Enterprise organizations that hope to better understand their business have more access to information than ever before. Data can come from a multitude of sources, with very little lag time between when it is collected and when it is delivered to the analysts who will leverage it. 

The problem that analytics and business users have faced for years isn’t the availability of data, but how they analyze and extract insights from that data. The current state of data analysis is manual, non-standardized, time-constrained, and biased. 

And the problem of data analysis isn’t limited to analytics and insights teams, it creates a snowball effect that affects the entire organization. 

In this blog we’ll talk about the current state of data analysis and how AnswerRocket aims to simplify data analysis and democratize insights for all. 

The Vicious Cycle of Data Analysis

Companies usually have no shortage of tools that they’ve invested in, dashboards to access, recurring reports to review, and other tools that they can review to evaluate their performance, but this typically only gets them so far. These views of data typically prompt follow-up questions and require further exploration to understand:

What is going on?

Why is this metric up or down?

These are questions that business users or key decision makers who are trying to take an action can’t easily answer on their own. This means they then have to involve an analyst or insight team member who has to dig into the data, compile supplemental data, evaluate it, interpret it, and then translate it into an answer that the business user can understand. 

Analysts, Business Users, Executives and more can all relate to the “vicious cycle” of gleaning insights from their organization’s data.

This then prompts more follow up questions, and the vicious cycle continues. This process is currently very time-consuming, and inefficient. 

Why Does This Matter?

Organizations that are stuck in this vicious cycle are likely experiencing one or more of these side effects as a result:

• Wasted time and resources
• Wrong strategies and tactics in the market
• Missed company targets
• Vulnerability to competitors

The challenge to analysts and insights leaders is, “How do we get people the answers they need, FASTER?” 

Faster insights means better, informed business decisions can be made more quickly, which can mean the difference between outpacing your competitors or getting left behind.

Who Does the Vicious Cycle Affect?

The vicious cycle of data analysis is a broad enterprise problem that doesn’t just stop at the analysts’ desks. At the core of this challenge is the fact that despite the vast amounts of data available, gleaning actionable insights remains a significant stumbling block. 

• For executives, this cycle means strategic decisions are often delayed or are based on incomplete insights, potentially resulting in poor market tactics or even missing crucial business targets.
• Business users, typically in need of quick answers for operational decisions, find themselves stuck in a loop of follow-up questions for which they are often reliant on analysts to help answer.
• Analysts, entrusted with the task of data analysis and interpretation, often find themselves overwhelmed by the demands of multiple stakeholders, which they juggle alongside business-critical initiatives.
• For data scientists, this process can be particularly frustrating as the tools and dashboards they’ve crafted, while advanced, still don’t yield the streamlined, quick insights the business desperately seeks.

What do these personas need to break out of the vicious cycle?

What Does the Vicious Cycle Look Like in Real Life?

Global Pharma Manufacturer

This pharmaceutical manufacturer has been trying to understand their performance in the market, as it relates to the sales team’s performance. 

They have a huge sales force that is in the field talking to doctors and hospitals on a regular basis, promoting the company’s prescriptions. The ability to analyze the efforts of this sales force to see what’s working, what isn’t working and change tactics on the fly is crucial to the success of the company.

Up until now, they have never been able to combine their market performance with what their sales time is doing. Questions like:

• How is our sales team impacting a change in the market?
• Does more prescriptions written equate to more share in the market?

AnswerRocket developed the omnichannel sales driver copilot, combining data that has never been combined in the past: market performance + sales teams efforts. Sales leaders can now see what’s working, what’s not, what’s driving growth, what to avoid, all with the goal to drive additional share gains.

By implementing this copilot and follow up questions, this allows them to pinpoint regions or districts that are struggling (to help) or pinpoint areas that are succeeding and implement what they’re doing in other areas. 

Global Beverage Leader                       

This multinational enterprise has prioritized working with brand equity data to drive their global marketing efforts and pricing strategies. This data captures how consumers perceive  their brands in the marketplace. 

In the past, someone like the CMO might ask “why is brand equity down in Europe?” and then everyone would drop what they’re doing and try their best to tell a story from very complex data.

Their data provider partner does not currently have great tools for storytelling. They have dashboards, but that’s it. 

With the Max brand guidance copilot, they can  glean insights and answer questions faster than ever before,  taking a 21-day process each quarter down to just 5 days (a 75% reduction). 

Previously, it was nearly impossible to keep a “real time pulse” on what was going on in the marketplace. But now that they’re able to cut out about 4 business weeks of wait time, they’re able to be as agile as they want to be. 

How are We Breaking the Vicious Cycle?

With our first-of-its-kind AI assistant for data analysis, Max, we put the power of data analysis in the hands of everyone in your organization, not just a select few. No matter your role, your technical experience, or what your insights goals are, Max is here to help. 

