Retail giant Amazon recently attracted attention for its plans to ship goods to customers using bird-like robotic drones.
Collective brows raised even higher when a patent revealed plans for an "anticipatory shipping" service, designed to send customers shipments of items based on the time they'd spent browsing products, their visits to a listing, previous purchase habits and other metrics.
Paul Gibson, Regional Director EMEA at marketing and analytics software company AgilOne, talks to TRPro about the technology that Amazon might use to achieve this seemingly psychic feat.
TechRadar Pro: What do you make of Amazon's recent patent announcement?
Paul Gibson: Amazon is one of the best users of predictive marketing, and utilise their vast resources and budgets to achieve a great understanding of their customers behaviour, and therefore deliver relevant, effective messaging.
This patent is yet another example of the impressive way they utilise this powerful medium, and would certainly overcome the often criticized delay between ordering and receiving goods. However, it could open the door for abuse, if people realize they can get free product just by browsing on the site.
TRP: Why is Amazon always leading the way on predictive marketing?
PG: Having vast budgets and resources has enabled Amazon to fully harness all the elements of predictive marketing, and has lead to them becoming experts in customer retention.
They have built systems, and algorithms that are able to collect data at every instance, from every touch-point, using a variety of campaigns logging your actions when you: search for something, purchase something, abandon your cart, send something back and even unsubscribe from their emails.
Amazon are then able to understand your preferences, behaviour and likely next steps and then turn all this information into personal messages tailored to your responses. This gives them a unique place in the market, as traditionally this has been out of the reach of smaller businesses.
TRP: Can other companies use predictive analytics to do this?
PG: Right now, yes, and this is really exciting! Few companies have the budget and resources of Amazon, so this capability has traditionally been out of reach for all but the largest enterprises.
However, due to the recent advancements in SaaS Predictive Marketing technologies, it is now possible for any-sized retail organisation - including those with a mix of on-line and traditional bricks and mortar operations - to harness this level of sophistication.
This helps them to gain a consolidated 360-degree view of the multiple customer touch-points and delivering unrivalled data science, which in turn leads to more effective marketing, loyal repeat buyers and ultimately improved bottom line.
TRP: Is software-as-a-service (SaaS) predictive marketing technology expensive to run?
PG: On the contrary, Cloud technology is helping to bring predictive marketing within reach of most businesses.
Companies do not need to have vast data centres, nor do they need to spend years developing new technology, and adoption of true customer insight and relevance now takes weeks, rather than months or years.
Historically, combining multiple data sources, and adding the required science was a time consuming, manually intensive and costly exercise, but SaaS predictive marketing solutions do all the heavy lifting, allowing marketers to concentrate on marketing.
TRP: How useful do you think the technology being used by Amazon will be to consumers?
PG: Of course, the actual idea behind delivering something someone has not yet purchased could confuse, and maybe even scare off a potential buyer who may simply be comparing and having already purchased the goods elsewhere.
But, in terms of receiving products which someone might need, for example if a retailer knew your printer cartridge needs replacing, this technology could prove to be very useful to consumers. It would mean you would never run out of a product which needs regular replacement.
TRP: What is the difference between SaaS predictive marketing and product recommendations?
PG: Many smaller retailers have taken the first step, which is great, by adopting product to product recommendations (such as, you purchased those red shoes, so if I give you 20% discount, you might buy this red belt). But this is really scratching the surface of what is possible.
True predictive marketing looks beyond what product someone is interested in, and understands the customer themselves, their behavior, and based on this understand things like their propensity to buy, driving more effective messaging.
For instance, if you know the person buying the red shoes has a propensity to buy additional products anyway, you don't need to offer a discount, just tell them about the product, saving 20% margin on that sale.
By combining data from every touch-point (email, web, transactions, POS, call centre, etc) you gain true customer insight and understanding and go far beyond simply reacting to what product they browse or buy.
TRP: Do you think more retailers will use predictive models to send customers freebies?
PG: Although this is very much a possibility, the suggestion that if a customer has not bought, they will get the product free of charge seems a dangerous precedent, as this may train a consumer to browse and not buy, thinking they will get a freebieâ€¦.and indeed if one doesn't then arrive, may feel the retailer doesn't appreciate them.
TRP: Do you think potential buyers will be scared off by the thought that they might have products turning up at their front door if they view a certain page?
PG: I think it could have an impact on some consumers, such as those already concerned with "big brother" watching their every move, and it would definitely make you consider where you click.
However when the technology is used at the point just before you make the decision to buy, such as recommending you buy something when you need or want it, it should be very positive both for the consumer and for the retailer.