Most marketers understand the importance of performance measurement and the need for ongoing optimization of all marketing efforts. While the need for this is obvious, the correct application of the right process is often not. Not all optimization efforts are created equal in terms of their potential return or the effort necessary to implement than. Additionally, some types of optimization are contingent on others, often with a relational hierarchy between them.
To help simplify this process, we have organized different optimization activities into a Maslow-type hierarchy of needs—with the basic or fundamental activities at the bottom of the pyramid and the most difficult, higher return optimization activities at the top. Organizing our efforts in this manner, we created the following optimization pyramid:
The idea here is to make sure that you are competent in the bottom of the pyramid before moving further up. Executed correctly, the activities at the top can deliver the greatest positive impact to your bottom line, but executed poorly they can cost you a bundle. Let's look at each of the stages.
Tagging and Reporting
The ability to accurately measure performance within your digital ecosystem is a key building block for most other optimization efforts—and tagging and reporting is a critical part of this. This encompasses properties like your brand website, product microsites, online advertising, your social media footprint (Facebook, Twitter, etc.), mobile apps, and others. The correct measurement of digital activity can help determine the demand for your products or services. A poorly or inconsistently implemented brand site tagging scheme will provide faulty data that could lead to improper conclusions or insights. While it may seem mundane, it is critical that what is being measured online is truly representative of the behaviors that are occurring. If your tagging is incorrect, then your reporting is incorrect and so will the conclusions you draw. In short, if you do the so-called small things correctly, you can make a big positive impact—rather than trying to do the reverse.
The next fundamental building block is reporting—which must be consistent and accessible to all key parties. The definition of what is being measured and how it is being measured needs to be clear and accepted across the organization. An ROI measure is of little value if key constituencies, such as Finance, don't believe in it. A lack of buy-in from other departments crushes the business case for further marketing investments, especially those in social media which doesn't yet have a commonly accepted ROI metric. Furthermore, good reporting is more than just site metrics such as "unique visitors" or "bounce rates"—it is also deep insights about what the numbers mean.
Campaign Optimization & Benchmarking
With consistent tagging and reporting in place, marketers can take the next step and start to optimize elements of an overall campaign. In the digital space, this can include items like creative messaging, key word selection for search and rich media placements. Benchmarks are critical to this process.
A historical view of benchmarks can be used to set goals for each element of the campaign. For example, knowing that for similar campaigns your landing page bounce rate for display media on your brand site varies between 15% - 25% would help you focus optimization efforts. If the new campaign was delivering a bounce rate less than 15%, then your optimization efforts are probably best focused on another aspect of the campaign. Over time you can start to build benchmarks for key campaign elements and narrow the focus of your efforts to the individual aspects that are underperforming.
Working to optimize each element of the campaign will improve the performance of the campaign overall. While applying iterative improvements to each of your campaigns, based on learnings from the past, will improve your campaign performance going forward. You can think of it as a marketing annuity, continually giving back increased performance over time.
ROI Forecasting & Predictive Models
The last two fundamentals we reviewed, while being critically important, are very tactical in nature. This next section is much more strategic and has the potential to deliver significantly higher returns. Using predictive models and forecasting to simulate the results of a campaign before it is launched is a powerful approach to optimization. In this case, we aren't waiting for results to make changes; we are "war gaming" hundreds of different campaign spend levels and mixes. Typically, a forward looking ROI is the ultimate measure for these efforts. I have talked extensively about how to create predictive ROI models in previous Fast Company posts (here and here).
Armed with these simulations, you can test the efficient frontier of effective spending before any actual spending occurs. Over the last two years, I have had several clients employ this approach and determine that they were going to over spend based on the amount of demand in the marketplace, and they saved tens of millions of dollars. One critical aspect of success for optimization at this level is a uniform and well accepted measure of success within your organization. Everyone must have the same definition of success before these simulations are run and the campaign is launched. You will need to have buy-in as to how your ROI is measured upfront. Otherwise, valuable time will be wasted arguing about whether a campaign was successful or not when instead you should be spending that time optimizing its underperforming elements.
Thinking back to the items in the bottom of the pyramid, it is easy to understand why having the basics right is essential. Your ROI models won't be very helpful if they are based on faulty or inconsistently applied measurement tactics. In fact, basing optimization on this type of faulty data can be very dangerous and result in the elimination or decrease in support for a campaign that in reality is driving good results. As industry leading innovation guru, Dr. Jeff DeGraff once said, "if you optimize when you don't know where you are going, you may be accelerating your way to ruin."
Marketing Enterprise Optimization
If you have one less dollar to spend where should it come from and, conversely, if you have one more dollar to spend where should it go—that's always the question, as ultimately much of marketing resource allocation is about tradeoffs. Once you have a holistic view of performance across campaigns you can make informed decisions about overall resource allocation for the entire Marketing enterprise. Consistent and truly comparable cross-campaign ROIs measures are powerful tools. This Meta view of your performance helps you understand what campaigns, products and marketing channels are driving the best performance.
Cross campaign and product ROIs are your Rosetta stone to understand how they stack up against each other. I have seen these types of measures lead to great innovation in terms of what gets funded and what doesn't within the marketing budget. This type of analysis is especially helpful in getting organizations to rethink continued support for those poor performing "sacred cows" in the marketing budget that no one has the emotional wherewithal to take on.
Another key factor in success at this level is the esprit de corps of your marketing and agency teams. In the wrong hands, these types of performance metrics can be used as a wedge to drive teams apart and create animosity or warring factions. Not everyone's campaigns will always perform well; the numbers won't always be good. Dealing with poor performance is just as important as dealing with success. It isn't personal, nor is it a judgment on someone's character. Strong leadership is the cure to make sure these metrics are used properly.
Finally at the top of the pyramid, we have optimization efforts that cut across functional groups such as marketing, product development, finance and purchasing. This is about a common language and measures of success that go beyond how the marketing department allocates its annual budget. At this level of optimization, the marketing department is working with the finance department, the R&D group, operations, etc. in an effort to understand the impact of different resource allocation decisions. This is a conversation typically lead by the CEO, where all levers are in play.
As an example, new product development is too often divorced from the costs of marketing that product. With a common set of performance metrics, based on if we spend X we will get Y, the organization can evaluate the costs of both developing and marketing the product in a common currency. Better decisions can be made about which products to invest in and how much support they should receive. Just as resourcing tradeoffs are made within the marketing department they are also made across the business as a whole.
This is about a process, a common language and set of success criteria that can bridge the gap of Six Sigma conversations from manufacturing and TRP's from media. It also creates an environment of shared responsibility and accountability. Just to be clear this language of performance is about more than sales or bottom line-revenue; it is about having performance metrics that are leading indicators of these measures and bring to light potential issues in the long casual chain of marketing and business activities. And this is not fantasy, companies such P&G, Southwest Airlines and Nissan are good examples of companies that operate in part or all of this paradigm.
Walk Before You Run
The potential cost savings and increases in performance for optimization within marketing and your business overall can be a competitive advantage. But remember, bad data leads to bad decisions. So it is paramount that you ensure that you have effectively built the bottom of the pyramid first. Remember to walk before you run and build on small optimization successes before tackling larger ones.
As you head into the holidays and the close of 2010, think about how much better your organization would have performed this year if you were one or two steps further up the pyramid. Sit down with your department and agency partners and ask a few questions:
1. Where are you on the Optimization pyramid?
2. Where do you want to be this time next year?
3. What actions do you need to take to more up the pyramid?
If you are feeling particularly daring, you can start by simply asking, "What didn't work?"
Steve Kerho is the SVP, Analytics, Marketing Optimization at Organic (www.organic.com).