How Marketing Specialists Are Using MongoDB to Upsell Faster and Further
Example 1: It’s a drizzly, cold December evening and you’re shopping for your favorite bottle of Laphroaig online. You see an ad for both a full bottle of 10 Year Old Laphroaig and a tester of their Quarter Cask, then follow it to Whiskybase.com to check out the deal. You think “that’s a good deal, but I’d like to check the prices elsewhere" and don’t complete your order. Before you even head off the website, there it is, a banner offering you 40% off any standard sized bottle + 2 free testers.
Example 2: “Happy Anniversary!", she exclaims. You return the smile, barely awake – it is only 7 in the morning after all – but your thoughts are frozen in fear: you forgot to get a gift. Luckily, an email just rattled into your inbox reminding you that there’s a new coding course on Udemy and it’s currently on sale for 50% off “if you order within the next 60 minutes". “Thank God for smartwatches," you think, as you quickly order and gift her the course.
The two examples highlight two key elements of real-time marketing: speed and relevance.
(Data Never Sleeps 6.0 : Domo)
Real-time marketing aims to give consumers exactly what they want before they realize they want it. We produce 2.5 quintillion bytes of data per day, and marketers are juicing and blending as much of it as they can to target consumers with relevant campaigns and products.
Real-time marketing has its roots in the 1990s with the rise of computers and their adoption in major industries, such as retail banking, and also by consumers themselves.
With the increased rate of technology adoption, marketers and businesses alike sought out new ways to reach their demographics through websites, emails, and even banner ads (the first clickable banner ad supposedly went live in 1994).
Traditional marketing campaigns were and still are, the results of months of planning and preparation. Real-time marketing, on the other hand, is the consequence of immediate decision making, capitalizing on the nano-seconds after a potential customer shows interest in a brand, such as Belize’s tourism board tweet offering free trips to the ‘Breaking Bad’ cast.
(Arby’s timely tweet at the 2014 Grammys is another a well-known example.)
Businesses of the 1990s recognized the benefits that real-time data could offer, but since the technological infrastructure was not yet in place, it didn’t take off as a mass-market revolution.
Meanwhile, the dot com boom came and went with the turn of the new millennium: wires were laid, data centers built, and e-commerce came roaring back to life. Thanks to the improvements brought in by this era, marketers and analysts now reap the “immediate” benefits of the data.
With the ever-increasing speed and volume of data produced, what is the best way to begin structuring data so that it can do useful work and yield some value to those whose servers it passes through?
Today, businesses utilizing any type of real-time analytics will run into issues such as unstructured data (like Youtube’s 46,000+ years worth of videos), fluctuating transmission peaks (as seen during Black Friday and Cyber Monday), and combining data from various sources(such as Appliances Online and MongoDB), to name a few.
To better handle such cases, a plethora of non-relational data platforms have emerged in the last fifteen years, that are often and confusingly referred to as NoSQL databases. Just ask Craigslist who recently made the switch to MongoDB from MySQL for several of the reasons above.
Everyone knows Black Friday, capitalism’s annual courting of consumers occurring every third week of November. It’s a busy time for the team at Fresh Relevance, who were snowed in by data, with thousands of events every second for some clients and hundreds of millions for the entire Black Friday weekend. When surges like these occur, how are marketing specialists able to cope?
According to David Henderson, CTO at Fresh Relevance, the increase and flexibility in computing power have been instrumental. “In the olden days, you’d have to project for future growth and provision a rack full of servers in a data center, with the associated capital expenditure and inefficiency of lightly-used hardware. Nowadays, the process is more efficient – letting you increase or decrease computing power when you need it.”
In the back-end, individual analytics such as bounce rate, browsing and purchasing behavior, and time spent on page are tracked, and marketers can easily increase processing power as needed. This translates to email alerts, ads for similar items, or time-limited coupons on the consumer end, sometimes triggered mere seconds after adding an item to their cart.
Inversely, marketers can just as easily turn down marketing efforts a notch when there’s a drop in activity, like off-peak times or during major sporting events like the Super Bowl.
Machine learning already heavily influences social media. Predictive behavioral targeting has been paying off for the likes of Amazon and YouTube for years – and this targeting is only going to improve. Within the real-time industry, AI innovations are happening as well.
A key lever for marketers is personalization in real time, consistently across all channels both traditional and new, which means gadgets like smartwatches, smartphones, and the ever-expanding range of smart kitchen appliances are becoming new treasure troves of data.
Take for instance Freshub, the company that makes the smart kitchen a reality by letting you create lists and shop for groceries through voice commands and simple movements, all in the midst of your cooking chaos.
“We turn any kitchen into a smart kitchen and any appliance into a shopping touch-point”, says Gena Minevich, VP of Research and Development at Freshub. “This solution called for creating and maintaining a database of over 1 million grocery products gathered through a detailed, granular-level mapping of online grocery product catalogs. Overall, the capability of the database to harness this data in real-time, and give users a satisfying experience, was make-or-break for the product.”
For Freshub, MongoDB offered not just a more flexible way of structuring the data, but “advantages in scaling when handling high volumes of IoT data is required, which can be accomplished simply and without performance bottlenecks." The nature of IoT applications is to be “always on", meaning that data is continuously created and chronicled. Simply put, MongoDB provides the scalability and accessibility to data that Freshub needs.
Real-time marketing is a dot com boom byproduct that’s changing the way consumers and retailers interact today. Gone are the days of waiting for weeks, sometimes even months, to judge a campaign’s effectiveness and adjust accordingly. These days, marketers can not only see the impact of an ad within seconds, but also personalize messages to the individual consumer like never before.
Personalization is a hallmark of real-time marketing that’s only getting better with innovations in AI, machine learning, and database management systems.
As I hold out hope for a smart fridge that only stocks my favorite food, is it too far-fetched to expect fridges of the future to be another data goldmine? How will the smart home change marketing as we know it?
One thing is for certain though, never before has marketing so immediately played on our impulses – why stop at one bottle of Laphroaig? – rescued us from domestic blunders, and helped us plan the perfect trip to Belize. Or anywhere really, for that matter.