How Gap Inc. Uses Attribution Modeling to Build Direct Connections

Understanding and connecting with an audience on a deep level is the Holy Grail of sorts for most marketers. Businesses with real insight into who their customers are can understand what truly motivates them. Customers, on the other hand, feel better about businesses (and marketing in general) when they feel they are being addressed as an individual, rather than as a group.
Personalized digital marketing created by the right attribution modeling company is one of the best methods for cultivating that kind of relationship with consumers — and one of the most interesting examples of that approach is being taken by Gap Inc., the global retailer behind The Gap, Banana Republic, Old Navy, and other brands.

How Gap Inc. is forging personal connections through advanced technology

In order to better connect with customers, Gap Inc. tapped into the power of deep data by building an advanced digital marketing stack centering around a proprietary customer data platform (CDP). By using both first and third party data, the company was able to create an in-depth and unified profile of its customers.

This approach allowed Gap Inc. to target these customers across digital marketing channels with creative and highly tailored ad content. Cutting-edge dynamic content optimization technology was the key to this operation, as it allows ads to be personalized for each customer at the moment of viewing, based on data collected by Gap Inc.

So what’s most interesting about the approach is taken by Gap Inc.? First, the company chose to invest in a proprietary CDP, giving them the freedom to build and customize a platform that best suits their needs. Next, Gap Inc. chose to integrate the first party data it already collects with information about customers interests and demographics, allowing them to create a fuller picture of who their audience is and what motivates them.

Finally, Gap Inc. is deploying sophisticated technological tools, similar to that of an attribution modeling company, supported by an in-house data science team. Content optimization technology helps deliver ads that are relevant, compelling and wrapped in the appropriate messaging context. Machine learning and AI tools, for example, can optimize the voice in which a marketing message is delivered based on the current mood of the customer.

Gap Inc. also leveraged the power of AI when building its CDP solution. Because the company has tens of millions of customer records, an algorithmic segmentation solution was necessary. The company used AI to create a unified view of its customers across channels and transactions, then deployed a separate AI solution to enable probabilistic customer matching, done in real time. This led to superior match rates and the ability to fine-tune marketing messages on the fly.

What brands can learn from Gap Inc.

Modern marketing is both an art and a science. The right creative work can move people and spur them into action, but it takes pinpoint distribution to maximize a campaign’s value. The Gap Inc.’s development of a proprietary CDP and its forward-thinking integration of AI and machine learning tools are examples of how technology can help brands connect with their audiences on a much deeper level.

At BIGEYE we’ve always tried to stay at the vanguard of new technology, while never losing sight of the creative human elements that inspire audiences. Contact us today if you’d like to hear more about personalized marketing and how the right attribution modeling company can take you to exciting new places.

Want Deeper Insights into Your Attribution Modeling? Ask a Model

Marketing used to be a lot simpler. Businesses had just a handful of channels to operate in, making assigning credit for purchases a fairly trivial task. These days, channels have proliferated and we have seen a vast increase in the number of consumer touch points. Attribution modeling is one of the best solutions for tracking sales costs and campaign success.

Attribution modeling explained

As the complexity of marketing processes has grown, simple yet critically important tasks — such as determining the source of sales and conversions — have become markedly more difficult. Customers typically don’t just go straight to a website and click the buy button. Sales are usually the result of multiple interactions across a variety of platforms prompted through a series of messages or interactions.

Attribution modeling is a framework that allows organizations to gain greater visibility into how credit for sales and conversions should be awarded or divided. There are multiple ways to model or map out credit. These model varieties serve as a set of rules for how credit is assigned to specific touch points within conversion paths.

To better illustrate how this works; let’s take a look at some basic attribution models. 

First and last interactions models

Under these rules, a single source earns credit for the sale. Using the first interaction model, the touch point that initiates contact (perhaps a downloaded white paper) would receive full credit for any sale or conversion. Under the last interaction model, the final touch point (something like a sales call) earns all the credit. While these are fairly easy to implement and track, they don’t always provide a full picture of how a customer arrived at a buying decision.

Linear, time decay and U-shaped/position-based models

Under these rules, multiple sources can receive credit for sales. The linear model attempts to give every touch point equal credit, while the time decay model gives greater weight to touch points that are later in the buying process. The U-shaped/position-based model gives greater weight to early and later touch points and reduced credit to those in the middle.

Choosing the right attribution model depends on a variety of factors, including the amount of data that is accessible and the overall complexity of the marketing campaign. Simpler campaigns with fewer moving parts may have enough visibility to make one-source attribution models a viable choice. Businesses may also create their own custom attribution models that assign weight to touch points based on their own specified criteria.

