Since 2019, when Gartner first predicted that 80% of marketers would abandon personalisation efforts by 2025, there has been a significant divergence in views between those who doubt personalisation can ever deliver benefits and those who see it as critical to delivering positive customer and value outcomes. So where does the truth lie?
Personalisation can have myriad meanings, from a name printed on a mug to a retargeting offer. In data science, as it’s applied in marketing, personalisation is the use of first-party customer data to deliver tailored offers and experiences, online and offline. By focusing on marketing communications and purchase journeys for existing customers at an individual level, an enhanced customer experience evolves and delivers long term value.
Using personalisation therefore seems like a no-brainer strategy. Why would brands not want to understand how and when to reach customers based on their personal preferences, and engage with them based on their interests and what they are most likely to purchase? Here, Gartner identifies two main obstacles – customer data management and a lack of ROI – both of which imply it’s not personalisation or its effectiveness as a strategy per se that’s problematic; the root issue is implementation.
Good quality – and, critically, complete – customer data, is key to delivering effective personalisation
Data about a customer can detail their every brand touchpoint, offering insight into a customer’s holistic view of a brand. When pockets of data are missed, it’s a game changer that fundamentally undermines a personalised marketing initiative.
Thankfully, data integration projects no longer need months of resource and development. For example, cloud-based data warehouse technologies remove up-front hardware spend and provide rapid scalability of capacity with proportionate costs. Furthermore, a properly thought-out data strategy, with unique customer identifiers used across all online and back-office systems, leaves no hiding place for data silos that can’t be joined up and attached to a customer record.
Technology is an enabler – not the solution
Delivering traceable ROI is often seen as being a technology challenge, where simply buying the right software licence will generate significant ROI. But technology is an enabler and nothing more. Personalisation requires data to be correctly interpreted.
Customers are invariably diverse and do not neatly conform to segmented groups, but they do have individual habits: some customers may buy immediately after landing on a website, while others habitually shop around and only buy on the second or third visit, perhaps over the course of a few days – and there are many possible scenarios in between. This level of complexity is simplified through data models that predict the likelihood of certain things happening by considering holistic intelligence. Data modelling is helping marketers to understand the likelihood of a customer performing a desired behaviour from a plethora of options, how and when they will do it, and what nudges are needed to get them to behave in that way.
In order to personalise content for customers, you don’t need to know everything about them.
Suddenly, the world of personalised offers starts to open up. You can send a communication when a customer is most likely to open it, or when the customer hasn’t visited the site for longer than usual. You can send it via the channel the customer is most responsive to. You can include the product they usually buy at that time of year, or the product they’ve never bought but are most likely to buy next. You can even understand how likely the purchase is and use an incentive when that likelihood drops away, thus preserving full value and not discounting products for customers who will buy anyway, while still increasing conversion where it matters most.
This also raises the question of how much content can be personalised? Ideally, a digital asset management system can be used to unlock existing brand content; failing that, there are various ways to focus on the most popular (or profitable) products. The ROI for this approach is only generated when all these things come together, so an important part of a personalisation strategy is factoring in the technology and budget for creating the models and the marketing automations that actually distribute the solution.
In our experience, shifting spend from the technology on its own into the models and execution has had a profound impact on the ROI, with many failed personalisation projects resulting from too much investment in over-specified, expensive technology, leaving scant resources available for the modelling and activity itself.
What do customers prefer anyway?
Another hurdle personalisation must often overcome is the privacy challenge. The media is awash with stories about the public’s privacy concerns and dislike of being spied upon. This can be a bit of a red herring.
In order to personalise content for customers, you don’t need to know everything about them – it’s very much a domain-specific challenge. If you are selling groceries, you need to know which groceries a customer prefers and when they choose to go shopping; customer behaviour demonstrates that they will happily volunteer this information to you and will usually expect you to know it if they have shopped with you before.
Being offered the things you are most likely to buy isn’t actually an invasion of privacy at all, it’s just using the information available to make the experience really convenient for the customer. Focusing on what is in the customer’s interests in this way really does remove privacy concerns, because it’s about acting on what the customer expects you to know anyway from their previous interactions with you as a brand. This is borne out in research, with customers persistently indicating that they prefer personalised journeys and content in a wide range of different scenarios.
So, to wrap up:
- Personalisation is about understanding customers and using that insight to provide them with a better experience but to achieve this requires high-quality, relevant and complete data at customer level.
- It’s using data, insight and expertise – enabled by technology – to provide customers with better and more mutually rewarding experiences. It is not a pure-play technology solution.
- Viewing personalisation strategies as a long-term commitment for unlocking value, instead of chasing short-term purchase objectives, is crucial. That means investing in a robust and agile data strategy that benefits both brand and customer.
In conclusion, if you are going to use personalisation, do it properly or don’t do it at all.
Join Jaywing’s session at the Festival of Marketing on 6 October, where you can learn how to elevate your customer data for greater commercial success.
Malcolm Clifford is customer experience director at Jaywing.