Partnering with HumanFirst, Infobip generated over 220 knowledge articles, unlocked 30% of their agents' time, and improved containment by a projected 15%.
Reviewing the state of employee experimentation and organizational adoption, and exploring the shifts in thinking, tooling, and training required for workforce-wide AI.
With a one-click integration to Conversations, Infobip’s contact center solution, HumanFirst helps enterprise teams leverage LLMs to analyze 100% of their customer data.
Partnering with HumanFirst, Infobip generated over 220 knowledge articles, unlocked 30% of their agents' time, and improved containment by a projected 15%.
Reviewing the state of employee experimentation and organizational adoption, and exploring the shifts in thinking, tooling, and training required for workforce-wide AI.
Most businesses are sitting on high-quality Voice of the Customer Data , generated across a growing number of channels in the context of sales, marketing, product and operations.
Examples of activities and channels generating this broad subset of natural language data include:
Voice of the customer (and employee) data is a goldmine as a source of both qualitative insights and training data for AI.
This data is actionable, and provides a direct, often unfiltered view into what customers or employees need or want, right now.
Some of the most useful applications unlocked by this data are:
What's fascinating, is how the same data can provide completely different butcomplementary value across teams of different technical skillsets:
Collaboration around this voice of the customer data is the key to significant improvements in customer experience.
Let's look at an example
A fintech we recently worked with was intent on improving their activation metrics: the last step of their onboarding flow required users to transfer money into their new account; until this step was done, users were essentially "lost" for this fintech.
Improving this last part of the flow would have a critical impact on the customer acquisition cost Â
The marketing department launched a proactive SMS outreach campaign to all users in the bucket "signed up but didn't activate yet": human agents initiated short personalized conversations with each and every user in that bucket, with the objective of understanding why they hadn't transferred money yet, and convincing them to do so.
Hint: SMS campaigns are a great way to gather immediate and actionable customer insights; however even without this type of investment, existing live chat logs, help center queries and email support tickets likely provide answers to many of the same questions!
The initial SMS campaign was done over 2 weeks and ~2k users, and led to important insights:
75% of the customers weren't aware that signing up gave them access to a virtual card that could be used to make online purchases (regardless of whether they had received their physical bank card or not)
15% of the customers reported not having sufficient money at that time to transfer (but signified intention of doing so)
A non-negligible portion of users banking with a specific institution reported issues when trying to transfer money: instead of reporting this to customer support, they simply gave up
What business value was derived from these insights?
The product team adapted the onboarding journey to prioritize the value of virtual bank cards for new customers, both creating a better entry point to this section of the app, as well as crafting content explaining scenarios in which these cards are better suited than physical ones.
The marketing team launched a post-signup email sequence geared towards users who are living pay-check-to-pay-check, showing the specific advantages provided by the fintech's offering compared to big banks for this particular demographic.
The customer support team addressed gaps in the Help Center by adding articles directly addressing the most recurring issues and bugs raised by users.
Generating these insights also led to building a highly custom training dataset that could be used to train a natural language understanding (NLU) classifier.
Some of the applications that could easily be built by machine-learning, data science and conversation design teams using this type of classifier include:
Tracking the occurrence of specific issues and user feedback over time, across the different voice of the customer channels, and reporting this via an NLU-powered dashboard.
Scaling and automating the SMS conversations, by augmenting the human agent with automatic responses and playbooks that help drive more conversion.
Triaging and automatically tagging users in the CRM based on what they say in the live chat or help center: doing so could for example automatically trigger a business process to re-send a bank card for any user who didn't receive it, without requiring a manual intervention by customer support agents.
Training a chatbot or virtual assistant that automatically recognizes these different customer intents and provides optimized journeys to convert them.
What next?
Until now, there was no general purpose solution to easily and quickly work with this type of data, andfor this reason most of the business value contained in this data has stayed locked away.
HumanFirst is unlocking this opportunity for teams of all sizes and across all industries, with a solution that provides completely flexibility and control over natural language data (and more specifically, voice of the customer data!). Try it out now!
HumanFirst is like Excel, for Natural Language Data. A complete productivity suite to transform natural language into business insights and AI training data.
Subscribe to HumanFirst Blog
Get the latest posts delivered right to your inbox
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.