Actually I will not spam you and keep your personal data secure
With the launch of its first digital edition of the annual report, L'Oreal steps into a new era.
The report is an impressive tool on its own, aimed at investors, shareholders and journalists. But the real change comes with the overall shift to digital as a tool to engage consumers.
For example, the "Digital" section of the annual report states just how important naming the first Chief Digital Officer actually is. This move shows L'Oreal as an up and coming major digital player. The company will probably focus on ecommerce, data technologies as well as engaging consumers both online and offline.
An example in the digital report shows just how promising ecommerce is, especially in China:
"In China – the world’s number one online-purchasing market(1) – e-commerce already accounts for 10% of L’Oréal sales, and more than 15% for brands like VICHY, LA ROCHE-POSAY and MAGIC(2). These promising results are underpinned by partnerships with online distributors like Alibaba and Tmall. On Singles’ Day, a very important day of special offers, L’Oréal’s brands performed well, particularly MAYBELLINE NEW YORK – the number 1 make-up brand in the country(3) – and MAGIC, which sold over 11 million face masks in 24 hours"
The shift towards omnichannel marketing AND ecommerce is spectacular. L'Oreal has traditionally relied on third parties to distribute products to consumers through retail shops. Could this shift be a change in strategy with a direct-to-consumer approach or will it be an improvement in dealing with online and omnichannel retailers? Nevertheless, the move will probably ripple trough and be adopted by others.
It may be a tectonic shift in manufacturers switching from traditional models to new digital models, engaging their customers, as well as providing them with the opportunity to purchase. How will this affect traditional partners remains to be seen.
A customer by any other name would spend the same. At least that’s what Shakespeare would say if he were to try to sell his plays online. But that’s not always the case and some customers are better than others. Some customers are also harder to engage and sell to.
How do we differentiate and how can we make a good medium and long-term decision about customer acquisition costs and returns? The answer lies in using two indexes well known two marketers: Lifetime Customer Value (LCV) and Customer Acquisition Cost (CAC).
Let’s take Lifetime Customer Value (LCV). The corporate world is often focused on quarterly and in best practices yearly results. That means that returns on investment have to happen quickly or they are useless. When short-term tactics prevail, customers are often mistreated. The innovation halts, marketing and customer relationship investments drop. Customer service is lousy. On short term executives may see a rise in profits but long term results are often slashed.
A lack of focus on LCV is often trouble for ecommerce companies and companies at large. Those employing the above mentioned short term tactics miss opportunities. The client base is often unimpressed, growing slowly or even decreasing. Let’s see why:
Say we get a new potential customer by encouraging him to visit our shop. John Doe likes what he sees, registers for a newsletter, but he’s not yet convinced to buy. Later on, next month, a product on a weekly personalized offer catches his eye. He clicks, goes online, buys the product – he is now a customer. But wait. John will hopefully continue buying from us, won’t he?
He’ll keep coming back, buying something every month, say for a period of two years or so, until one day – something happens. He stops receiving his weekly personalized offer. Somewhere along the chain of command somebody decided personalized offer are too expensive. The overall operational costs dropped but so did mr. Doe’s orders. His Lifetime as a Customer is over.
In this scenario we can identify the following:
Any customer demands to be treated as a human being. That’s easy to say but when companies such as these handle millions of customers, that takes hard work to get done.
First, it takes a change in perspective. You have to understand and quantify probably the greatest asset any retailer has: the customer relationship. Ecommerce has made it easier for dissatisfied customers to jump boats. The leaders know this and they use it to their advantage.
For instance – Amazon is not making any profit when it sells a Kindle. The company supports costs so they can get more customers aboard. Those customers turn to happy customers and get to spend roughly $2400 during their lifetime as customers. So what Amazon loses in hardware sales, makes up in eBook sales and other product sales. Such a strategy is not possible without a clear understanding of Lifetime Customer Value and Customer Acquisition Cost, two of the most important indexes online retailers have to work with.
Previously we had an example of Lifetime Customer Value and how we could better understand the concept and estimate the customer’s value. Those numbers being a crude model, we have to reevaluate and get a new perspective on this value. Here it is:
We have some variables (such as customer expenditure value or purchase cycle) and constants (such as retention rate or profit margin, which are less likely to change dramatically). But don’t worry, once you get the hang of it you’ll have a great and easy way to understand wether you’re spending too much or not enough on keeping your customers happy.
Let’s start with the variables. Feel free to adapt these to your own company metrics:
These variables are defined here as weekly variables but you can change those to monthly values, if it fits your business model better. You will obtain the values above by estimating median values for all your existing customers.
When you have estimated your variables you will have to take into account some constants. They will help you predict your estimated customer lifetime value. These are:
So now we have the variables, we have the constants, let’s get busy with the equations, from simple to complex.
