We think CMO’s and CFO’s are trying to solve two big problems:

1.    Lack of Historical Insight

This is the Yesterday Problem. With all the complexity of running multiple campaigns across several different media, advertisers are unable to accurately determine whether their overall mix of online advertising investments generated a sufficient return. Attribution is important because it allows marketing folks to roll up performance by categories (or as I call them, asset classes) while taking into account their influence and interdependence within the overall portfolio.

Attribution is a requirement for accurate performance measurement. If you want to know how much your banner campaign contributed to the bottom line, for example, you have to understand the role display ads played in introducing or influencing the ultimate sales. The ads might have been more, or less, valuable than your initial assumption.

Without attribution, you’re just guessing at the past performance. Guessing is okay for explaining why the Cleveland Indians haven’t won the World Series in the last hundred years. But as an historical financial tool, it’s generally frowned upon by CFO’s.

2.    Lack of Actionable Intelligence

This is the Tomorrow Problem. Historical insight is all well and good, but what marketers really care about is how it can be used to predict what they should do next.  Attribution provides the data they need to optimize the financial performance of their portfolio. It provides the inputs for budget allocation models and helps marketers decide into which advertising bucket they should drop a marginal dollar.

Attribution is a requirement for accurate valuation – in other words, figuring out what an advertising asset is really worth. Without accurate valuation, allocation analysis is pointless. And even the most sophisticated optimization and simulation tools won’t help. If you’re asking, for example, whether to allocate more spend to search or display (given current market conditions, business objectives, etc), you’ll just get a highly refined and sophisticated answer that happens to be completely wrong.

Without attribution, you’re just guessing at the future performance. Guessing is okay for picking which team other than the Cleveland Indians will win the World Series this year. But as a predictive financial tool, it’s generally frowned upon by CFO’s.

The way to solve the problems of Yesterday and Today is to build tools that accurately measure and optimize returns across a complex, cross-media advertising portfolio. Attribution Management is a critical piece of the solution.

Photograph © Jack Wilgus

I’ve read several articles and blogs recently that talk about the trend towards quantitative marketing.

A few weeks ago, for example, New York Times reporter Stephanie Clifford noted:

The shift to data-based campaigns is forcing marketers to learn new skills and drawing a new breed of worker to Madison Avenue. While most data executives now in the field came from media backgrounds, they are recruiting Wall Street math geniuses because the job requires hourly adjustments in strategy based on numbers.

This week, in a Brandweek column called “The Evolving CMO,” Todd Wasserman observed:

While crunching numbers was always part of the job, new sources of data—from emerging media like interactive TV, mobile and social media—along with a need to show ROI have CMOs awash in data and newer, often younger marketers have adapted. Sometimes derided as “spreadsheet jockeys” they are a break in tradition from the “creative savants” of old who dreamed up big ideas and multimillion-dollar TV campaigns.

Given that we started ClearSaleing in 2006 to create a Wall Street-like tool for enterprise marketers, it probably won’t surprise anyone that I agree in full, or at least in significant part, with what they’re saying.

I know I’m not exactly climbing out on a limb here, but I’m pretty sure that we’re in the early stages of a rapid transition of marketing into a mostly left-brain activity. And I want to go on record with a few can’t-miss predictions so I can eventually (and accurately) use the phrase “profoundly visionary” in my future bio.

Besides, I need to get my batting average up after being slightly off on my prediction that GM would return to profitability in 2009.

So here are 4 predictions that I think are safe and easy:

