Now available for a song
by Joe Dysart * Updated Jan. 30, 2022
Legions of workers stumped by company charts and graphs that make no sense to them can take heart: A new, dirt-cheap AI software tool from Microsoft enables you to trigger a text explanation of a chart or graph with just a mouse click.
Moreover, the new tool also enables you to string together text explanations of multiple charts and graphs, which together can create a report on a company’s performance — completely and automatically written by your computer.
Dubbed ‘Smart Narrative,’ the new tool is included in the latest version of Microsoft Power BI, Microsoft’s business analytics software.
The new Microsoft tool is also emblematic of a number of software solutions growing in popularity at businesses.
Those solutions are also using AI-generated writing to explain what’s going in a chart or graph.
Or, they’re bypassing business visuals altogether and auto-writing text-based company reports — which are virtually devoid of complicated chart-and-graph visuals.
There are many reasons why companies and organizations should be fascinated by Microsoft’s foray into auto-generated text explanations of charts and graphs.
But one of Smart Narrative’s biggest draws is that it’s so intuitive to use.
For example: If you’re looking at a chart in Power BI you don’t understand, just click on it with your mouse and Smart Narrative will auto-generate a pop-up text box that explains in plain English what’s going on, according to Ancy Philip, an AI specialist at Microsoft.
You can see for yourself how this ‘click-and-learn’ feature works using the Smart Narrative tool — and how it makes sense of a sample scatter chart — in this four-minute video from Microsoft.
Specifically, the video will show you firsthand how sales in the hypothetical cities of Crystalcastle and Esterwyn are doing — as illustrated by a scatter chart.
And then it will show you how to get a clear English explanation of what’s going on in that scatter chart with a simple mouse click.
Essentially: Any time you’re confronted by a chart, graph or other visual in Power BI that mystifies you, you can use the new tool to click on the image and have it explained to you in plain English.
Not surprisingly, the Smart Narrative tool is expected to be a boon to a wide swath of workers who embrace business charts and graphs with all the enthusiasm of an impending root canal.
Like similar tools offered by a number of companies specializing in AI-generated writing, the Smart Narrative tool works its magic by drilling down into company databases, crunching the data it finds there, and rendering the resulting insights it uncovers in plain English.
Essentially, these are ‘Company reports that write themselves’ — or precise, instantly understandable analyses of company trends, performance and similar insights — auto-written by a machine.
Until recently, this kind of techno-wonder — also known as Natural Language Generation (NLG) — used to cost big bucks.
But as with most things tech, the price of AI-generated written reports — in some cases — is dropping dramatically. These days, a subscription to Microsoft Power BI, for example — which includes the new Smart Narrative tool — goes for as little as $20/month/seat.
Observes Ehud Reiter, chief scientist, Arria NLG, a pioneer in AI-generated writing and auto-generated reports: “Narratives (simple to understand sentences) can be better decision-aids than visuals, because they deliver focused and powerful insights.”
In contrast, forcing people to rely on charts and graphs to discern business insights can actually undermine understanding, Reiter adds.
“Many people looking at graphs get distracted by patterns that aren’t actually significant or relevant,” Reiter observes.
Sharon Daniels, CEO, Arria NLG — another pioneer in AI-generated writing — agrees: “While visuals (charts and graphs) paint a picture, they’re not the complete picture.
“With the addition of NLG (natural language generation), business intelligence dashboards are transformed,” Daniels says.
“The ability to access key information in near real-time — communicated as if written by the company’s top analyst, without bias, at an NLG writing-speed — is truly astonishing,” Daniels adds.
How The Auto-Report Writing Method Works
Generally, the AI companies that pioneered auto-report-writing method use an AI-populated, ‘report template’ to auto-generate their reports.
These templates work on the same principle as the old Mad Libs game you probably played in grammar school.
Back in those days, Mad Libs furnished you with the template of a story — say, how pizza was invented. You asked your friends for random nouns, verbs and adjectives to plug into blanks that Mad Libs left in the template.