Given the challenges businesses face in the realm of digital marketing, developing a useful model for assigning credit is an essential part of generating accurate ROI assessments. Companies that make this a priority will earn a significant competitive edge when designing sophisticated multi-platform marketing campaigns.

Modeling the next step

If your business is aiming for more ROI clarity and better multi-platform campaigns, a top Florida advertising agency like BIGEYE can help make this a reality. We have the technical chops and marketing expertise to help you deploy the most accurate and effective

Omni-channel, Cross-channel, & Multi-channel Attribution, oh my!

Cracking the customer code shouldn’t feel impossible. We’re all customers, after all. Yet making marketing decisions can sometimes veer too far into Mad Men-style gut checks and intuition, or too far toward a clinical analysis of data that leaves out that emotional element. Attribution modeling — whether you call it omni-channel, cross-channel, or multi-channel — begins to solve this problem by blending data with common sense and human behavioral patterns.
Striking this balance is important in today’s fast-paced, always-on market, where consumers have limitless choices, and unfettered access to information about your brand and products. As Mark Braydon, the Content Marketing Director at Barclays UK explains, “I really like the concept that the modern day marketer needs to be part-artist and part scientist. That blend is what I look for in my team. You need to know all the things that drive short-term interest and intrigue but you also need to be able to understand what impact that is having on some of your ongoing trading and performance metrics, and that is where the deep analytical knowledge is required.” Braydon also agrees that cross-channel attribution modeling is one of the most important tools marketers have at their disposal when trying to solve for the customer journey. For us, there are three main reasons to invest in attribution marketing.

1. Attribution is the new digital currency: Traditional KPIs only reveal a piece of the puzzle as we explain in our recent blog post about how attribution unlocks your customer journey. While web analytics may reveal what your customers are doing, attribution helps you move beyond clicks and page views to a deeper, more intimate knowledge of your customers’ motivations and preferences. There are still some limitations when creating attribution models (specifically, linking multiple devices and online and offline channels), but these gaps are getting smaller every day. As more people lean in to attribution modeling as a way to support their analytics programs, these gaps will become smaller and smaller until they disappear and attribution is the new king.

2. Diving into big picture marketing: Even with gaps in attribution modeling, we already know customers are starting and finishing their shopping experience on different channels and across different devices. While this was occurring five or even ten years ago, the number of devices the average person has, and time spent on each of these devices, is increasing every year, making it even more important to understand the big picture customer journey rather than assuming that where someone starts their interaction with your brand is where you will earn their business. We predict that big picture marketing will be one of the fastest growing trends of 2017 and biggest drivers for success in 2018. And attribution is the way to understand that trend.

3. Breaking away from a one size fits all approach: Attribution modeling comes in many flavors: last touch, linear, decay, u-shaped, weighted … you get the idea. Because attribution modeling attempts to understand customer motivation, you can adapt your models to fit the actual behavior you’re seeing in real time. Attribution weighs and balances many variables, so, unlike static KPIs, you can adjust these variables to more accurately reflect what customers are doing and make decisions based on that information. It’s true that no model can perfectly capture each and every customer journey, but these models do present a more sophisticated and nuanced picture of the purchase path. You can even use a combination of attribution reports and models to get closer to a full picture of your customer experience.

You probably already have several tools in place that can inform the foundation of an omni-channel strategy. Let us take a look at your current data and we’ll help you link these to your customers’ motivations and passions, so your clients are always a name and not just a series of numbers.

Attribution marketing unlocks optimal mix for on & offline efforts

When people think about attribution marketing, they often think of it as a digital tool. Although this mindset is somewhat misleading, it makes sense because many of the earliest attribution models focused primarily on tracking digital touch points. As our understanding of attribution has become more refined however, we know that the digital world is not the only channel that plays a part in the customer journey. Understanding your attribution model can help you plan for the best marketing spend possible and show you where to invest and where to scale back.

Online and offline channels feed each other in a constant loop. We know that television ads drive consumers to their mobile devices, and that many purchases start online and conclude in-store. One cannot exist without the other. This can lead marketers to believe that they need to be everywhere at once and, as we’ve said before, that simply isn’t possible for most brands.

Attribution modeling allows you to test for the optimal marketing mix on and offline no matter what your budget may be. With a little bit of patience, you can test your way toward a successful balance between your initiatives. Finding the right mix starts with understanding your baseline success metrics. In our recent blog post about managing multiple sources of data (found here), we explain the four key pieces of data you need to begin tracking to successfully start building an attribution model: online data, CRM information, social metrics, and offline performance. Establish your data tracking methodology, generate a baseline of data, then set goals against that baseline. You’ll begin understanding the right marketing mix as you decrease and increase spend across these channels and watch all your KPIs adapt in concert. While it’s sometimes difficult to separate which channel is driving impact, attribution marketing begins to solve that question by assigning value to each channel along the conversion process.

