Marketers cannot rely on fate. A customer makes 9.5 visits to a brand website on average before buying, according to Rakuten Marketing. In between they will research, chat to friends and undoubtedly check the competition.
This makes for a complex, multichannel, multi-device customer journey and that means brands need to use data to ensure their marketing is effective in a number of contexts, including the traditionally opposing objectives of brand awareness and direct response.
As consumers’ purchase journeys get longer, the chance that they will find a better match elsewhere increases, but so too does the amount of data at a brand’s disposal. From that first interaction – a click on a website, advert or social media post – marketers need to use this data to ensure that people come back.
“Retailers frequently say they have ‘no data’ on their customers, particularly for first time visitors at the top of the funnel, but this isn’t true,” says Jenny Parker, co-founder and marketing director at Country Attire. “First-time visitors to a website come with a whole trail of useful data to help the brand produce relevant content.”
Balancing brand and performance marketing
A difference when using data to tailor marketing at the early stages of the journey, compared to conversion-driven efforts nearer to the purchase, is that bombarding potential customers with sales messages from day one is a sure-fire way to turn people off.
However, taking things slowly is not always favoured by company boards. “They want faster movement [down the funnel],” explains Steve Mark, brand director at vehicle leasing firm Vanarama. “They don’t want to hear it will take five or six months to warm people up.”
Getting the balance right is not easy. Frequency, type of interaction and the level of interaction are all key considerations – as is the content. As an example of low-pressure marketing to boost engagement and brand objectives, as opposed to new subscriptions or renewals, dating site eHarmony turns data into ‘tips’ to help users be more successful on dates. “That’s the type of content consumers are looking for,” says UK marketing director Romain Bertrand.
Country Attire’s Parker agrees: “Nearly all areas of a website can be improved by personalising content to consumer preferences, removing irrelevant content and promoting content that fits the consumer’s needs.”
This subtle personalisation of the offer, which can be as simple as replacing the word ‘trousers’ with ‘pants’ for shoppers browsing in the US, can encourage engagement. And the more engagement there is, the more data the consumer reveals, which can be used to refine the statistical models and conversion-driven messages later in the journey. Consultancy OC&C found that consumers are 26% more likely to respond to messages when they are tailored to personal interests; and 22% more likely to engage when specific to their location.
Northern Rail, who worked with Amaze One, recently generated a 131% uplift in sales volume, a 112% increase in revenue and a 21% rise in conversion to sales as a result of a better understanding of when, why and where customers interact with the brand. Phil Gates, head of marketing and sales, explains how it was achieved.
“We identified three core groups which were critical to develop a communication strategy and foster a more rewarding relationship with families, students and commuters. To further streamline our strategy, each category was split into high- and low-value consumers, enabling a greater level of personalisation,” he says.
“Adding location data into the mix meant that we could [understand] why these groups of customers were travelling by train. That insight proved invaluable for maintaining a strong brand presence with our customers, so that the next time they had plans to travel, they chose our network.”
The company also used the data to select regional events to communicate “timely and relevant promotions”.
For Parker at Country Attire, who works with Monetate, personalisation means increasing the relevance of an experience to customers. Good personalisation should be “almost invisible”, she says. “You don’t want customers thinking ‘this has been personalised for me’ but you do want them to see only relevant products in their preferred way,” she adds.
But how soon can personalisation move on to the next level, seeking direct response instead of engagement? Matt Curry, head of ecommerce at Lovehoney, which works with optimisation agency PRWD, says: “Once you know interest, and reservations, you should start looking at personalising product and messaging. Before that, you should be convincing them that you are the right place to shop with, without trying to sell them specific products. You can obviously try to personalise with very little data [but] experience has shown me it’s not worth it.”
As customers move down the funnel, their commitment levels increase, as does their propensity to share more information. Personalisation can then be modified accordingly.
A survey of CMOs last year by Deloitte, Salesforce and ExactTarget found that ‘personalised experience’ is among the top three marketing priorities, right alongside customer acquisition. Data is central to this. “We must move from numbers keeping score to numbers that drive better actions,” explained David Walmsley, director of M&S.com, in the report.