We will be using 1 year as a reference timeframe and we will be estimating how much will we be making in a year on any given customer. There are two main variables involved – the average customer value / week (a) and the average customer lifespan (t), expressed in years.
Limitations: this is a pretty crude estimate so it will only serve as a base for further examination. It does not take into account the retention rate and attrition (loss of customers), the discount rate, not even the profit margin. It just tells us – how much would we be expecting our customers to spend with us, during their customer lifetime.
The formula is:
So now we know roughly how much will our customers will be spending with us. But that’s not actually our money, isn’t it? That’s the revenue, not our profit. So let’s step a little further and take into account our profit margin and double check the figures, by using the Customers expenditures per visit (s) and the Purchase cycle (c) value.
Remember – this is the not the final form – we will still have to think of a future projection of our lifetime customer value. However, the second formula would be this:
This formula has it all – Gross Margin per Customer Lifespan (m), discount rate (i), retention rate (r). It is also one of the oldest and simplest ways to estimate customer value (well, as simple as it can be).
Let’s have a look at it:
You can see there that this is directly proportional with two of the values. First – Gross Margin per Customer Lifespan (how much will you profit from your customers during their lifespan as customers). Second – retention rate. So do what you can to extend your customers lifespan and the retention rate.
So now we have three formulas. Each outputs a different value. Which is the right one?
Answer: all. And none. Remember – this is an estimate. The best you can do with these three is find an average and try to stick with it. Once you have a number you now know how much should you be spending on your customers. You want an example? Head over to this info-graphic and see these formulas in action with a fine aroma of roasted coffee. Starbucks has a Lifetime Customer Value of $14099 so as long as its spending less than that to turn you into a customer and keep you one – they’re profitable.
What is LCV good for? First off – telling you which customer to keep and which not. When it comes to ecommerce data is anything but scarce. You have the info – now use it. Find out who are your best customers. Analyze your data, split customers into marketable groups and … action! Drop the marketing on unprofitable customers (that doesn’t mean you should treat them worse – just spend less on acquisition). Engage your profitable customers.
But be advised – you have to have a long enough timeframe to analyze data. Sometimes those negative LCV’s might turn out to grow in time. Use predictive analytics and extend your search to see where are your customers going, not just where they are right now.
If you enjoyed reading this post as much as I’ve enjoyed writing it go ahead and share it with those you know might want to read it. Comment on it. Like it. Anything as long as you can show me you wan’t to know more about it. Next stop – Customer Acquisition Land. How much, where and how would it be better to spend on new Customers.
Did you know that stores use smartphone WiFi and Bluetooth connections to track your movement? Turns out that’s kind of a growing trend right now. Showrooming is ever on the rise so traditional retailers need to act on understanding customers better. Tracking phones is one way to do it.
There are some companies out there (their number increasing) that provide tracking technologies. One of them is Shopper Trak and I had the pleasure of meeting one of their representatives this week. The company uses a combination of WiFi and Bluetooth signal detection to count, profile and report on customer behavior. How do they that? By registering the smartphone’s MAC address.
What are MAC addresses? Good thing you asked. These are unique identifiers for your smartphone. Kinda like your IP, except they don’t change. That’s one great feature if you’re going to track returning customers. Of course – all of these informations are anonymized and encrypted, as Bill McCarthy of Shopper Trak convincingly told me a couple when I had the pleasure of chatting with him.
Working in tech for some time now – i’m not really so sure about anonymous data but the technology is pretty interesting and its applications can work wonders for multichannel retailers.
Being a online-first type of guy, I was surprised to see the kind of tracking you get with Google Analytics in brick and mortar stores. The first question that popped into my mind was – “Can you compare store tracking data with online analytics data?”. Apparently most of the companies that provide such a service do provide a form of data export that can be used to understand online-offline behavior.
The second question was “Isn’t this thing a little intrusive?”. Probably.
Last year Nordstrom decided to find out more about its brick-and-mortar store shoppers. They thought they can get valuable intel by tracking who comes in the shop, which products customers buy more, what’s the return rate and others. You know – the kind of stuff all online shops track so they can improve customer experience and increase sales. Except they did this by tracking customer’s smartphones.
But Nordstrom did something that online stores don’t usually do – they posted a sign announcing shoppers they were being tracked. And the shoppers were not happy at all. You can see in the image on the right the kind of feedback they received.
Fearing increasing frustration with their tactics, Nordstrom discontinued the program.
Atlanta based Brickstream uses a 2d /3d type of cameras to track shoppers inside stores, reporting on queue length and customers behavior.
Brickstream uses path tracking to understand and report customer routes. It also uses height splitting in order to differentiate between different demographics (male, female, child) and 3D technologies to “see behind obstacles”.
Their video intel is, of course, pretty efficient. Used together with mobile tracking- even more so. It is also a little scary for customers inclined to privacy concerns.