And then, as you read the completed story aloud, everyone roared in laughter at the nonsequiturs — mostly because all your buddies at the time were still in pigtails or chewed bubble gum for a living.
Auto-report templates used by AI writing companies work the same way. They use a skeleton of a story — say a report on daily sales — and leave blanks in the template.
Those blanks are designed to be populated with the latest numbers on a company’s sales numbers, computational analysis of those sales numbers and similar.
Once the auto-report template is rigorously and repeatedly tested to ensure it generates flawlessly accurate copy about daily sales, it’s activated within the AI-generated system to start auto-generating reports.
A daily sales report template, for example, will continually tell the story on your company’s daily sales anew as fresh data is added to your database each day.
Observes Akanksah Ramchandani, an associate product manager at vPhrase , another AI-generated writing firm: “Such accurate analyses and insight generation minimizes the possibilities of data misinterpretation — and makes decision-making effective and fast.”
In terms of AI-generated company reports, the big difference with the Smart Narrative tool in Power BI — compared with some other AI report writing software — is that Power BI still generates a chart or graph on virtually any aspect of company performance.
But simultaneously, Power BI also enables users to auto-generate a written analysis of that chart or graph with a click of the mouse.
The decision to stick with a dual expression of insights from company data — charts and graphs as well as text explanations — is a clever move by Microsoft.
It ensures that the importance of the charts, graphs and similar illustrations in Microsoft Power BI is preserved — the bread-and-butter of the analytics program, really — while simultaneously offering graphics-averse users plain English descriptions of what the charts and graphs are trying to say.
In a phrase, the Smart Narrative tool — along with similar AI-generated writing software tools — can be a game-changer for companies that use business analytics software like Microsoft Power BI.
Equally impactful: AI-generated reports analyzing company performance also solve another major problem that has plagued companies using business intelligence programs like Power BI, Microstrategy, Alik, Spotfire, SAP, Tableau.
Specifically: Getting company performance insights unearthed by those programs into the hands of more people, more quickly.
For years, companies using business analytics software have had to deal with the fact that despite its many advantages, many of the valuable insights embedded in business charts and graphs are lost on workers averse to such visuals.
Plus, these same companies have also wrestled with the hard fact that they often need to hire highly specialized data scientists to work with programs like Microsoft Power BI — simply to get their business visuals created in an accurate and timely manner.
Even more nettlesome: Many of these same companies find those same expert data scientists often must be tasked to translate what’s going on in the charts and graphs, by offering live presentations to company personnel.
Or, the experts must spend time on their keyboards writing their own reports to explain the same insights to workers.
The result? Mission critical information from business intelligence packages like Power BI is often trapped in the hands of a relatively few expert users.
Observes Shruti Vasudevan, senior content strategist, vPhrase: While “BI (business intelligence) tools help businesses to analyze their data, data scientists and technical teams have to spend considerable time and effort in preparing data and drawing and explaining insights from the data.”
The good news: Not only does Smart Narrative solve the data insights logjam.
It also ensures that many of the templates needed to auto-explain the meaning of a chart or graph generated by Power BI come standard with the program, already baked in.
Essentially: You’ll find while you’re using Smart Narrative that you don’t need to write special templates to generate text explanations of many — if not most — of the charts and graphs you rely on when working with Power BI.
Instead, you’ll discover Microsoft Power BI and its Smart Narrative tool has already pre-fabricated those templates for you.
Other Helpful Automation With Smart Narrative
Still another benefit of using the Smart Narrative tool is its ability to quickly unearth subtle insights in the data — which may get lost when Power BI tries to express it graphically.
Puzzling over all the dots and lines and colors on a sales scatter chart, for example, you or another user may find it difficult to instantly discern that during the past 30 days say, sales were flat for say the first 18 days of the month — but rose during the last 12.
But if you were to click on that chart in Power BI, you would instantly trigger that insight in plain English: “During the past 30 days, sales were flat the first 18 days, but rose during the last 12.”
