Retail Personalisation

When we think of personalisation in Retail we obviously need to think of data. Traditional data in the form of Demographic, Profile, Behavioural and Transactional; but we need to consider the new boys on the block, Contextual and Emotional.

Contextual personalisation will allow marketers to deliver messages, content and experiences based on where the customer is. This might mean based on type of location – a retail outlet, whether that be on a high street or a shopping mall or other factors based around location, for example weather. For the former digital displays could now reflect the status of the store as well as the context of location and surrounding stores, whether they be competitive or complementary. In fact, those messages could also be based on the individual and their previous purchase history (Transactional Data).

The actual weather in that location can obviously be used to entice you into buying that North Face waterproof or a warming latte.

Understanding the emotions that a consumer is exhibiting in response to products (Joy, Surprise, Anger, Confusion?) will allow brands to target communications either real time or as a follow up.

For most consumers, omni-channel experiences are now a given, but the issue marketers have is recognising a customer when they are ‘offline’ in a bricks and mortar environment. The solution needs to focus on the thing we have with us 24 hours a day. No, not mobile phones, faces!

If we want to look to the future and how this might all come together then need to look East.

Brands in China and Hong Kong are using cameras with Facial Recognition capability (yes AI) to identify individuals and then combining for example online browsing history to allow informed conversations with store staff.

Alibaba and Guess have Fashion AI that combines facial recognition into ‘magic mirrors’ in changing rooms. It allows customers to essentially see what clothes look like on without trying them on. Recommendations on other options can also be based on local weather, gender and age .

But of course aiding a sale can also be delivered by making the transaction itself easier. Alibaba has launched an experimental cashier-less store called “Tao Café’’ where customers give permission for facial recognition to essentially to be used to facilitate payments without queuing. (Amazon Go’s queueless shopping experience has people queuing to use it!)

And a little bit less East, Zara is already experimenting with automated in-store order pick-up at its new flagship store in London’s Westfield Stratford shopping mall. Customers can scan their order QR code or provide a pin number to activate an automated warehouse behind the store, which uses robots to find the package and drop it into a collection mailbox.

The key is to use data and technology to deliver a customer experience based on who they are, where they, what they need.

Value of Data

A subject being debated at the moment at the DMA (I sit on the DMA North Council) is the value of data.  

Brands have had a place on the company balance sheet for a numbers of years now. One of the most influential measures of the value of a brand comes from Interbrand who define brand valuation with three key components: The

  1. Financial performance of the branded products or services

  2. Role the brand plays in purchase decisions

  3. Brand’s competitive strength.

So why not data on the balance sheet?

Should we place a value on the size of the database, or the customer details it holds, or an organization’s capabilities in using it. In my world of CRM it is second nature to place a value on an email address or a customer’s lifetime value. Value can also be derived from the CRM data itself to generate sales as anyone who received easyJet’s anniversary email from a couple of years ago would remember or who get the Trip Advisor email informing you that you are in their top 10%. 

Of course, value can also be viewed in another way. With the recent GDPR coming into force, organizations need to understand the implications of not having their data and systems in order. Data breaches are getting bigger and bigger (Are you one of the Facebook 50 million?)

 No matter how you value it, data is the lifeblood of most organizations. 

Which then suggests we need to answer the question as to what is the value consumers should place on their data? 

We all know that there is no such thing as a free lunch, and that the so-called free services we get from Google, Amazon, Hotmail come with a hidden cost. Whether we realize or not (and we do realize it don’t we?), there is an exchange taking place. We get free search/email/shopping experiences in exchange for our data. This exchange is vital to the economies of the aforementioned brands. Its estimated that Facebook collects $240 from every American adult. This estimate was made by Wibson, a decentralized data marketplace that provides individuals a way to securely and anonymously sell validated private information in a trusted environment. 

So, what if brands changed their model and put a tangible price on your data. Not so far fetched. Shiru a Japanese coffee chain  provides free drinks to university students in exchange for personal data that they share with companies interested in hiring those student. In fact, according to Shiru’s website  ‘We have specially trained staff members who give students additional information about our sponsors while they enjoy their coffee’’

 Who would have thought that one pillar of the cashless society would be one in which the exchange of data would play a part

ePrivacy Directive: combining modern marketing and privacy

I was invited to attend a debate hosted by FEDMA on the future of ePrivacy and held at the European Parliament in Brussels on February 8th.