Are you are one of those customers? Than you may want to scan through info on the 8 major players in this growing market, Brickstream being one of them:
In-store traffic traffic tracking is an industry lead by these 8 companies, with other minor companies quickly growing. The list is provided by “Future of Privacy”, a think tank based in Washington DC, focused on “advancing responsible data practices”.
One of the younger companies providing in-store analytics, Nomi, which recently received a $10 million funding, mentions the length they go to in order to insure customer privacy. The privacy principles they list on their website are:
So everything is cool right? Well…
So far there have certainly been some concerns regarding privacy. Retailers usually addressed them as quick as possible. And when that was not the case – customers could just turn off their WiFi and Bluetooth connection so they won’t be tracked.
As mentioned earlier the technology only works when there is some type of WiFi or Btooth connection that beacons can track. Without it – smartphones are basically invisible. But than Apple thought – hey, let’s change that.
One of the often left out features when it comes to Apple’s new iOS 7 is the iBeacon. The iBeacon is Apple’s response to NFC (near field communication). When an iOS 7 device comes within range with an iBeacon it emits a BLE (Bluetooth Low Energy) response. It becomes trackable even when the above mentioned connections are turned off.
And Apple is really committed to using it:
The technology laid dormant during the past months since it was announced. Now Apple will instal iBeacon transmitters in its stores. When walking past such a device, iOS users will be notified of additional information they can read and save on their mobile devices.
The technology will offer in-store analytics to Apple, push ads and info to customers, assist in queue lines at the genius bar and of course help with purchases and payments.
Numerous other possible uses come to mind, mostly location based enhancements… Things like door opening for the blind, customized ads, personalized offers and many others will act as an usher in a new age of technology.
This new age, however, does not leave place for privacy.
Tomorrow is all about Big Data and how best can you handle it. See, companies don’t need more data. Most medium to large companies either have the data or ways to get it easily available. The problem is – most of them don’t know how to handle it.
Here comes the boom: Predictive Analytics is *the thing* nowadays. Long gone are the days when merely registering data, processing it and acting upon the findings in the next fiscal year was enough. Right now the fastest growing companies register data, analyze it and respond to it in real time.
We all leave trails behind. Our shopping habits, our marital status, our social groups, the shows we watch and gadgets we buy – all these and much more are trails and they are in some database, somewhere. Using this data, or whatever is available at any given moment, predictive analytics software can determine our future actions through two types of programmed responses (it’s a little bit more complicated than that, but you’ll get the picture):
1. Rules Based Personalization – “If this than that”. Basic personalization. Ex.: Customers click on an ad, enter our website and we can determine they are from New York. Let’s show them our stores in New York first. They click on our product catalog, select the high-priced products. Bang! We now know they have a medium to high income. This kind of responsive personalization does not really make use of any kind of predictive analytics. It just reacts to actions. It does not try to predict them. This is a job for…
2. Predictive Personalization – this is something we, humans, can do easily. Machines, not so much. Let’s say our sports store has a sales person with a decent IQ who’s at least a little bit interested in the customers checking out the merchandise. He notices customer X has tried on at least a dozen of sports shoes in the last hour. He walks to the customer and asks him “Hey, can I interest you in this brand new snowmobile? It’s 10% off“. Oh, wait that is stupid. That just what old-time ads would do. He would actually ask the customer if he can help him find some shoes that fit and look good. That’s basically what Predictive Personalization is all about: 1. Analyzing the data real time / 2. Using context to pinpoint the best potential recommendation and 3. Personalize the output.
In case you were wondering – yes, there’s a little bit more science to it but the previous example shows what the buzzword stands for. If you are interested in the subject or you’re a future Predictive Analytics Expert you can have a look at “Personalized Recommendation on Dynamic Content Using Predictive Bilinear Models”, on how Wei Chu and Seung-Taek Park of Yahoo Labs used Predictive analytics to recommend better content on Yahoo’s front page.
You know why Facebook stalking is so easy? Because people want other people to know about their interests. The Millennials, the digital natives, generation Y – they are today’s youth and they are born and living online. They offer their info, they share their interests, they make their photos public. No more mass message. Each and everyone expects to be treated as an individual.
Companies that do not “stalk” their customers are going to be left behind: Amazon is personal, Facebook is personal, Google is personal. Most of the top online retailers are personal and they make customers’ shopping experience unique.
How about offline? Yes, 5 years ago we couldn’t have had any kind of Predictive Analytics or Personalization offline but the iPhone changed that. Now smartphones fill the gap between the data stored online and offline activities. Companies are now tracking consumer behavior through mobile activity and make use of predictive analytics to address individuals needs and wants … well .. individually.
Acting on data is not enough anymore. It’s acting on data NOW that’s important.