Deus Ex Machina - Machine Learning is your friend

Why? Well, in a nutshell, machine learning is the use of computers to refine patterns at a level beyond human ability. It’s kind of like if our brains were supercharged and able to store thousands of years of data, keep it all in memory, isolate the most important information, and then use all that information to predict results in real-time.

Let’s look at an example inspired by an earlier piece written by Quora:

Say you’re looking to buy the sweetest oranges. You ask your mother, who always has the ones you like, and she tells you that the biggest oranges from the local market are the sweetest.

So, following mom’s directions, you go to the local market and you buy ten of the biggest oranges you can find. Sadly, only about half of your oranges are sweet. Further investigation reveals that the oranges that happen to be the sweetest, also have bright spots.

You come to the conclusion that the sweetest oranges are the ones with the bright spots… and that your mother is a liar.

Armed with this knowledge, the next time you go to the store you only get oranges with bright spots. Once again, only half of your oranges are sweet, but this time, the differentiating quality is that the biggest are the sweetest.

You realize that the oranges that are both big and have bright spots are sweetest.

Now you can go to store and always pick the sweetest oranges…after a quick apology to your mother. 

Next time you go to the local market, however, your roommate asks your to pick up the juiciest oranges. But you don’t have any information on how to buy the juiciest oranges, so it’s back to the drawing board. It would take multiple trips back and forth from the market, trying different combinations of oranges, and maybe even stores, before you would ever find the right mix of attributes that equaled a “juicy orange”.

That means a lot of wasted time and money… unless, of course, you have a machine with a good algorithm to predict results based on a sampling of millions of different oranges across thousands of different stores. Then it would be able to use this information to find the desirable oranges (whether sweet, juicy, or sour) by generalizing and isolating all the combinations of attributes (in the case of oranges, might be big, small, smooth, lumpy, spotty, soft, hard, pale, etc.) that are of statistical importance among the infinite sets of possible variations. By paying attention to only the attributes that are known and show statistical significance towards a specific outcome, the algorithm would be able to pinpoint what matters and calculate results in a highly efficiently manner, providing results before even leaving for the store.. and without having to eat a ton of ”bad” oranges!

But this isn’t just about oranges. Let’s look at how this process may effect an advertiser:

The graph below shows the daily spend of a fashion brand that ran on our network. Follow the paths of daily spend.  Over the first few days of the campaign we’re spending more with MSIE, and less with Safari. The CPA on MSIE was high and burning most of the budget too quickly, so the system stopped buying it as much. In this case, the machine was able to anticipate the outcome by recognizing trends in the data.  Eventually, we were essentially spending nothing against MSIE.  

The best part? The client asked us to stop spending against MSIE on August 22, but looking at the graphs you can see that we were automatically decreasing spending against it.  

The machines were a step ahead, and this gave the client great confidence that our system – machine learning, works.  

A Reaction to Quentin George’s presentation at IAB Ad Ops Summit

At the IAB Ad Ops summit, Quentin George, cofounder of Unbound, gave an insightful reality check into the obstacles and challenges that lay ahead for the digital advertising industry in his talk “Perception vs. Reality: What Value Does Automated Advertising Have to Brand Marketers?”

In an ever-evolving industry, Mr. George laid out changes/solutions that are necessary in order to improve certain negative perceptions that currently surround digital advertising.You can watch the video here.

It was nice to see that several of his “solutions” coincided with a lot of LiveIntent’s current offerings and benefits:

Solution – “Reduction of Waste”: the industry needs to do better creating transparency between brands and publishers making sure there the market is efficient.

Reality: Before our technology, email was an inefficient market. Ads had to be bought and sold on send without a guarantee of anyone ever opening, let alone seeing the ad. We allow advertisers and publishers to serve ads dynamically, on-open, guaranteeing users are seeing the right ad at the right time.  

Solution – “Differentiate” and “Contextualize: Personalize your message to each user, do not have just one standard creative that fits for most of your audience while also making sure the contexts of your message fits endemically with where it is being served.

Reality: What he is essentially referring to, is A/B testing – an archaic form of optimizing creative campaigns. Through LiveIntent’s algorithmic optimization tool, L.I.P.S., we use a multitude of data points such as age, newsletter, gender, location, etc to serve ads to the users who are mostly likely to be interested and engage, optimizing to each individual impression; fostering a conversation that will delight every user, not just the majority.

Solution – “Sequence”: Make a user experience that evolves over time, tell a rich narrative do not keep showing users who have already purchased your product the same standard banner ad over and over again.

