I’m constantly amazed at how businesses ignore the mountains of data they have, settling for measuring ROI, rather than IMPROVING it. They act like they’re scared of predictive analytics, which seek to improve your bottom line.
What is this, some game of CYA so your boss doesn’t fire you?
Sure, knowing the ROI you get lets you determine which campaigns or tactics bring in the highest return, which helps you optimize your spend — spending money on what works.
That’s great. But, it’s not enough.
What you really need are tactics that IMPROVE the ROI you get from campaigns and that’s where predictive analytics comes in. Predictive analytics tells you how to make money, not just which tactics you’re already using work best.
INC had this to say about predictive analytics:
Predictive analytics is the use of statistics, machine learning, data mining, and modeling to analyze current and historical facts to make predictions about future events. Said another way, it gives mere mortals the ability to predict the future like Nostradamus or Carnac the Magnificent (but without funny hats).
Here are some possibilities for using predictive analytics to improve your bottom line:
- Recommendation algorithms
- Manage the customer journey
- Segmentation based on CLV (customer lifetime value) or other variables
- Optimize deployment of company resources
- Hire the best employees for a job
- Detect fraud
Let’s explore a few of these options in detail.
Recommendation algorithms
Even folks who couldn’t identify an algorithm in a line-up, know the word courtesy of media attention to algorithms such as those used by Netflix and Amazon to get you to buy more of their stuff or Facebook’s algorithm determining who’s lunch selfie will show up in your Newsfeed (BTW, Twitter is introducing its own version of an algorithm to keep your Twitter Feed from working like a fan, from the breeze of all those Tweets scrolling down your feed.)
Fortune attributes much of Amazon’s 29% sales increase to its recommendation algorithm.
Algorithms are great — until they aren’t.
Ever get a recommendation from Amazon to buy something your just purchased FROM THEM? I get this all the time — hence the title of this post. You really think I need another pair of the same hiking books I JUST BOUGHT FROM YOU? Seriously.
The right hand (marketing) doesn’t seem know what the left hand (sales) has been up to. That just makes you look stupid.
Even worse, you’re wasting money on ads to get me to buy something I just bought from you because the cookie you put on me didn’t update based on my purchases.
Effectively integrating marketing and sales data to provide a better recommendation algorithm is the closest thing to printing money a business can get.
Even low tech business can get into the recommendation business. For instance, a company might use their loyalty card to track purchases then email recommendations based on an algorithm. Wouldn’t that be great — no more ads for diapers when my baby is 23 ! She’s been out of diapers for 21 years.
Managing the customer journey
Unless you’ve been hiding under a rock for the last year or so, you know the customer journey is the hottest thing in marketing today.
If you need a refresher, here’s a good intro to the customer journey.
Predictive analytics allows you to predict where a customer is along their journey so you can provide the right information at the right time to optimize that journey — and result in a sale.
For instance, maybe my cookies show I’ve been searching for boat shoes. I’m likely on your site looking for boat shoes, so show me them on your home page. If I’m not interested, I can easily search for what I am looking for. If I’ve been looking at mortgage rates, maybe I’m also interested in living room furniture, instead of boat shoes. Don’t laugh, online retailers increasingly sell a wide range of products.
You don’t have to rely on cookies, either. Google can sell you anyone interested in anything, especially if you use Chrome and log in with your Gmail account. Then Google has a treasure trove of information about you.
Now, to some this might seem a little like stalking or kinda ‘Big Brothery’, but I’d much rather be approached about something I’m interested in buying than something I’m not — like diapers.
Segmenting consumers
Segmentation is a bastion of marketing, but it was usually pretty anemic based on demographic or geographic variables. Using predictive analytics, firms can segment their customers or prospects on much more influential variables.
Take CLV. Not all customers are created equal. Some are more important because they spend more with you. For instance, Wal-Mart is a bigger customer for P&G than is the local mom and pop just as a person who buys your brand every week is a bigger customer than someone who only buys your brand occasionally.
Segmenting consumers based on CLV optimizes your marketing efforts. For instance, you’re better off spending your marketing budget reaching high CLV customers than spreading your budget across all your customers.
Allow Facebook sign-in and you now have a wealth of information about prospective customers because you know who their friends are. Now they’re not JUST like their friends, but they probably share some attitudes, beliefs, and behaviors that give you clues as to how to best market to them.
Optimize deployment of resources
Company resources are expensive and not deploying them correctly has a very high customer piss-off quotient, when customers have to wait for service.
Predictive analytics can help you optimize deployment of these resources, whether they be physical or personal. Using data, you can determine what days and times require more personnel or less. If you provide social services like police, fire, or EMS, predictive analytics can predict where services will be needed so you can stage personnel for quick response. For instance, I saw a case study where there was a correlation between sporting events and needs for emergency services that allowed management to reduce response time without hiring more providers.
Travel providers can use predictive analytics to determine when they’re likely to experience problems from weather or mechanical failure. Knowing what’s likely to happen helps businesses plan better and reduce the negative impact on customers.
Other uses for predictive analytics
Now, I’m not an expert in HR or fraud, so I won’t go in to those areas, but predictive analytics can help with those and many other aspects of managing a business.