5 Questions to Ask When Using Natural Language Generation to Automate Your Business Reporting

March 8, 2017 Gretta Hermes

automated deployment

For most people, the concept of Advanced Natural Language Generation (NLG) is new. Once understood, its impact becomes clear. While data can help us understand what’s happening in the world, people learn best from stories -- from narratives. NLG is a form of artificial intelligence that turns data into everyday language to improve our understanding of the world.

Using Advanced NLG to automate routine data analysis and communication can greatly increase workplace productivity and allow employees to focus on higher value tasks that can further transform their business.

At Narrative Science, we help our customers evaluate the business impact of automatically transforming their data into narratives. Our experience with customers has allowed us to create some best practices to help your business get the most value out of natural language generation technology.

Start with the Business Goal

The first step is to determine your business goal. While that sounds obvious, people often have a tough time defining what they want from data. A common misconception is, “I'm going to throw my data into this black box, and it will magically uncover all the answers to my toughest business problems.”

A better approach to NLG is to think about how your star analyst would approach finding an answer to a problem. He or she would start with the question they are trying to answer and then use data to find and support the answer.

Advanced NLG technologies function in the same way, but in a scale, consistency, and accuracy not possible with your best ‘human’ analyst. Once an NLG system is taught what you want to communicate, to whom and why, it can write that report as soon as the data becomes available.

For our clients, typical business goals that can be addressed by Advanced NLG include:

  • Automating an existing process that humans do today, freeing up resources and enabling a business to scale in new ways,

  • Creating a new product or feature that can be monetized by expressing the knowledge locked within proprietary and formerly inaccessible data, or

  • Adding personalized insight to each and every customer interaction, enabled through the natural language expression of data

The next phase is to think through the intention of the narratives you want to automate by considering the questions below. Once these questions are answered, the automation piece is easy:

1.Who is the audience?

  • Who is consuming the narrative? Does there need to be versioning and personalization for multiple audiences?

  • Why does this person need the analysis of the data? What action will he or she be able to take after reading it?

2. What type of narratives are you automating?

  • What is your vision for the narratives that will be automated?

  • Is there compliance to consider when teaching the system?

  • Is variability going to be important to the success of the language?

3. Where will the narrative be distributed?

  • What is the typical process for getting this information?

  • Will this be incorporated in an existing dashboard?

  • Will this narrative be sent via email?

4. When will the audience be interacting with the narrative?

  • Will the narrative be refreshed regularly - a weekly, monthly, or quarterly report - or does it need to be available “on-demand”?

  • Does the frequency of the reporting make it inefficient for people to manage today?

5. Why are you automating?

  • How will you measure the success of this new automated process?

  • What higher-value tasks are you going to have your employees focus on, now that they don’t have to spend time writing these reports?

Once you know answers to the 5 W’s, implementing a successful NLG solution is easy! You’re ready to teach our system what you want to communicate and bring efficiency, consistency, and scale to your business.

Visit our Solutions Page to learn more about the potential applications of NLG for your organization >>

Previous Article
What is Natural Language Generation?
What is Natural Language Generation?

Ever wondered what Natural Language Generation (NLG) is? Here are answers to the top five questions regardi...

Next Article
Data Storytelling: 4 Common Myths Debunked
Data Storytelling: 4 Common Myths Debunked

Data storytelling is a powerful asset but common myths prevail around the practice. This blog post debunks ...


Get Narrative Science blog posts in your Inbox

Keep an eye out for your confirm email!
Error - something went wrong!