Hyper-personalisation: what, how and why?
Hyper-personalisation: what, how and why?
Hyper-personalisation is a proven way to achieve customer centricity, but the choice of methods to achieve it can be overwhelming.
Customer Data Platforms can feel like an easy solution, but many organisations struggle to deploy them effectively. An alternative option is using bespoke AI tools.
Which one is right for you depends on what you want and need from hyper-personalisation.
Never mind the hype
No customer-centric business or digital marketeer is immune to the allure of hyper-personalisation. The benefits it can bring are tempting, but the route to effective deployment is littered with pitfalls.
One such pitfall is the rapid proliferation of marketing technology tools that promise to achieve hyper-personalisation. The cost and benefits of these tools have to be weighed up alongside the growing limitations on access to consumer data due to increasingly stringent privacy regulations, web-browser and in-app restrictions, consumer actions, and the walled and closed ecosystems of some ad tech platforms.
This doesn’t mean that companies should give up on looking to understand their customers. The challenge resides in how to make it happen.
The true value of hyper-personalisation comes from a better understanding of where a differentiator offering becomes an advantage over competitors. This means having a clear purpose for it. Are you personalising product features? Are you choosing to create personalised marketing communications and sales offers? Are your retention offers based on individual customer needs? All of the above?
Once you’re clear about what you want to achieve with hyper-personalisation, you can plan an effective route to success.
Are CDPs the answer?
Customer Data Platforms are well known to be great tools for driving personalisation, but they are not the only viable option.
Simply put, CDPs are software tools that take data from many sources, organise it and send it somewhere to be put to use. Data sources are primarily first party and can include websites, mobile apps, and email platforms – anything from the buying cycle that helps to offer a single view of your customer’s interactions with your products, services and marketing activity.
The data stored in a CDP can help you to predict the optimal next move for individual customers and customer groups. This enables the creation of innovative use cases at scale to, for example, re-target a group of customers who are close to conversion or orchestrate a single, personalised customer journey across multiple channels or devices.
Why wouldn’t I invest in a CDP?
Choosing the right CDP isn’t straightforward and the market is growing fast: as of July 2022, there were 161 vendors in the market(1). They come in all shapes and flavours so having the right framework for determining requirements and evaluating vendors is essential, but not simple.
Once past the choice hurdle, getting the best value from your CDP investment will be affected by two things often overlooked: the quality and availability of your data and whether you have the ability to deploy and embed the solution. Getting the multiplicity of technical and business stakeholders aligned behind the idea and working together to deliver a new solution is an unenviable task.
CDPs rely on first-party data for things such as identity resolution and personalisation, and this comes with significant privacy and compliance requirements. CDP programmes must consider user consent, what the data is being used for, how long it is retained for and so forth. Robust data governance and security protocols are essential in successfully deploying a CDP.
Delivered effectively, CDPs can be massively valuable. However, there is an alternative route…
(1) Source: CDP Institute, accessed in July 22, https://cdp.com/articles/basics/customer-data-platform-market-size/
Doing data differently
Another route to hyper-personalisation is to start small and focus on select outcomes and rapid time-to-value with a targeted application or use case.
Some organisations are leveraging bespoke AI tools to hyper-personalise critical steps of the customer journey. The decision to do this starts with understanding the way your business creates value. You can then set about exploring the most strategic challenges and opportunities across the value chain and how AI and data could be leveraged in each case. This leads to a collection of AI use cases aimed at targeted interventions, allowing the business to prioritise and focus on what is most critical first, then leveraging those capabilities and learnings in the following use cases.
Even though these bespoke AI solutions can break straight into uncharted territory, their experimental nature means they can be riskier and not always as scalable as off-the-shelf solutions.
As with CDPs, AI models rely heavily on data, so the same privacy considerations apply. In this field, federated learning is emerging as an exciting privacy-preserving way of implementing smart and personalised solutions to customers.
So which one?
In reality, CDPs and AI use cases are not mutually exclusive. They tap into personalisation opportunities from slightly different – and sometimes complementary – angles. Either one can reap huge benefits when applied to the right business problem or opportunity.
On the one hand, a CDP enables experimentation and personalised interventions at scale which is difficult to replicate with bespoke AI (or Machine Learning) applications. On the other, the simplicity of solving technology problems in an agile fashion is a winning recipe, especially when off the shelf market solutions don't fit.
Methods in action 1: Deploying a CDP
We helped a leading software company reset an ambitious programme to reap $30m in incremental benefits delivering hyper-personalisation through a Customer Data Platform that combined experience management capabilities.
Methods in action 2: an AI solution
We helped an organisation with millions of customers globally deliver individual, bespoke engagement strategies for each one of them, tailoring when to engage and through which channel.
The AI tool also determines the optimal content and incentives to drive up new conversions, upsell, cross-sell and boost renewals across the customer base. This model represented a novel application of reinforcement learning into this space.