Getting a leg up on your competition:

 Once you establish a baseline and preliminary hypothesis about how your channels are influencing each other, you can begin getting a leg up on your competition by balancing your on and offline efforts for maximum impact. The good news is that, according to Contently, only 49% of marketers are successfully serving content aligned to the customers’ purchase journey. That means, if you can unlock how much time customers are spending on and offline, your business will be ahead of the curve … way ahead. Serving the right content to your customers based on their needs increases the likelihood that they will purchase and increase their lifetime customer value.

That’s why we believe that it’s better to have four quarters than ten times when assessing great marketing campaigns. When you serve the right content at the right time, the amount of content or marketing spend is far less important than the impact it’s having. Attribution marketing reduces costs by honing in on where to spend so you are focusing on that quality rather than quantity. When seeking a balance between your online and offline efforts, focus on how each of those channels can feed each other. Nothing you do online should be disparate from your offline efforts and vice versa. By creating content that works across channels, you can repurpose your assets as your mix becomes more intelligent.

 Click here to learn more about our attribution services and begin your journey toward the perfect marketing mix today.

True or false: Myths about cross-channel attribution marketing

Over the next three years, more than 73% of brands plan to increase their marketing analytics spend in an attempt to understand the ubiquitous notion of attribution marketing, according to Forbes . Attribution, or cross-channel marketing is a relatively new trend that has taken the digital world by storm and attempts to assign value to each channel touch point your customers interact with across the purchase journey. You can learn more about what attribution marketing is and why it’s important for your business in our recent blog post, here. As our clients delve into this new discipline, we want to dispel a few common misconceptions about what attribution marketing can and cannot do so you can make the most of your analytics efforts.
Myth 1: Attribution can’t be accurately measured, so it can’t accurately forecast sales: It’s true that there are blind spots when trying to accurately measure attribution. When consumers switch between devices, are drawn to action by offline advertising efforts, or revisit digital content across a variety of channels before purchasing, it can be difficult to track their journey or assign an exact value to how those channels impacted their decision making process. But just because we can’t measure everything, doesn’t mean it isn’t valuable. Or accurate. Despite these shortcomings, attribution modeling still presents a more accurate picture than traditional analytics, which only assigns value to the first or last channel a customer touches. In this way, attribution takes us one step closer to perfection — even if we aren’t quite there yet.

Myth 2: Predictive models aren’t accurate enough: To that point: predictive modeling based on cross-channel attribution can be upwards of 80-90% accurate depending on your sample size and the type of data used to create a forecast model. While we may not have perfect models, testing can get us pretty close. Even at lower confidence intervals, attribution modeling provides directional insight into your customers’ behavioral patterns, purchase path, and communication preferences. All this information can inform other, more traditional tests with easily measured KPIs. No testing is completely accurate, or can guarantee that customers will always perform the same way given the same set of variables; but it can help marketers invest more money in the right channels at the right time to meet their customers’ needs.

Myth 3: Adaptive marketing and attribution marketing have nothing to do with each other: We’re always shocked at how infrequently cross-channel marketing is used as a tool to improve adaptive marketing. As automated testing and targeting becomes more common, adapted marketing helps businesses adjust their advertising spend on the fly as real-time data finds the perfect marketing mix. This automated fine-tuning process can be improved using attribution data to balance out your spend across channels. Adaptive marketing optimizes spend in each channel silo, while attribution marketing zooms out to optimize your entire marketing budget. Linking your data across channels is the number one challenge when using attribution to improve adaptive marketing, so work with your agency partner on the best ways to do this quickly and effectively.

Myth 4: Attribution marketing means you have to be everywhere at once: Today’s marketers feel a constant pressure to be everywhere at once. There are so many marketing tools — from print and television, to social media and PPC, to earned and event coverage — it’s hard to know how and when to spend. This can tempt even the best marketers to try to do a little bit of everything. The problem is that very few companies have the resources to be everywhere at once without breaking their budget or tanking their ROI. Attribution marketing decreases the pressure to be everywhere to everyone by helping you understand which channels are most valuable to your best clients and where your biggest areas of opportunity to invest are.