Reality: Through the use of hashed email addresses, LiveIntent enables advertisers to have a much deeper conversation with their users.Lets say an advertiser has 3 potential users: 

User 1 has no relationship with the brand.

User 2 has signed up for the brands newsletter, but not purchased their product.

User 3 has purchased the brands product.

Through the use of hashed email addresses, we can deliver three different messages to all three users that will evolve as they continue to engage with the product.  We can show User 1 the standard banner ad to get them interested and interacting with the product. Since we know User 2 has already signed up for the newsletter, we can show them a special promotion for the product. With User 3 we can suppress against, only showing new products they might be interested in without inundating them with products they have already bought.

With more consumers buying smartphones, now is the time to optimize

If brands aren’t already fine-tuning their messages for mobile devices, now is the time to begin doing so. A new report from Juniper Research suggests that more than 250 million smartphones were shipped in the third quarter of 2013, which marks a 50 percent year-over-year growth rate as well as a new quarterly record.

A variety of factors have driven this growth, from the greater availability of low-end smartphone models for people who don’t want to pay for high-end devices to the gradually improving economy that is allowing customers to spend more of their money on luxury commodities. However, all of that is irrelevant to marketers and advertisers – the important takeaway for them is that more people are using smartphones on a daily basis, so brands need ensure their messages are being tailored to these devices.

The importance of mobile optimization

Businesses are beginning to realize the importance of mobile as a channel, but few are actually devoting the proper amount of resources needed to utilize smartphones and tablets effectively. This results in content that can be displayed on mobile devices, but may not look right or take advantage of the many features unique to the channel. As many advertisers will admit, this kind of approach can often hinder the results of campaigns.

For example, say a company used email advertising to engage relevant customers. If customers opened these messages from their phones, the ad unit might not be displayed accurately on the screen – it could be entirely too small or bleed off the edge of the screen, rendering it unreadable. In essence, this brand has just purchased a wasted impression, as the customer in question isn’t likely to purchase a product or service after seeing this advertisement.

This is why optimization is so important. If that company had optimized their email advertising efforts for specific smartphones and tablets, they could have reached a high-value prospect with an incredibly relevant display. They could have even gone one step further by adding functions such as “push to call,” thereby enhancing the interactivity of their efforts by catering to the special features of smartphones.

More people may be buying smartphones, but that doesn’t mean advertisers should just throw this channel into the mix – advertisements need to be optimized to these devices to maximize success.

Optimizing ad campaigns for mobile devices to engage affluent prospects

It’s no secret that affluent consumers are the drivers of economic spending. While people with less disposable income buckle down and reserve their income for necessary budgetary expenses, affluent individuals have more freedom to spend their money as they like. As such, they make for prime targets when it comes to planning marketing and advertising campaigns.

If businesses are targeting affluent households, they may want to consider optimizing their email advertising campaigns for mobile devices. A new report from comScore and JumpTap suggests that 40 percent of tablet owners earn more than $100,000 annually. Although it comes as no surprise that people from affluent households would be able to afford the latest tech gizmos and gadgets, the percentage of tablet owners belonging to this demographic is much higher than even smartphone and computer owners.

“Smartphone and PC owners have a similar income distribution, although PC owners skew slightly more to the low-income side. That is, 41 percent of PC owners are in the [less than] $60,000 income group, compared to 33 percent of smartphone owners,” MarketingCharts added, citing the report.

When including users with an income of between $75,000 and $100,000, the number of tablet users jumps even higher, to 57 percent. The message should be clear to brands: if you’re seeking affluent customers with money to spend, your campaigns should be utilizing mobile components – particularly tablet devices.

Optimizing email initiatives for tablets

None of this information should come as a surprise to advertisers, yet many still don’t take the steps required to optimize their campaigns for mobile devices. This is particularly the case when it comes to email, with an Econsultancy survey reporting that 32 percent of respondents have “nonexistent” mobile optimization initiativesand 39 percent saying their strategy was only “basic.” While this is an improvement over 2012 (when 76 percent had “nonexistent” or “basic” mobile optimization strategies), the survey suggests advertisers are still underestimating the reach of mobile.

Mobile devices have much different screen sizes and resolutions than standard desktop computers, and breaking that down even further, there is a significant difference between tablets and smartphones as well. Not optimizing emails and advertisements to smaller screens could impact the message brands are trying to send and make it difficult to see what products are being promoted.