The Marketer's Latest Tool: Age And Gender Detection
Something of enormous impact is happening in marketing right now. The era of "big data" is here, and for the first time in history marketing is becoming a mathematical proposition rather than one of art and instinct.
Something of enormous impact is happening in marketing right now. The era of “big data” is here, and for the first time in history marketing is becoming a mathematical proposition rather than one of art and instinct. Up until now, marketers have almost exclusively used the idea of branding to attract customers, then turned the customer over to salespeople. For decades, salespeople have used personality and the psychology of human behavior to strike deals and sell products.
Now they can all lighten up on the soft skills and begin relying more on numbers. That’s because we now have such powerful data capture capability on consumers that it’s totally changing the way people do business…especially online.
The newest development is facial recognition, which allows marketers to collect age and gender data on visitors and consumers. It will soon be the greatest tool in the online marketer’s toolkit. In fact, with this kind of data, we’d have to upgrade that toolkit to full-blown arsenal. For now, its best application is for brick-and-mortar businesses. We can watch and see how they use this amazing technology…and learn.
Facial recognition technology
For age and gender detection from actual photos or video, the science of facial recognition is the key. There are now artificially intelligent computers that are “learning” how to tell whether someone is young or old, male, or female from photos, much like a human would do.
This depends on lots and lots of data to feed the mega computer systems which are actually “learning” how to recognize gender and age (if you read my post on machine learning, you’ll know that deep learning requires humongous amounts of data).
The first practical use of facial recognition comes in the form of an app, How Old Do I Look?.
To drive the data input for their “How Old Do I Look?” app, Microsoft launched the app at the Microsoft Build conference a few weeks ago, where it would get a lot of play. The more people who use it, the more data the program has to work with, and the better their app becomes at guessing age and gender.
That’s good, because right now, it has a really long way to go. Also, practical applications for online marketers are still a long way off. Website visitors, unlike brick and mortar store visitors, are “invisible”, visually. But take a look below: physical stores are already using facial recognition and getting valuable data on their store visitors, most importantly: age and gender.
Someday, somehow, this will be available to online marketers too. How? For example, do you have people who come to your site via a Facebook login? Facebook is full of profile pictures. Connecting the dots yet? This makes it easy to see how this technology may soon allow you to start gathering age and gender data about your own visitors.
Retailers are already using this in Japan
Japanese electronics giant NEC has developed a ready-to-go service that uses facial recognition to help retailers analyze trends. Wondering who visits your store and when? For almost $900 per month you’ll get data on age, gender, and more. You’ll also get whether you have repeat visitors, and how they behave within your store. Now imagine this sort of data for your online store…simply wonderful.
Deep learning powerhouses compete to turn the technology into products.
Microsoft isn’t the only tech company amassing great stores of data a feeding it to computer networks so they can “learn”. Google’s doing it too, as are Baidu and Amazon. It’s just that Microsoft beat everyone else to actually producing something usable from it (that would of course be their How Old? app).
So, what is available to online marketers right now, as far as age and gender detection go? Plenty!
Google Analytics provides age and gender data
Microsoft may have beaten Google to the finish line with a practical application of facial recognition, but Google still rules the day with its Analytics product. For the past two years, web developers may have noticed that Google Analytics gives “audience reports”. This is something new, and it returns age and gender data on your website visitors. Age and gender data (AKA “demographics”) is extremely helpful for online marketers since it opens up whole new worlds of opportunity.
it allows the setup of targeted landing pages for different genders/ages
it allows you to see who’s buying your product
it allows you to see who’s visiting but not buying
We’ll leave it at that: I’m sure you can come up with several reasons why it’s good to know age and gender of your online traffic!
How is gender determined by a computer?
Facial recognition is not currently practical for online marketers because not every visitor to your site comes providing a photo (that might come someday). But there is a way for web developers to determine gender pretty accurately: and that’s by name.
When a visitor enters his or her name, there’s a way for software to tell whether it’s male or female. It’s called the DOTS Name Validation 2. Provided by Service Objects, this is a web service API that takes what you feed it and returns gender. Working from what’s essentially a giant database, the service can also tell you whether any name you supply is fraudulent.
To envision this in action, you could let’s say feed it your email list. Assuming you collect names along with email addresses, you could feed the list through the DOTS system and find out what percentage of your list is female. That information, in turn, would of course help you to target your customers more effectively. You could even split your email campaigns…one version for men and another for women. The possibilities are endless.
Gender and age data is more than handy… it’s essential
The possibilities are not only endless, they’re essential. While an online presence for any businesses is hardly considered optional any more, there are some of the finer points of online marketing (i.e. using demographics) that are still considered optional. Big mistake! Some marketers are failing to put age and gender data to work for them, even when it’s right there with tools like the DOTS Name Validation 2. It’s only a matter of time before we all come to see that as a huge oversight.
Someday, using demographic data in your marketing strategy will be non-negotiable, just as having an online presence has become today. Start collecting it, tweak your analytics software to make use of it, and you’ll be putting yourself at the forefront of online marketing.
Original Post here
Find out more on the future of Tevhnology in marketing at the DLUK Trends Briefing 24th September 2015