The event, ePrivacy Directive: combining modern marketing and privacy was hosted by Member of the European Parliament, Axel Voss.

Claire Bury, Deputy Director General of DG CNECT, at the European Commission set up the session by outlining the EU’s intention to align the ePrivacy directive with the recently adopted GDPR ( General Data Protection Regulation). ( Learn more about GDPR on the DMA’s website )

For those of you not in the know..this is Directive is more commonly known as the Cookie Directive.

In the first panel, ePrivacy: the right balance between business & buyers, Wojciech Wiewiórowski, Deputy European Data Protection Supervisor and Harald Lemke, Senior Vice President, Special Representative for e-government and e-justice at Deutsche Post DHL shared their views on what should be the focus of the ePrivacy Regulation proposal. Although everyone agreed that yes, the  confidentiality aspects of the ePrivacy Directive should be respected there was a disagreement on its impact on the growing digital and data economy in Europe. Mr Lemke, obviously representing the commercial world’s interests warned against implementing the Directive without first studying its impact on this economy. Bizarely Mr. Wiewiórowski claimed he had not seen any studies that flagged any impact on the industry (I’m not aware of any that having been commissioned)

Mr. Wiewiórowski also stressed his optimism that the Directive could be introduced in time to mirror the introduction of GDPR in May 2018. This not created a wave of sceptical smiles through the audience but also prompted Mr Lemke to flag that his legal team were so busy with GDPR that he doubted that they could then also cope with ePrivacy

Some mention was made of research carried out by the Norwegian national data protection authority, which found that consumers prefer random versus targeted advertising when given the choice. Although I’m not sure how much weight this should carry considering the maturity of the data and direct marketing in Norway.

Diana Jannsen from the Dutch DMA, presented some highlights from their recent study What consumers think about data. Interestingly one highlight is that 75% of consumers are willing to share data, but 89% of them state that business currently benefits most. The study is available to download in English here

In the second panel ,Judicael Phan, Senior Counsel at Criteo presented the different technical tools that already exist on the market to provide users with ways to express their preferences and in fact made a very strong case for how being much more transparent about the use of data can provide a business advantage

What’s the difference between a Directive and a Regulation? A regulation means that each of the member states in the EU must adhere to the exact same laws and ways of implementing them. Whereas each country in the EU can implement whatever version of a directive works best for their individual markets – usually a reflection of the maturity of that market

The annoying banner pop-ups that appear on any website we visit, asking for consent to collect cookies? That was the product of the last (and existing) EU cookie directive update. The new Directive actually drops these banners – as they are annoying – but actually then essentially insists that anyone who wants to drop cookies onto a device will need to go through more hoops to collect permission. More banners anyone?

And it’s not just websites – messaging apps also get roped into the new legislation! This Guardian article gives you a flavour

As mentioned above, the EU aim to get this place alongside GDPR in May 2018

To really feel your customers' pain, you have to stand right by their side

I’m going to make a bet with you. At some point last summer the sun properly came out (no, really), and the mercury started to rise. When it did, sales of ice cream started to rise. As it got hotter, they continued to rise… until, at about 25.3C, when they suddenly plateaud, before falling steadily as the temperature continued its hypothetical march towards the traditional “London hotter than place X!” headlines.

Why am I so sure? Big data tells me so.

Now, as Zone’s head of CRM, you might expect me to tell you that – as big data can tell you everything. However, I’m happy to admit that it can’t tell me why ice cream sales plateau at 25.3C, because it doesn’t know. You actually have to speak to people to understand the real-world dilemma that prompts this behaviour (apparently, this is the temperature at which concerns over melting outweigh a craving for refreshment).

Big data is incredible: it can help us map out customer journeys in a granular way, particularly when thinking about online experiences. But often the problem the consumer is trying to solve occurs offline, leaving a disconnect between the pain point and the big data. To bridge that disconnect, you have to talk to the customer – and that’s when you might find an opportunity.

Take washing powder. Retail sales data tells me that washing powder is bought on specific days early in the week (possibly after a weekend of multiple washes), or simply once a month when consumers do their big non food-related shops.