Clear. Simple. To the point.
No guesswork.
Besides generating basic insights, the Smart Narrative tool also is also able to slice-and-dice data portrayed in a chart or graph.
For example: You may be looking at a chart in Microsoft Power BI that depicts company sales for the last 30 days.
But if you highlight just the last 15 days of sales portrayed by the chart and click on it, Smart Narrative is astute enough to sense you’re looking for insight on just the last 15 days.
Consequently, it serves up its take on just those last 15 days in easy-to-understand text.
Another key advantage of the Smart Narrative tool is its ability to auto-update the written reports it generates, as new data is added to your company database.
For example: If you have a chart tracking company sales by region over the last 30 days, Smart Narrative will generate a fresh, new, written report on those sales every day — given that the Power BI software is always tracking the last 30 days of your sales.
Yet another powerful feature of the Smart Narrative tool: Its ability to simultaneously auto-generate text explanations of multiple charts and graphs.
This capability will prove especially handy for users who program Power BI to open-up to a visual dashboard that simultaneously tracks say 6-12 different kinds of charts and graphs.
Instead of attempting to discern insights from all those charts and graphs by trying to decipher colors, shapes and lines, you’ll find the Smart Narrative tool will do it for you by offering pop-up, text explanations of what each graphic is trying to say.
Need custom templates to auto-generate text explanations of chart and graph insights that currently are not available with the Smart Narrative tool? That’s also no problem.
Expert users of the Smart Narrative tool will find the software offers them the ability to easily create such highly customized templates to ensure they can generate the precise kind insights they need.
Power User Tip: Adjusting Smart Narrative to Write in Your Style
Once you’re comfortable with the Smart Narrative tool, yet another powerful feature you’ll find is that you can customize the writing style Smart Narrative uses to auto-generate text explanations of charts and graphs.
This capability — in the hands of a professional writer — is in no uncertain terms a game-changer in its own right.
The reason: While Smart Narrative is plenty smart and gets the job done using plain English, its writing style is a bit plain and barebones.
But if you punch-up that writing style by adding some sparkle and wit from a pro writer, you’ll find the text reports Smart Narrative generates will grab a worker’s attention more forcefully, hold onto that worker’s attention much longer — and ultimately better ensure Smart Narrative’s key insights will stay top-of-mind for your company’s decision-makers.
Essentially: A gripping writing style is why key decision-makers at organizations across the globe regularly quote and rely on the reporting in the Wall Street Journal, The New York Times, the Financial Times of London and similar — and have long-since abandoned lesser competitors.
Moreover, a pro writer will also come in handy if you plan to use the Smart Narrative tool to string together a number of written insights from a number of charts and graphs to auto-generate a longer report that offers deeper, more significant analysis of what’s going on with your company.
In a sense, hiring or tasking a pro writer to customize the writing style of the Smart Narrative tool reinforces a time-tested maxim: Hire the right person for the right job — and you’ll never go wrong.
Essentially: You wouldn’t hire artist-of-the ages Michelangelo to pen an epic poem for you, because Michelangelo’s gift was painting — not writing.
And in the same vein, it makes much more sense to hire a pro-writer to punch-up the writing style of Smart Narrative — rather than assigning the task to ‘no one in particular.’
Equally important: Your best choice of a pro writer for slam-dunking the task of writing gripping report templates for you is a business feature journalist.
Business feature journalists — or people who can write like business feature journalists — are trained to simultaneously engage and entertain a business audience.
And they’re trained to do so without resorting to hype, distortion or fabrication.
As for the overall take-away on the advent of the Smart Narrative tool and how it’s helping change how business analytics is done?:
For companies looking to get a taste of how AI-generated company reports work and how auto-written analysis of company insights can turbo-charge the speed at which business analysis can be easily dispersed throughout an organization, Microsoft’s Smart Narrative tool is a great place to start.
But over the long term, you’ll definitely want to take a look at similar AI auto-report writing tools offered by pioneers in AI-generated writing, including:
These companies have been working at commercialized AI-generated writing longer than Microsoft.
