Chief Technology Officer, Martin Paton discusses his top tips for those implementing a new CMS/CXM this year.
I've been travelling the world helping organisations implement CMS/CXM platforms in some way for the best part of 15 years now and with the landscape as diverse and complex as ever, I hope the following will help guide you towards a successful CMS/CXM implementation in 2019.
Focus on highest-value, measurable outcomes not lists of features the platform should support
Working in an agency, I have personally been on the receiving end of unending lists of MOSCOW prioritised requirements a platform must/should/could have which are often just lists of features with no qualitative or quantitative measures that prove implementing the feature will have a positive effect on the user.
By way of example:
The platform must support WYSIWYG
A slightly better outcome-driven example of the same requirement:
We believe implementing a visual editor will allow authors to create a page with fewer defects during the approval process which should reduce the overall time to create a page meaning our editors can create more pages per day at a higher, more consistent quality. Success will be measured as an increase in the number of pages published per day with a downward trend in the number of defects at approval
Broken down this translates to:
[hypothesis] - [desired outcome] - [how it will be measured]
Thinking in an outcome driven way over lists of features will sharpen your thinking towards developing and prioritising hopefully fewer backlog items that can be measured correctly and add genuine positive benefit rather than be a waste of time and money.
Don't rely on the sales demo
Sales demos have been very carefully tailored to artificial scenarios over a long period of time. They will undoubtedly look impressive, but may be gloss over hidden issues that surface at time of implementation which in the worst case could leave you high and dry.
Focusing on outcomes over features again plays a key role in avoiding this pitfall. Work the vendor and ask for a tailored demonstration and gain proof that the implementation will support your desired outcome. If this proof requires development that is beyond fair investment by the vendor it may be worth setting aside a small budget to allow for your implementation expert partner to create spikes or more involved POCs.
This approach will almost always pay off in the long run - it will significantly reduce the risk of a failed project and also allow you to see in more detail how the vendor works and where any issues or concerns may arise so you can head these off before getting into the meat of the project delivery.
Cost everything out in detail
Implementing a platform has many cost facets. It's tempting to focus purely on the ticket price of the platform plus the agency implementation cost.
However there are a lot of hidden costs that will need to be budgeted and set off against any ROI predictions - items like staff costs, lost opportunity costs, staff reassignment, data cleansing, migration, business process re-engineering, retraining and staff ramp-up, new ways of working alignment, etc. All must be accounted to give as much accuracy and visibility to the C-suite. Plus with modern platforms being almost entirely SaaS, you need to play the black magic game and predict your usage sensibly to both avoid over-specifying on one side, and a hefty bill on the other.
Work out a realistic implementation roadmap and don't underestimate the implementation itself
Today's platforms come bursting with promises of "out of the box" end-to-end marketing automation, personalisation, AI, etc. Whilst the power this places into the hands of the marketing and development team is undeniably immense, it doesn't all happen from day one and the phrase "out of the box" should be treated with extreme caution. It may well be in the box, but the implementation that fits your business ambition almost certainly won't be.
Be realistic about what you can achieve in the first 6-12 months.
Getting the architecture right is crucial to allow scalability and avoid cul-de-sacs. Capturing enough data to make AI systems effective is constrained mostly by time and optimising your CX will require much manual experimentation before you can unleash automated AB optimisations. Don't be afraid to change things that obviously aren't working but do allow time for the platform to bed in so you can start to see trends emerging against your outcomes, and finally remember to get your measures in a dashboard that are transparent to all stakeholders.