Customers also want to know that data is used responsibly. In 2007, the scope of data held on Tesco Clubcard was detailed in the book Tescopoly. It said: “Tesco probably ends up knowing more about a cardholder’s comings and goings than the holder’s husband or wife.” At the time there were 10 million active cards in use; today there at 16 million.
But that is just one set of data. Brands today have access to structured data from transactions as well as the less structured data attached to videos, emails, photos and social media updates. This has enabled brands to move on from personalisation to contextualisation – rather than showing a customer the products for sale it shows shoppers what they want to see; the messaging also happens in real time rather than planned in advance.
OC&C’s research suggests that consumers are 21% more likely to respond to contextualised messaging but brands must exercise caution. A much-publicised example involves retailer Target, which analysed data to calculate a ‘pregnancy prediction’ for shoppers. Mother care has used data in a similar way to predict what its customers need on their journey through parenthood before they have realised themselves.
Second-guessing customers is never going to be an exact science though, not least because it requires less structured data, such as that gleaned from social media. The technology is improving all the time.
Earlier this year, researchers at the Cambridge University created an algorithm that can use Facebook ‘likes’ to judge personality traits more accurately than friends or even family members. Feed in enough data (300 ‘likes’) and only a spouse can rival the computer’s understanding, they claimed. “The ability to judge personality is an essential component of social living – from day-to-day decisions to long-term plans such as whom to marry, trust, hire or elect as president,” said researcher David Stillwell at Cambridge University.
Dating sites have long been sparring about how accurate and successful their algorithms can be. But matchmaking is only the half of it. Bertrand at eHarmony says if a customer has offered enough information and they are far enough down the funnel, “data can be used to determine when they’re likely to leave us”. This means CRM can be ramped up, with data science, messaging and marketing teams working the data together to determine what promotions or discounts might trigger the customer to recommit rather than churn.
Bertrand admits that “areas of darkness” remain. “There are offline factors we don’t know; there are areas of darkness [in the data] that no computer will be able to tell us [but] you don’t want to freak people out.”
At Direct Line, when devising our recent Guaranteed Hire Car campaign, we used data to drive relevance and engagement by informing the messaging and context of our advertising.
Ahead of the campaign activity, we set up a number of social listening projects that monitored all online conversations around hire and courtesy cars (not just insurance-related). We used the data gathered from this social listening to draw out a number of trends.
Two trends shone through. First, there was more conversation on forums rather than on big social media platforms. We used this insight to change our advertising plan to include sponsored activity with Mumsnet, one of the biggest forums in the UK.
Second, we found that going on holiday was a big topic surrounding conversations about hire and courtesy cars. We used this insight to inform our social content development and created a short form video asking the question ‘How would you go on holiday if your car was written off?’.
Marketers have traditionally looked at the customer journey as a funnel: they start at the top and are fed down through a number of stages, including awareness, consideration and purchase. But that model has been transformed by the way we use technology, impacting consumer habits. The journey is now multichannel, multi-device and complicated. But marketers shouldn’t fear this.
Today’s customers leave a trail of data everywhere they go. This information shows us what they want and need. So rather than shout from the rooftops at the many, brands can choose to use data to talk to the specific few that match their target profiles. You know whom to target and with what, as well as how they would like to be engaged and when. What is more, you’ll know if you feed them a message that doesn’t resonate – and quickly.
But bringing all this data together for meaningful insight can be intimidating. Indeed, we help a lot of companies collecting a lot of data to decide what is relevant based on their marketing objectives, as many are not sure what to do with it or what it shows them.
Take the vehicle manufacturer we worked with recently. We analysed its CRM data, together with that from a third party, to delve a bit deeper into who its loyal customers really are. We discovered some things we probably already knew: those looking at the sports utility vehicles had families, while those opting for saloon models tended to be younger with more disposable income.
However, we also found out that the saloon drivers over-indexed three-to-one on Android smartphones versus iPhones – and yet the existing advertising was focused on the iPhone compatibility of the cars. With this insight, the brand was able to quickly rework its creative messages to the market and get better bang for its marketing buck.
For a luxury retail brand, we found that their best customers were Apple advocates – those with Macs and iPhones were, respectively, six and seven times times more likely to convert. Using this insight during the campaign helped increase sales by 171%.
With the amount of data available, marketers don’t have to shoot in the dark. With data you can personalise and perfect messages at any stage of the customer journey, across both branding and performance objectives.