersection of technology, digital, and advertising. She is an integral part of the Lotame marketing team, participating in product marketing, social, events, and corporate marketing efforts. When she’s not talking data, Haley spends her time traveling the world or playing soccer. Haley holds a BA in Public Relations from Wilkes University.