5 pitfalls of personalisation and how to avoid them

02 September 2016
Sharukh Naqvi, Vice President, Analytics, Barclays
Frink_Cognifide_2016_HeaderImages_0117

A recent Digital Trends survey revealed that 73% of consumers prefer to do business with brands that use personal information to make their shopping experiences more relevant.  So it’s no surprise that personalisation is the hot topic in marketing in 2016.  Nowhere more so than at Barclays where we have clearly defined our mission when it comes to personalisation.  We want to make sure that we anticipate the needs of the customer before they know themselves.  After all, the only reason to personalise is to provide value to the customer and doing it well means having an informed proposition around each customer segment.  


Personalisation - Sharukh.JPG


At Cognifide’s Unconference earlier this year, I had the opportunity to lead a discussion around the issue.  It was great to be able to share experiences - success stories but also common pain points - with peers from so many different organisations.  It’s always good to know that you’re not alone on what can sometimes be a difficult journey,  so I thought I’d share some of the common themes of the discussion.  As the conversation often veered towards lessons learned through trial and error, these 5 pitfalls seemed to strike a chord with most.

Don’t let technology dictate strategy

Technology tends to be a key driver - largely because of the financial commitment that it represents.  Critical enablers such as a business intelligence infrastructure, data feeds, analytical resources and a multi-channel campaign execution platform are all essential tools.   However, make no mistake that strategy should come first and don’t imagine that a comprehensive personalisation strategy will jump out of your overall digital strategy - it won’t!  In defining your personalisation strategy make sure that you are answering the following questions:

  • What is the business objective and is it achievable?

  • Have you identified and segmented your customer base.  Who are you targeting and what actions to do you want to trigger?

  • Have you defined the actual steps and put it into a time frame?  Who is responsible?  

  • Have you defined the technology, processes and people that are needed?

  • Execution - how will you make it happen and ensure ROI?

Don’t think that personalisation is a platform feature

Successful personalisation for most organisations actually depends on a big shift in the way that they produce and manage content.  Unless you are ready to invest time and effort in changing your ways of working, you won’t be able to take advantage of the personalisation features that your platform offers.  It’s not all about automation and machine learning, there is a vital human element and it is impossible to personalise properly with flat and unstructured content that hasn’t been well thought through and subjected to some very human analysis.

Don’t overcomplicate - it should only be about providing value to the customer

Given that switching to a personalised content strategy requires such a seismic shift, the sheer enormity of the task ahead could well be your biggest blocker.  But remember you don’t have to start with 10 different customer segments all needing tailored content.  Try thinking of it as customisation, rather than personalisation, and start with one segment.  Use the technology tools available to you to test, iterate and evolve.  It’s much easier to tailor messaging to one group of customers, based on simple criteria such as their location or time of day, than to get bogged down with multiple variants initially. Think big, start small and iterate.

Don’t blindly trust your data

The data you’re using for personalisation needs to be accurate. Sounds obvious?  However,  geolocation is not always accurate, neither is device detection. And people change! If you’re using data given by users, be aware that it will have a limited shelf life.  In short, a huge slice of common sense needs to be applied to the analysis of data.  Continually sense check the information that you are getting and be on the lookout for any abnormalities.  Most importantly, don’t confuse analytics - the tool, with analysis - the application.

Don’t forget that content is still king

You can apply all the clever analytics, segmentation and targeting tools in the world but if you don’t deliver relevant and meaningful content to your identified audiences on the right devices, the entire exercise is rendered pointless.  Personalisation will increase the amount of content production your company does. You need to have enough capacity to cope, or you won’t be able to take advantage of your technology.  Plan ahead and think very hard about how personal you get - aim for cool, not creepy.

This list is by no means exhaustive, it’s just a snapshot of some of the most common themes that came up when a handful of Cognifide clients got together to share ideas and debate the on-going challenges.  Air your views with Cognifide on Twitter - it’s good to share.