Myth 5: Attribution marketing is a digital term: Although cross-channel marketing is easiest to track between digital channels (i.e., email to social to lead capture, to close), it’s important to remember that offline influencers such as store promotions, merchandising, and traditional advertising can all influence customers to return to their devices or begin searching for your product for the first time. Instead of thinking of cross-channel marketing as a digital term, think of attribution marketing as a digital tool that transcends the physical world. We know…pretty deep, right? The offline channel is just as valuable as your online content when creating a holistic picture of your customer, so don’t fall into this classic attribution pitfall.

Myth 6: Assuming your agency is already doing this: Many clients think their agencies are already modeling attribution as part of the standard testing process. Unfortunately, this isn’t true.At BIGEYE, we are pioneering attribution data capture and testing because we believe it is one of the most powerful tools to help small- and medium-sized businesses spend their marketing dollars more intelligently. When choosing an agency partner, don’t assume they are aware of or have the tools to track this emerging trend. Contact us today about how we can help your business or how to know whether your current agency is the best fit.

What is marketing attribution & why it matters to your business

Imagine that you launched a new retargeting campaign yesterday that serves website visitors a mobile Facebook ad after they browsed items in your e-commerce platform on their laptop. Now imagine that today you emailed your newsletter database a 20% off coupon code to kick off the spring season. When a customer comes into your store to purchase that product and use their 20% off code, do you credit the website for educating them about their purchase, the Facebook ad for reminding them to buy, the email marketing campaign that served them the discount, or the store for their point of sale support? The answer to this question is important when planning where to invest your marketing dollars.
According to research, 90% of consumers switch channels and devices — such as cell phone to tablet, or online to offline — to “accomplish a conversion goal, with 67% of people using multiple devices to subsequently shop online.” Marketing attribution allows you to understand the impact each of these devices, and their corresponding channels, have had on the customer journey. And understanding this impact allows you to optimize your marketing spend around the channels and tactics that are most beneficial and cost effective. 

Single source attribution marketing:

Most marketers use what is called “single touch” attribution modeling, which gives full credit to either the first or last channel a customer “touched” before purchase. In our example, that would be either the website or the physical store venue, with no credit going to the Facebook ad or email campaign. Yet, it is unlikely that the brick and mortar location, or any other single channel really, deserves full credit for this purchase when assessing their return on investment.

Single source attribution marketing is commonly used because it is simple and easy to track. Stitching together how customers move between devices and marketing assets is difficult and sometimes limited, so we — as marketers — settle. The problem with settling is that many of us wrongfully believe that the data will normalize across channels because customers are always entering and exiting the purchase flow differently. Unfortunately, single touch attribution often leads us to believe that one marketing source is more valuable than others and doesn’t inform you about your most valuable marketing channels. This makes it difficult to accurately assign marketing spend and can lead to over or under investments in a certain channel.

Multi-source attribution modeling:

A common solution to the single touch dilemma is to systematically weight each channel in the customer journey. There are lots of ways this can be done: linear attribution, time decay, u-shaped, and full-path attribution are just a few of the models striving to systematize omni-channel marketing. Bizible provides a detailed overview of some of these models in their blog post, here. Depending on your sales cycle, each multi-source model has pros and cons. Multi source attribution marketing is certainly more accurate than single source attribution, but still has blind spots. Namely, the difficulty of tracking users across and between devices or accurately standardizing which channel has the most impact on customers when using a “one size fits all model.” If you are just dipping your toe into the attribution world, this is a great place to start. You can easily test and compare weighted attribution models to see which one is the most accurate for your business and use that as long as it suits your needs. While it may not be perfect, it will provide directional insight into where to scale your spend in a way single touch attribution cannot.

Algorithmic and fractional attribution:

The last, and most nuanced, form of attribution modeling is algorithm-based attribution. The goal here is to leverage big data from a variety of sources such as your CMS system, web analytics platform, social metrics, traditional campaign ROI, and even physical POS information when it’s available, to create a living, breathing attribution model that adapts based on how your customers are actually shopping.

Agencies in particular have the power to do this because they can partner with large data organizations that help paint a complete picture of how, when, and where customers are switching between devices. This model assigns weight intelligently based on the amount of time customers spend on each channel, the frequency at which they use a platform or device, as well as where the sale and lead starts and ends. We firmly believe that as attribution modeling becomes a more standardized part of the budgeting process, this model (and the tools that support it) will become more prevalent.

Even with the most comprehensive algorithmic attribution system, we are still learning about and adapting to the shifting nature of consumer purchase behavior. Realize that perfection is not the goal — yet. In our recent blog post debunking myths about attribution, we shed light on why this methodology is so important regardless of its inherent limitations. This new and critical level of insight into your customer journey lets you serve your clients while doing right by your business at the same time. Get in touch today and we’ll show you how.