| ||||||||||||
PRINT
COMMENT
SUBSCRIBE
RSSOn all these fronts, I agree with him (I've actually said these very things in this column over the past few years).
So when I caught up with Schmidt in the hallway after his speech, I asked two questions:
1. How credible are these uber-models likely to be if they fail to account for "non-marketing" variables like operational changes affecting customer experience, and/or the impact of ex-category activities on customers within a category (e.g. how purchase activity in one category may affect purchase interest in another)?
2. At what point do these models become so complex that they exceed the ability of most humans to understand them, leading to skepticism and doubt fueled by a deep psychological need for self-preservation?
His answers:
1. "If you can track it, we can incorporate at into the model and determine its relative importance under a variety of circumstances. If you can't, we can proxy for it with managerial judgment."
2. "That is the big challenge, isn't it?"
So, my takeaway from this interaction is: Google will likely develop a "universal platform" for market mix modeling, which in many respects could be more robust than most of the other tools on the market, especially in terms of seamless integration of online and offline elements, and Web-enabled simulation tools. While it may lack some of the subtle flexibility of a custom-designed model, this platform will likely be "close enough" in overall accuracy, given that it could be a fraction of the cost of custom, if not free. The tool will likely evolve faster to incorporate emerging dynamics and variables, as the company's scale will enable it to spot and include such things faster than any other analytics shop.
If Google has a vulnerability, it may be underestimating the human variables of the underlying questions (e.g. how much should we spend and where/how should we spend it?) and of the potential solution.
Reflecting over a glass of cabernet several hours later, I realized that Google's developments are generally good for the marketing discipline, as the company will once again push us all to accelerate our adoption of mathematical pattern recognition as inputs into managerial decisions. Besides, the new human dynamics this acceleration creates will also spur new business opportunities. So everyone wins.
68 people recommend this article.
RSS FEEDS
I don't feel like Eric's tone or diction indicated he had the perfect answer. He's merely stating that the things that CANNOT be measured, should not be measured.
I mean, marketers have an ignorant tendency to criticize Google in defense of their own ego. Instead of relying on the data, we want to say "hey, I'm a marketer!!! I know what I'm supposed to do" rather than say "hey I'll rely on Google for 50% of what's measurable and make a solid conclusion based on assumptions that are usually 50% accurate." We'd rather criticize the data and go blind. We make dumb assertions about how Google shouldn't make conclusions about the "human variable." But we live in a world where VIRTUALLY everything is measurable. Dinosaurs just don't want to admit their "experience" is shadowed by information and talent.
Not only am I a firm believer that raw talent eclipses old school talent any day of the week....
I like to think that going into a marketing campaign with 50% measurable data is BETTER than trying to create a marketing campaign where "my assumptions" are the best answers. You don't get to curb or control consumer behavior, and you don't get to dictate how the end user consumes our branding content.
So let me get this straight, if we're criticizing Eric for ATTEMPTING to measure our marketing variables, then that means we actually HAVE the answers to criticize, right?
Us "kids" laugh at old school marketing mavens trying to discredit multivariate testing without realizing that it's an absolute statistic.
Let's just all admit it. Old school marketing is a wanna be. We WANT TO BE a science where we create value out of thin air, we sell to companies, hoping they pay us for an idea WE THINK is good. And if the idea fails, it's not our fault, and we don't care cause we got paid.
Old school marketing is a universal puff piece for ignorant people like us. We want to talk about branding equity, but none of us can calculate the economics behind branding equity without Googling it. I majored in econ, and it still gets me sometimes. I mean, feel free to tell me what YOU think "marginal opportunity cost" is for consumers in your NICHE markets.
Look I'm one of you guys, but we can't judge a company that's sole mantra is "never make assumptions, let's take a look at the data."
Don't take what Eric says out of context. Being right about EVERYTHING 50% of the time, is way better than pretending to be right 100% of the time.
I'm excited by the possibilities and am actively engaging in trying to do similar analyses (on a non-mathematical-genius level to be sure) and fully believe it can be a great benefit to our company. But let's not get carried away with how revolutionary Google's efforts and learning’s will be and the undeniable accuracy of their tools. At the end of the day it is still people deciding what to show/sell/advertise to other people; and people at both ends are always unpredictable.
Now if Google can just perfect a model that could choose to buy the products being advertised for us, we could live a life of leisure, and they could dispense with the chore of trying to predict us consumers altogether.
I wonder if Eric, Larry and Sergei remember that once upon a time, leading companies like AT&T, GM/Chrysler/Ford, United Airlines, IBM and Microsoft were very much admired by most Americans. Today, not so much.
Ubiquity leads to fear and discontent.
No number of mathematicians can fix the puzzle when the pieces keep changing, the data collected is a giant sparse matrix of non-statistical data, and behavioral targeting cannot even tackle the problem emerging from social media.
Requiring management input is ok and it works in credit card fraud analysis since fraud always morphs into a new method once detected. How that comes into play in ad metrics and measurements of effectiveness is the elephant in the room. Google has the brains but this problem is not as simple as optimizing search engine traffic analysis which in retrospect is and was trivial even before Google's domination of the market.
When you have to have people involved in making the model continually adapt you introduce new unknowns that do not help solve the problem at hand.
I know what has to be done as the CTO for a behavioral targeting ad media platform. We had wonks from Yahoo, blue Lithium, Right Media and others working on it to no avail.
Good luck Eric...marketing hype will not cut it.