And they may have AI writing solutions more to your liking.
Those include AI-generated writing add-ins of their solutions.
These are tools designed to seamlessly integrate with a number of business intelligence programs, including Microstrategy, Alik, Spotfire, SAP, Tableau — and Microsoft Power BI.
Key Features to Look For in AI Generated Writing Software
Before opting for any one of these solution providers, there are some key features you should look for in any AI-generated writing package.
You should know going in that while a tool like Smart Narrative will auto-generate a quick text description of a chart or graph — or a number of quick text descriptions of multiple charts and graphs — getting AI writing to auto-generate longer and more sophisticated reports for you will most likely require a bit more effort.
Essentially, if you’re looking to go beyond simply translating charts and graphs into short text-explanations-on-demand, you may need to task one or more people at your firm to become expert at manipulating and customizing any AI-generated writing software your company buys.
Observes Brian Everett, senior partner, MindShare Strategies, a marketing firm, regarding more sophisticated uses of AI-generated writing: “People must invest a significant amount of time and resources to set-up and perfect NLG Software on the front-end.”
Plus, “After reports and articles are generated, people still need to review and tweak the text (template) to ensure it makes sense and is error-free,” Everett adds.
Another critically important note before you sign on the dotted line: The companies mentioned above are able to render extremely precise company reports because they rely on the strict data-into-text (Mad Libs) method.
This approach differs markedly — and fundamentally — from another AI-generated writing method that has mushroomed in popularity since 2020: Automated writing based on AI-driven predictive analytics.
Essentially, this alternate approach — popularized by OpenAI, the makers of the GPT-3 auto text writing engine — generates copy using an extremely different methodology.
Instead of drawing info from specifically chosen data tables, the GPT-3 auto-text-writer and its ilk simply request a user input a few lines of text into the software — triggering the auto-text generator to ‘guess’ what kind of text the user is looking for — often with great success.
Undisputedly, this alternate approach to AI-generated writing has stunned many computer scientists and others across the world with its ability to auto-generate usable text using just a bit of input from a user.
And already, a number of companies have licensed the GPT-3 auto-text writing engine to create apps that sit atop GPT-3 to auto-generate emails, create short ad slogans, whip-up short blog posts and similar.
Even so, OpenAI and many of the researchers championing this alternate method of auto-text generation are among the first to admit that supercomputer-driven writing engines like GPT-3 are often less-than-accurate.
In fact, such auto-text generators are often extremely ham-fisted about coming back with what a user is looking for.
And so far, GPT-3 like solutions are no match for the auto-writing precision offered by the company report writing solution providers above.
Truth-be-told, GPT-3 and its ilk often need a human-in-the- loop to do a final edit of any long-form writing they come up with — which is fine for the kind of writing they’re offering customers.
But when it comes to the automated writing of company reports that live-and-die on the accuracy of the numbers and analysis featured therin, GPT-3 like auto-text generators are currently no match for the Mad Libs school of automated writing.
Essentially, when it comes to automated company reports, business users overwhelming want set-it-and-forget templates, which can be relied upon to generate flawlessly accurate reports once those report templates have been fully tested and refined.
(For an in-depth look at GPT-3, check out: “GPT-3 and AI Writing: Stunning, if Imperfect,” by Joe Dysart.)
Given those caveats, here are the key features to look for in any decent AI-generated writing software specifically designed to auto-generate flawless company reports:
*Custom Analysis: Currently, scores of AI-generated writing toolmakers have the ability to transform data into simple text. But fewer offer the added ability to manipulate that capability to unearth highly specific insights.
The best AI-generated writing tools offer the greatest flexibility for unearthing those insights.
*Granular Focus: The best AI-generated writing solutions also enable you to easily produce reports that can be tailored to unearth highly specific insights – even insights that may be of use to only one individual in your organization.