By integrating with the power of GPT with AnswerRocket, Max allows users a chat-based analytics experience that allows them to ask questions in basic, conversational English, and get answers back in plain English. By getting “beyond the dashboard” and avoiding complex reports, there is no longer a steep learning curve. 

This element of self-service allows business users and executives to  investigate changes on their own, and ask follow up questions on the fly for deeper insights. 

Check out this 90-second Max demo below to see more of what Max can do. 

Bringing Max to the Enterprise

At AnswerRocket, we have been working with each individual customer to thoroughly understand the problems they’re experiencing. Each customer situation is nuanced, and has a workflow that we need to fit into.

We like to understand:

• What is the current state of data analysis in the organization?
• What should the future state of data analysis look like?
• What is their vision for what an AI assistant could do?
• What should the output look like?
• What insights are they missing today that they would like to have?

The process is very consultative and individualized. 

We are trying to understand the skills that are required to solve each customer problem. Information such as what are the data sources and formats, how do we insert our tool to make it a seamless part of their existing workflow?

In this way, we can co-create a purpose-built analytics copilot that addresses our customer’s needs head on. 

Max is helping to achieve the AnswerRocket mission more than ever before, automating meaningful insights to help decision-makers take action, faster. 

The post Breaking the Vicious Cycle first appeared on AnswerRocket.

Watch the video below or read the transcript to learn more.

Is Max dependent on GPT or can other LLMs be used?

Mike: So it’s 100% flexible to use lots of different versions of GPT, or lots of different locations where the language models are stored, or lots of different language models. So we look at the language model very much like a database, something that over time will become faster and cheaper and more commoditized and we want to be able to swap in and out whatever those models are over time, so that we’re not dependent on it, on any one. We do use every capability that’s available to us from the language models, things like completions and embeddings, these are technical terms of the capabilities of the models and we will look for those same capabilities as we expand into additional models. But it’s not a dependency for our solution. And in fact there is a mode where AnswerRocket can run, in fact has run, until about six months ago when these language models were introduced. 

That does not rely on external language models at all, right? It relies instead on the semantics of the database, on the ontology that’s defined by a business and how they like to use their terms. And so it does not rely on having to have a GPT source. But when there is a language model in the mix, you get a more conversational flow to the analysis which makes it feel a lot more comfortable to the user. It’s clear that from a foundation model perspective, the providers of the core algorithms behind these models, there will be models that are specific to medical, that are specific to consumers, that are specific to different industries and different spaces. And so we very much expect to be able to multiplex across those models as appropriate for the use case and again treat them like any other component of infrastructure, whether that’s storage or database or compute. 

These models just become one more asset that’s available to enterprise applications that are really putting together productivity suites for the end user. 

Conclusion: AnswerRocket is not solely dependent on GPT; in fact, it initially operated without relying on external language models, using database semantics and business-defined ontologies. However, when language models are incorporated, it enhances the user experience, enabling a more conversational flow in data analysis. The focus is on leveraging the diverse capabilities of language models while treating them as components of infrastructure alongside storage, databases, and compute resources. Analytics and Insights experts like Mike foresee a future with specialized language models catering to various industries. The aim is to provide enterprise applications with enhanced productivity suites for end-users by multiplexing across different models as needed for various use cases.

The post The Future of Language Models in the Enterprise: A Multi-Model World first appeared on AnswerRocket.

ChatGPT Enters the Scene

ChatGPT launched in November 2022 and within 5 days it had reached one million users, shattering every previously held record by an online service. Users piled on to create free accounts and dive into the new technology headfirst. Some were just curious, others wanted to understand the rapidly changing artificial intelligence landscape. Still others, like AnswerRocket, wanted to understand how to utilize this new technology to the benefit of their customers.

Making Waves in AI and Beyond

There’s a concept from marketing guru Seth Godin that talks about the elements that make up the much sought-after “brand crush.” A brand crush is a love affair that consumers have with a brand that they can’t live without. They consume said brand frequently, talk about it fervently, and no substitute will do.

The elements that create a brand crush? Magic and Generosity. 

The magic of ChatGPT is easy to see. OpenAI has taken the power of artificial intelligence, plus billions of data points, and put them together to provide users answers within seconds. It was unlike any tool any of us had ever seen and we couldn’t wait to show our friends. 

The generosity piece of the equation is that ChatGPT is completely FREE. That’s right. Some of the most powerful technology in the world, boxed up with a ribbon on top and handed over to the world. 

THAT is how they reached over one million users in just 5 days. They created a brand crush between ChatGPT and the world, and it went viral. 

What Can Analytics Leaders Learn from ChatGPT?

In a world where a 25% adoption rate is accepted as the norm, analytics leaders can borrow a few tricks from ChatGPT’s playbook to spark adoption of analytics tools within their own organizations. 

