Using Multi-Channel Attribution to Improve Marketing Effectiveness

August 17, 2017    

By Daniel Kehrer

Multi-channel marketing attribution – the science of assigning value to each marketing or brand touch point across all online and offline channels – is evolving quickly and becoming more effective all the time. But many companies – including some major brands – still use antiquated tools and ineffective methods.

Slowly, advancements in attribution tools, technology and tactics that combine digital attribution with mix modeling, optimization and the offline world are writing a new chapter in smart analytics. Five core advancements are changing the attribution landscape:

1. Market- and consumer-level integration
   Consumer-level attribution and market-level modeling – including online, offline and non-media factors – are merging. As this happens, what Forrester Research has called "adaptive marketing" is replacing old planning and measurement processes, transforming how major brands deliver on marketing goals and connect with customers. Integration of attribution and mix modeling tools and functionality is making the old campaign mindset obsolete and enabling real time optimization.
2. New capabilities to measure and predict brand impact
   Unlike older attribution approaches that can't incorporate longer-term brand advertising impacts, advanced tools can begin to factor brand impacts alongside shorter term, consumer level impacts. That's significant since failing to account for brand can throw your marketing measurement and ROI calculations off by a wide margin.
3. Improved predictive accuracy
   Model refinements have greatly improved predictive accuracy. The best models don't merely analyze historical information; they examine relationships between market factors and have the ability to look forward. They account for all paid/owned/earned influences and analyze customer behavior across a full range of both online and offline activities that ultimately lead to purchase. Most importantly, they can transform those insights into simulations that answer what-if questions.
4. Programmatic connections
   For some companies, one of the most dramatic refinements is the ability to integrate attribution insights with programmatic buying. In essence, this turns an attribution platform into the 'brains' of a programmatic effort, helping inform everything from how to allocate spending across DSPs, to which creative works best and where.
5. Improved insight adoption
   Accumulated experience with advanced attribution at scale for major global marketers has also produced learnings and refinements in how insights are applied internally to make better decisions, faster. The most successful adopters are recognizing that analytics is not a marketing-only solution.

At one large hospitality company, early attempts at digital attribution proved to be entirely inadequate to the needs of a major global marketing organization. Prior efforts had relied on notoriously inaccurate 'last touch' attribution, and marketing's impact was difficult to measure or compare across multiple tactics, channels, brands, geographies and messages.

One gaping hole was the company's inability to measure the impact of long-term branding effects and, for example, to accurately predict whether it should spend more on advertising its entire portfolio, or focus on individual hotel brands. Companies with large portfolios often struggle to understand how to achieve both brand and enterprise objectives. Planning, allocation and measurement all take place at separate times and locations, causing major inefficiencies.

Now the company is deploying more advanced marketing analytics to gain new customer insights, merge digital and offline into a single view, and deliver results that are being noticed at the highest levels of the company. They've connected several digital data, CRM, targeting and web analytics solutions to provide real-time recommendations for programmatic buying, plus planning and optimization capabilities. Cost-savings from the planning process alone were substantial.

Adding new capabilities to track brand advertising impacts is providing the company with insights into the best way to allocate marketing investments across portfolios and specific hotel brands. The key to revealing these vital interactions is the new ability to comprehensively, accurately and continuously link market- and consumer-level insights with predictive capabilities that tell them where marketing investments will yield the greatest impacts.

By employing these advances, the company has gained the ability to not only understand halo effects, but can now translate that knowledge into spend allocations and specific messages down to the individual level. One key refinement is the ability to interpret results across different time frames, as brand equity takes more time to determine than typical digital-only attribution.

Meeting New Competitive Threats with Analytics

Business-to-business technology companies are also achieving dramatic results by adopting and scaling sophisticated analytics programs. Consider a $3 billion, U.S.-based leader in mobility, virtualization, networking and cloud services. Since the company first embarked on its analytics journey in 2012, it has by its own description "completely transformed the way we work."

For several years, the company was growing rapidly at an annual rate of over 25%. But growing competitive pressures caused it to seek new ways to capture competitive advantage. Analytics provided an answer.

The company was using a patchwork of CRM systems but had little evidence of real return and only a sketchy understanding of whether its marketing investments were paying off. With annual media spend growing 5-15% and market share under pressure, the company got more serious about analytics.

"Applying B2C modeling approaches to a purely B2B business was groundbreaking at the time," says one marketing VP at the company. "The ability to optimize marketing spend across the entire mix delivered bottom-line impact, including a 5% lift in sales with no additional marketing spend in the first quarter of use."

Multi-Stage Methodology
This B2B tech brand took a multi-stage approach with four key goals:

• Examine the full marketing ecosystem – including both external (non-controllable) and internal (controllable) factors – to better understand marketing's contribution to sales.
• Understand the interplay between all touch points in the path to purchase. The model they use identifies both marketing's direct and indirect Leaving out the latter can result in underestimating impact.
• See how traditional media ignites online activity and how online and social media amplify traditional media.
• Analyze each pathway through the customer funnel and link them in a system of models.

Some findings confirmed what the company expected. But others were surprises that challenged the company to make significant and sometimes counter-intuitive changes. The company began to account for diminishing return effects and reduced customer acquisition costs by shifting some funds from above-the-line channels such as TV and radio and into digital.

Once market-level analytics tools were in place, customer-level attribution was used to reveal the interplay of individual digital investments. The company could then assess results against an independent library of response curves that serve both as a starting point for evaluating unexplored media, and a benchmark on the company's own progress.

Now they're able to evaluate spend down to individual products and geographies. Product-specific insights also clearly reveal halo effects. In addition, geographic insights have helped the company understand for the first time if and when it is overspending in the U.S., for example, and how it can optimize investments elsewhere.

And like many organizations, this company had self-imposed constraints on where it could spend its marketing budget. They'd set floors and ceilings for allocations based largely on gut feel and experience. But attribution modeling has changed that. Instead of imposing arbitrary limits, the company now plans its allocations based on insights derived from the data.

Improving Programmatic Effectiveness

Refinements in attribution are also providing new insights into 'digital body language' – the aggregate of all the digital activity you see for a given customer. With those insights, brand marketers can reap big benefits from continually optimizing their programmatic efforts. For example, brands that use numerous DSPs have been able to boost returns by 10% or more by syncing attribution tools with programmatic systems to achieve greater efficiencies.

The latest refinements in integrated tools and modeling methods allow companies to account for variables such as changing vendors or publishers, along with the effects of changes to creative and placement levels. And the software tools let them simulate the impact of changes and make mid-course corrections – on a weekly, daily or (soon) a real-time basis.

Source

"Using Multi-Channel Attribution to Improve Marketing Effectiveness." ANA Data Analytics Center, 8/17/17.