Digital Marketing in an AI World: Futureproofing Your PPC Agency by Frederick Vallaeys
Author:Frederick Vallaeys [Vallaeys, Frederick]
Language: eng
Format: epub
Publisher: Modern Marketing Masters
Published: 2019-05-27T18:30:00+00:00
Dr. House, Forensic Pathologist
In Optmyzr, we have a tool called the PPC Investigator that allows an advertiser to ask general questions about why their business outcomes have changed from one period to the next. A typical question might be, “Why do I have fewer conversions this quarter than I did in the same quarter last year?”
The tool does a root-cause analysis and comes up with one or more diagnoses. Perhaps you got fewer clicks because fewer people are searching for what you are advertising. You may have lowered the amounts you’re willing to bid and have gotten fewer impressions as a result, while your competitors got more of the impressions you would otherwise have received. Your conversion rate might also have gone down because you made a change to your landing pages that makes them less compelling than before.
The client then logically asks, “You just pointed out three significant problems. I’ve only got a limited amount of time and money at the moment. What’s the one thing I should do that would have the greatest effect?”
In one case, the client was a startup that had been able to make changes to its landing page very quickly and on the fly. However, the account manager also knew the company had recently been acquired by a large corporation where policies were much more rigid. Unless that information had been input specifically, it’s unlikely a machine learning system would be able to figure out that the business had been recently acquired by a new owner whose priorities were different from the former management team.
If the question had been asked a few months earlier, the answer might have been: let’s make some changes to the landing pages. However, the account manager knew that the new owner had a very rigorous process for changing company web pages. In fact, the corporation had a complicated, fifty-page brand-guideline document that covered the specifications for every change down to the pixel, including image content and color-coding.
Given current restrictions, the account manager knew better than to revise the landing pages. The suggestion then became, “Let’s change the keyword mix, because that’s what we can do right now that will have an impact and start moving things in the right direction.” Down the line, other changes might still be made. At the moment, it was critical to prioritize action based on feasibility. This is the kind of flexible approach that account managers, who speak with their clients frequently and take a real interest in what is happening to their businesses, are able to take.
Another important aspect of “bedside manner” or “hand-holding” is the ability to explain the rationale behind a recommendation. Once machine learning is applied to a problem, you can see if it was resolved or not, but the system is generally very bad at explaining exactly how the solution was arrived at. If the client asks, “How do I apply this solution to another business where I’m encountering a similar problem?” the system can’t respond intelligibly.
The classic example here is asking a machine learning system to look at a photo and tell you if it is a cat.
Download
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.
Computer Vision & Pattern Recognition | Expert Systems |
Intelligence & Semantics | Machine Theory |
Natural Language Processing | Neural Networks |
Algorithms of the Intelligent Web by Haralambos Marmanis;Dmitry Babenko(7835)
Hadoop in Practice by Alex Holmes(5650)
Jquery UI in Action : Master the concepts Of Jquery UI: A Step By Step Approach by ANMOL GOYAL(5495)
Life 3.0: Being Human in the Age of Artificial Intelligence by Tegmark Max(4473)
Functional Programming in JavaScript by Mantyla Dan(3712)
The Age of Surveillance Capitalism by Shoshana Zuboff(3395)
Blockchain Basics by Daniel Drescher(2868)
Big Data Analysis with Python by Ivan Marin(2830)
The Rosie Effect by Graeme Simsion(2689)
WordPress Plugin Development Cookbook by Yannick Lefebvre(2523)
Applied Predictive Modeling by Max Kuhn & Kjell Johnson(2460)
Hands-On Machine Learning for Algorithmic Trading by Stefan Jansen(2443)
Dawn of the New Everything by Jaron Lanier(2423)
The Art Of Deception by Kevin Mitnick(2278)
Rapid Viz: A New Method for the Rapid Visualization of Ideas by Kurt Hanks & Larry Belliston(2174)
Human Dynamics Research in Smart and Connected Communities by Shih-Lung Shaw & Daniel Sui(2167)
Once Upon an Algorithm by Martin Erwig(2135)
Building Machine Learning Systems with Python by Richert Willi Coelho Luis Pedro(2047)
Test-Driven Development with Java by Alan Mellor(2038)