OK, fine – but now personalise it. If you’re like me, you realise that you have run out while standing at the washing machine (the dilemma). Standard operating procedure in the Cuzziol household is to jot it down on the shopping list ready for the next Waitrose order… and thus contribute to that standard big data analytic.

But in that pain point, there was an opportunity to bypass the shopping list. An obvious e-commerce solution might come in the form of a simple one-step order interaction with my mobile. But I don’t keep my mobile in my pocket when I’m at home. What I need is an ever-present, dedicated digital solution present at the pain point – such as Amazon’s Dash buttons.

The identification of these dilemmas – and their solutions – happens when we get closer to the customer than big data might allow. Essentially a kind of ethnography, it’s about gaining insight just by being with consumers, in their own environment, as they perform tasks. Focus groups and surveys can ask these questions, but don’t happen where the real action takes place – and are unlikely to discover that I actually don’t have my smartphone with me next to the washing machine.

We can use all the web analytics data to describe the journey a consumer takes when trying to make a purchase on an e-commerce platform, but it’s only by sitting with them and recording their struggles with the credit card section of a checkout that we get a richer understanding of what’s going on.

That’s why we place such a premium on user research, observing them on their digital journey to observe where the speed bumps are, and working on ways to smooth them out… even to the extent of mapping people’s facial expressions to the movements of a mouse on a test screen. A grimace, for example, can be an incredibly valuable data point.

Yes, big data can get us closer to customers, but we only get really close to them when we are literally close to them… and can share the pain point.

First published via the DMA

Do as I do, not as I say

Or how data based on behaviour delivers better customer experience than data based on what customers tell you

Versions of this have appeared via Zone's weekly Digital Distractions email and the The Wall blog

“If I had asked people what they wanted, they would have said faster horses’’ Henry Ford (allegedly)

History is littered with examples of how actually talking to customers about what they want has resulted in shall we say less than productive outcomes for organisations. Malcolm Gladwell’s book Blink cites numerous and entertaining examples of how consumer research lets us down.

This boils down to many factors and in fact doesn’t mean we shouldn’t listen to customers, but try to understand the context in which they are making those decisions or providing us with that information. I wonder how many restaurants collect information from customers signing up for discounts who tell them that Aberdeen, Aberaeron, Abingdon or Acton is there favourite restaurant; how many retailers see people who register for an account who have their birthday on January 1st; or market research companies have questionnaires completed by individuals who claim their annual income is one range higher than it actually is.

You could argue that all data is made equal, but some more equal than others.

In general I have a rule of thumb that Demographic data is less reliable than Profile data is less reliable than Behavioural data is less reliable than Transactional data.

Demographic data – Does my postcode really flag me as being similar to my neighbour

Profile Data – Even if accurate when I gave it to you, that was 12 months ago

Behavioural Data – Yes, I told you I was interested in Politics but the funny pages online are so damn appealing. Anecdotally I’ve heard that Film Four audiences in research groups told them they should be running French art house films and it turns out everyone just tuned into Dumb and Dumber 2

Transactional Data – Ok, so now I’ve put my money where my mouse is

But of course context is important and that can often be the most important factor that needs to be considered. Amazon were great at suggesting products similar to the cake boxes we recently bought for our daughters christening. But to be then bombarded with emails and given amazon.com recommendations all about cake boxes and related products doesn’t really work when I’d rather be getting recommendations based on the constant stream of comedy DVDs and CDs I’ve bought over the last few years and watch on Amazon Prime.

It’s confusing isn’t it. And that’s before we start talking about how customer behaviour might give us a clue to future spend. I hear that Homebase can predict that a major home project is on the cards based on your purchase of a random packet of bird seeds when in one of their stores.

All of this data can be useful, but perhaps it’s the changes in a customers behaviour or transactions that matter the most. I may be constantly engaged with your content and with your products but what does it mean when I start to show a decline in watching your TV programmes or eating your pizzas?. Is that an indication of future churn?

Conversely what happens over a period of time if I become even more engaged with your content? Is that the real time to deliver a member get member offer rather than just running it to everyone who currently has an engagement score of 8. It’s the rapid movement from a score of 5 to 8 that might offer more insight

Stringing together data and the customer journey both off and online can be a powerful tool but a journey that can be quite heroic . Don’t believe me? Check out this ad by software company Thunderhead