1. The numbers trend will accelerate – and won’t ever reverse. As enterprises demand more financial accountability out of the marketing department, marketing in general, and online marketing in particular, will become more numbers-driven. ClearSaleing and other marketing technology companies will respond with new, sophisticated tools and methods. Marketing conferences everywhere will become much more boring.
2. Marketers will start to speak with a financial accent. As marketing becomes more quantitative, marketers and agency executives will begin to adopt the language of finance. They’ll say things like “derivatives,” “asset class,” and “portfolio optimization” without feeling the least bit self-conscious.
3. Smart quants will develop derivative instruments to help manage risk. I don’t know what they’ll ultimately look like, but as marketers try to manage risk, derivative instruments (such as futures) will be introduced into the marketing world. Skepticism and incomprehension will quickly fade once marketers become fascinated with the cool hand signals on the trading floor.
4. Quantitative marketing will require different skill sets. The trend towards quantitative marketing will result in one less career path for English majors. Quants from Wall Street and aerospace engineers from NASA will invade marketing departments and ad agencies. Despite the initial embarrassment over their marketing titles, they’ll eventually dominate the marketing world and often refer to their non-quantitative colleagues as “standard deviations.”

Check back in about a year or two to see how I did with these. Or just read my bio in a few weeks.

You recently passed the Google test, and you’re eager to show off your new SEM expertise to a civilian. You pull up the following information for Bruce:

Based on the data, you confidently explain that at $48, “30 year mortgage”has the lowest cost per lead. The answer’s obvious, and you get that “I’m bored with you now”look on your face. “That’s terrific, Bruce says. “So how many mortgages have we closed as a result of that keyword?”

“Ummm” — you fidget like a 3rd grader giving his first oral book report. “I have no idea.”

Bruce asks you politely to pause the campaign, as he tries to remember if he has HR on speed dial.

Sound familiar to anyone?

Fast forward a few days. Let’s assume that you did a little research (between frequent trips to Monster.com) and discover that the sales information you need resides in your company’s CRM system. But you don’t know how to combine that information with your PPC data. So you close your eyes and pray to the IT gods. The IT gods sprinkle a little code dust on your system and close the loop between your PPC application and your CRM system. Now when you generate a particular lead from your Web site, you can determine whether the lead eventually turns into an offline sale by looking at the sales stage. You can also identify the source of that sale (search engine, keyword, and ad text).

So with this additional, integrated information, let’s revisit the scenario:

You know the cost per lead and now you also know the number of sales generated for each keyword. A little simple math gives you the new answer to Bruce’s question: at $2,080, “fixed mortgage”required the fewest number of leads to generate a sale and therefore had the lowest ad cost per sale. So it must be the best performing keyword.

You email Bruce with the answer. A short time later he strolls down to your cube and apologies for giving you a “needs improvement”on your performance review. But as he starts thinking about the data, he points out that while ad cost per sale is a valuable metric, what he really needs to see is the profitability associated with the keyword and sale. He just completed an expensive management seminar where he learned that unprofitable sales are bad.

You vaguely remember reading about profitability on a blog somewhere and, genius that you are, realize that once the IT gods integrated the CRM system into your PPC application, you had all the information you needed to answer the profitability question. You explain to Bruce that if you just apply the actual revenue minus costs (which you know from your CRM data), you can figure out the gross profit of the sale. Then you just have to back out the ad cost and, a la kazam, you have true profitability.

“Now,”you say to Bruce, “I can identify exactly which keywords are generating profitable sales.” And you show him the following data:

What appeared initially to be the correct answer, “30 year mortgage,”actually turns out to be the worst (-$100) when profitability is applied. “Fixed mortgage”is better at $2,020. But when you have the additional revenue and cost data, you can see that it isn’t quite as good as “mortgage.” When taking profitability into account, it’s clear that at $2,400 “mortgage”is the undisputed champ.

Bruce eyes you suspiciously, realizing that the little twit he almost fired will now be asking for a raise.

So as a result of the PPC & CRM integration, you can automatically match sales to leads and calculate profitability. Each time you close an offline sale and record the sale in your CRM system, the sale gets matched to the lead where the sale originated. With this data, you can identify which ads are responsible for your most profitable sales. Then you’ll be able to maximize the profitability of your account by focusing your ad spend only on keywords and search engines that have proven to turn a profit. You can eliminate the rest.

For those of you who can’t find “praying to the IT gods”in your company’s project management manuals, stay tuned. We’ll have more information on how to execute a flawless PPC & CRM integration project in an upcoming blog. We’ll probably call it something like “How to Execute a Flawless PPC & CRM Integration Project.”