*Real-Time Updating: Many AI-generated company report writing solutions offer a ‘set-it-and-forget-it’ option. This enables users to design a pre-defined report that is updated unendingly with real-time data.
Continuous updating is an extremely handy feature, for example, if you’re monitoring critical data that’s streaming into your organization on an hour-by-hour — or even minute-by-minute — basis.
*Ability to Port Insights to Text-Chat Bots and Voice-Chat Bots: While not every organization is currently using chatbots to communicate with customers – or to foster better communication among employees — you may want that option down-the-line.
As chatbot technology matures, for example, a chatbot backed by an AI-generated writing system will ideally be able to carry-on a conversation with a customer in a very natural, intelligent and innovative manner.
Ditto for chatbots your employees may want to start using to query your database for easy answers from the insights buried there.
Such AI-powered chatbots would be a welcome change from conventional chatbots, given that too often, conventional chatbots come off as overly robotic.
There also often flummoxed by even the seemingly easiest questions.
*Pre-Fabricated Add-Ins: A number of AI-generated writing pioneers now offer pre-fabricated add-ins you can use to integrate their AI writing capability into your favorite business intelligence software.
A number of solutions providers, for example, offer pre-fabricated, AI writing add-ins for Power BI, Microstrategy, Alik, Spotfire, SAP and Tableau.
A pre-fabricated add-in — which has been rigorously designed and tested to work with your favorite business intelligence software the moment you add it to that solution — saves you all the time and effort you’d ordinarily need to custom-integrate an AI-writing solution into your business intelligence software.
*Custom Writing Style: As previously noted, basic AI-generated text has a reputation for being a bit humdrum.
Less commonly known is that the best AI-generated writing solutions enable you to customize the style of the company reports and/or other text they generate for you.
Essentially: if you want to go beyond the pedestrian, you can bring in a highly creative copywriter — ideally well-suited to business feature writing — to tweak the language style of the templates driving your AI system to your preference.
On The Horizon
Overall, given the great value AI-generated writing offers organizations, look for a number of new AI-generated writing providers to emerge during the next few years.
You can expect them — as well as current players — to offer you ever-more powerful and ever-more versatile systems for automated company report writing.
Such solutions will come in the form of stand-alone AI writing tools like those from Automated Insights and Arria NLG.
And others will emerge as tools already integrated into comprehensive business software suites like SAP and Salesforce.
Still others will be built into application-specific software. For example: AI-generated writing tools are already proliferating in highly specialized programs that auto-write contracts for lawyers, auto-create exams for educators and auto-generate research summaries for academics.
James Richardson, a specialist in business analytics at Gartner — and author of the June 2021 Gartner report, “Augmented Analytics: Teaching Machines to Tell Data Stories to Humans” — sees great promise in this kind of increasing sophistication during the next few years:
“Automated data storytelling software is becoming available that can monitor a huge variety of data and find not just anomalies — but deeper patterns in datasets.
“It can deliver the information to decision makers in the form of rich narratives that have an understanding of what decision makers are interested in.”
“The approach has the potential to analyze far more data than would be possible by an analyst population — and to arrive at insights from data that are less colored by human attention and bias.
“Effectively, the data itself tells the story.”
Richardson’s personal picks of companies he says are ahead of the curve on this evolution of AI-generated writing include Stories.bi (now owned by Workday), Narrative Science Lexio, Datastories.com and Yellowfin Signals.
The bottom line on the Smart Narrative tool and similar AI software that can auto-generate company reports: Even if you’re not in the market for AI-generated writing at this moment, you’ll want to track this tech closely.
In coming years, AI-generated writing is poised to dramatically alter the way organizations and companies work with, understand and leverage data.
Share a Link: Please consider sharing a link to https://RobotWritersAI.com from your blog, social media post, publication or emails. More links leading to RobotWritersAI.com helps everyone interested in AI-generated writing.
–Joe Dysart is editor of RobotWritersAI.com and a tech journalist with 20+ years experience. His work has appeared in 150+ publications, including The New York Times and the Financial Times of London.