Because Google has transformed its free Google Product Search into the paid Google Shopping service, retailers must re-evaluate their comparison shopping engine (CSE) strategies. Retailers must ask themselves: How should the budget be redistributed? Where should the money come from? Which CSEs have the most impact? What will this all cost per click? And, most important, what is the average revenue per click (RPC)?
Harness the Big Data
Those are Big Data questions that marketers must be able to answer. Like the Oakland Athletics coach who put together a winning team on one of the league's smallest budgets by harnessing granular data, marketers have to capture the right information to support more strategic, more profitable media buys on CSEs and elsewhere.
Call it the Moneyball approach to cross-channel marketing optimization and media attribution.
In that 2004 book, Author Michael Lewis tracked Major League Baseball Coach Billy Beane as Beane eliminated the biases and assumptions that were the norm in scouting players and building a team. Instead, he demanded hard data, and acted on it. The result was a lower budget and a whole lot of wins.
Marketers could learn a lot from the Moneyball effect. As CSEs evolve and marketers struggle to determine which touchpoints influence conversion, they must find ways to capture, track, analyze, and take action based on the hundreds of thousands of marketing actions their customers take. That online and offline data delivers an understanding of what really drives conversions. The insights from such in-depth data far surpass what we can gather from aggregate data.
Shoppers encounter advertisements across channels, devices, and time. The marketers who capitalize on those interactions, and take actions based on what they know works, reduce their overall spending and register greater profits.
Let go of misconceptions
So why do so many marketers insist on measuring success based on cost per click, click-through rate, or audience size? Perhaps because they are holding on to common, expensive misconceptions about aggregate data. Surface-level data analysis seems simple and good enough, but it no longer looks that way when you consider the cost of lost opportunities and media buys based on incorrect information.
By contrast, attribution registers every user interaction (clicks, views, conversions, and other actions) in a database and then pulls insights out of that data and makes recommendations based on the insights derived from it—invaluable to the marketer who might otherwise rely on incomplete information.
For example, Facebook provides customers with the capability to see which visitors who click on Facebook ads eventually convert. However, for marketers to know whether Facebook is worth the investment, they must use a multi-attribution tool to take into consideration all of the marketing events that lead to those conversions, since most users who convert from Facebook will undoubtedly have interacted with other marketing channels, such as search, display, and television.
Ask hard questions and demand answers based on evidence
The same holds true for CSEs and any other touchpoints. Consider the television performance of one of our subscription clients. The company's media plan includes a variety of cable broadcast stations, from small outlets to large ones. The biggest stations provide the most revenue, but they also cost the most. On the other hand, the smaller stations don't provide the same volume as the larger ones, but they are often more profitable investments. Is that valuable if the conversions delivered by these smaller stations are in the single digits?
The marketer who follows an aggregate data approach might never be able to answer the question, since her process would capture the sale, but not its source. All she would learn is that, in aggregate, television either works or it doesn't. The marketer would not gain any insight into whether the larger or smaller stations were worthy of her budget. From media attribution data, however, the marketer could...
- Learn that one particular station is not only profitable but exceptionally profitable—perhaps more so than the bigger station the performance of which initially seemed better
- Decide to act on recommendations and deploy more money to smaller channels
- Multiply such insights across hundreds of smaller stations, greatly improving the return on investment
A Big Data marketing optimization system can help marketers see and act on recommendations from granular-level data, and directly improve their results.
Pay attention to every interaction, every time
We know that targets need to have numerous interactions with a company before any conversion takes place. Customers make those contacts across channels, so marketers must be able to monitor fully without restrictions based on outlets, devices, or time.
For a purchase that costs more than $200, the average number of touchpoints is greater than eight. Marketers need to know where all of those touchpoints are and which ones tip the scales most significantly toward conversion. That's the key to linking user behavior across multiple channels and devices.
Aggregate-level data cannot get the job done. As CSEs become more important and customers rely on a greater number of outlets to make their purchasing decisions, marketers need to know more about those interactions. In such an environment, media attribution is an essential tool for informed budget allocation and higher conversion rates.
