4 steps to customer transformation: a series
Step 2: Optimization
AI-assisted transformations deliver incredible results in short time frames. Even a few hours of analysis of your existing output proves the benefits of reimagining your customer communication portfolio. This next step delivers more powerful insights. Once our clients see the results of our initial analysis, Quadient and our clients work together on a larger sample for a few days to uncover even more opportunities to accelerate the transformation of your customer communications.
We spend time training Machine Learning modules to separate and categorise your communications into groupings that are meaningful to your business. We go through a few iterations, often discovering an “AI Surprise,” which is a pattern that people, even experts, would not have found. Quadient, the AI and our client discern patterns in the content at the micro level of content fragments and the macro level of templates and applications. These patterns lead to new ways of organizing communications.
Looking at reports, we re-evaluate the relationships between content, templates and departments. Using the “pilot console” we adjust the sensitivity of the model to variations at the block and template level to give a variety of options to consider as the future structure of the portfolio takes shape. Within a few iterations, we find the optimal model.
Together, we generally work through analysis at the rate of about four hundred inbound communication types per week, and process around one to three million communications per night at capacity during this phase. This allows for a comprehensive analysis by covering as large and as diverse sample of communications that we establish with our clients.
Once we decide on the best possible path forward, we run some larger samples through InspireXpress to ensure that we have caught as many variations as possible. At this point, we have templates, variables and content established. We also have a map of the relationship between blocks of content within templates.
We work together, using cloud-based collaboration technology to make final modifications on templates, add detail, track expert knowledge, describe rules, make comments and optimise structure of the portfolio. At this point in the process, we work to manually override any results the AI suggested portfolio. For example, we combine templates that are similar to reduce ongoing management costs.
We also concatenate blocks to reduce the amount of content fragments that are migrated to the Inspire Customer Communication Management system. As a final step before export to the CCM system, we work to ensure everything is named in a way that reduces the burden for all of the people who will be using the new system, as contemporary CCM technology divides work amongst the relevant stakeholders in the most impactful way possible.