4 Steps for Successful Personalisation
Personalisation was a ubiquitous topic at last week’s Festival of Marketing held at Tobacco Dock in London, with plenty of case studies in evidence to show how companies are reaping tangible benefits from a more tailored approach to communications with prospects and customers.
Among those talking about this topic was Nick Mottershead, data scientist at analytics consultancy Lynchpin, whose presentation as part of the festival’s data and analytics track was packed with tips for organisations embarking on personalisation activities.
Here were his four recommended steps for success.
#1 Make personalisation a strategic objective
Lynchpin’s first recommendation was for the business to embrace personalisation as a strategic initiative and competitive differentiator for the business, rather than as something tactical that doesn’t sit within a data-driven and customer-centric approach to website optimisation.
Research by Econsultancy has shown that 93% of companies see improved conversion rates from website personalisation while, according to Forrester, 89% of digital businesses are investing in this area, including the likes of Netflix, Coca-Cola and Wells Fargo.
As well as directly impacting the bottom line, companies should consider the growing expectations consumers have for content and communications that are tailored to our needs.
#2 Align with business objectives
Businesses were advised to set personalisation objectives that align closely with overall business goals within the framework of the commercial business plan. For this to happen there is a requirement for clear key performance indicators (KPIs) and metrics for success, so that the assessment of how a test has performed is based on objective and agreed criteria. The advice from Lynchpin was to keep the approach ‘simple, quantifiable and measurable’.
Companies will typically use revenue generated or volume of sales as their primary success metrics, but there may also be a focus on customer experience and satisfaction metrics. Changes to your website ultimately need to enhance the user’s experience, so brands need to avoid boosting revenue in the short term to the detriment of the customer experience and subsequent damage to the business in the long term.
#3 Understand your audience
The third step to success advised by Lynchpin was the need to understand your audience, with a clear understanding of the customer journey and different needs and behaviours exhibited by different groups of customers.
An array of tools and data sources are available to companies seeking to learn about their customers, ranging from digital analytics to tools such as session replay technology that give a more qualitative understanding of why consumers might be acting in a particular way.
Although the customer journey has in some respects become more complex as touchpoints and devices have proliferated, it is possible for companies to identify key triggers and points in the path to purchase, and how they vary for different types of customer persona.
#4 Be data-driven
The fourth and final pillar for success highlighted by Lynchpin was the need to be data-driven as an organisation, something that all businesses should be striving for, even if gut instinct will continue to play a role within most companies to a greater or lesser degree.
The test-and-learn mindset needs to be embraced by the whole organisation, with c-suite executives leading by example and showing a willingness to experiment and learn from tests that don’t necessarily lead to the results that were anticipated.
A challenge for many organisations is that data is often very siloed and spread across different business units, for example CRM, transactional and web interaction data. The key to successful personalisation is unifying different sources of data, though this might be a gradual process rather than happening overnight.
The good news for companies embarking on personalisation is that that a more advanced model is not necessarily a better one. According to Mottershead, ‘personalisation should be an evolving process, not a one-time solution provided by one model’.