1. K.I.S.S. (Keep It Simple Stupid)

Your non-technical users are more likely to be intimidated by complex processes and dashboards that stand in the way of getting the answers they need. So make it unbelievably easy to use, right away. 

Upon signing in, ChatGPT users are greeted with a simple dashboard that outlines some basic examples, capabilities and even limitations of the platform, followed immediately by a search bar where users can enter their prompt for the tool. 

Recent activity is saved in the right hand column, making it easy for users to resume previous activities. Also on the right side near the bottom are the standard account/settings/help options. 

That’s it. No fancy bells and whistles, no setup, no training, just create an account and you’re in, ready to start using the tool. 

Analytics leaders can lean on this example when preparing their teams to use a new analytics platform.

→ Identify what content is the highest priority to review, and limit yourself to only showing that content. This way, you avoid overwhelming users with too much information to start.  

→ Focus on meeting people where they are now, instead of what they could be doing with the tool in the future. A gradual build up of question complexity grows confidence. While ChatGPT 4 has the ability to write code for a website, you can bet most users are not starting there.

→ Start with simple functions and walk users through the process, step-by-step at a slow pace. This allows time for and creates an environment that welcomes open dialogue and questions.

2. Build Off Existing Skills

Part of why ChatGPT is so successful is because the general public has already learned how to interact with a chat interface. If you handed someone who’s never had a computer an open laptop with ChatGPT up – they’d really struggle. But ChatGPT shines because it knows that its most common user base is made up of folks who are familiar with this kind of setup. It’s natural to find the text box, use a “send” button (paper airplane) and continue to interact.

→ For analytics leaders, find what’s familiar to your users and their skill sets. This could be building off of presentations they’re used to reading, or referring to another tool (like Excel or PowerPoint) to ground explanations in your analytics tool.

3. Deliver Immediate Value

In the same way that users could create a free ChatGPT account and immediately get a recipe for that evening’s dinner, users want to see immediate value in the tool.

Because of the simplicity of the interface, ChatGPT users can choose to dive right in if they decide to. 

When preparing their own teams to use a new tool, analytics leaders can lean into the ChatGPT experience by offering a thoughtful first experience. Loading data, testing, and training prep all go a long way to make sure the solution is polished and functional when your team sees it for the first time.

→ Keep training sessions streamlined and focus on delivering a quick win. This helps create excitement and drive adoption.

→ Address a couple of key use cases in your training sessions and demonstrate how the tools improve upon the current processes.

→ Offer resources for users to dive-deeper on their own AND be available for one-on-one problem solving sessions. This allows users to get answers in the best environment for them.

4. Make it fun!

While it may seem counterintuitive for data analysis to be fun, delivering a useful tool that makes the 9 to 5 a little bit easier should make anyone happy, be ready to share your excitement!

→ Offer prizes for participants who ask or answer questions during your training.

→ Sprinkle humor into your presentation and deck.

→ Take frequent breaks, encourage open dialogue and ask about current pain points. You’ll learn a lot about the team you’re working with, and create trust. 

→ Remember people want to be talked to, not at. 

5. Find your raving fans

As users piled into the ChatGPT platform and discovered how easy it was to use and how much fun it was, excitement grew and word spread quickly. In a sense, ChatGPT made mini-influencers out of all of us.  

We’ve found some of the best ways to drive adoption of our augmented analytics solution in various organizations is by spotlighting your most excited, most passionate users. This goes beyond your own role as an analytics leader, as you’ll need to find advocates on each team that you’re working with. 

→ These individuals talk the talk, understand the hurdles and headaches, and can help relay unvarnished feedback back to you.

→ Showcase the success stories of your users and the different ways they’re getting value from your analytics tools. When users can see their own peers succeeding, it makes it seem less intimidating and within their reach.

In Conclusion
In just 5 days, ChatGPT launched and broke records by acquiring a million users. Its overnight success can be attributed to a viral brand crush born of the generous sharing of “AI magic” for free. 

The good news is: brand crushes aren’t just reserved for leaps in AI technology. No matter the industry or the audience, leaders can fuel crushes for their own brands by embracing magic and generosity.

Magic can be as simple as a tool that works, and makes life easier. 

Generosity may not be in the price of the platform itself (free won’t work for most business models) but can come in the teamwork, relationships, and resources of a mutually beneficial partnership. A thoughtful, intentional roll-out will help teams focus on the immediate value your platform brings. The continued collaboration of an engaged team makes all the difference for success.

The success of ChatGPT serves as an inspiration for analyst leaders who are trying to create a similar following in their own organizations.

The post What Analytics Leaders Can Learn from ChatGPT first appeared on AnswerRocket.

From just a year ago, when I first attended the Summit, to this year, the BI and analytics space is moving at breakneck pace. Analytics vendors are rapidly enhancing their features and capabilities. At the same time, enterprise businesses are investing in new platforms and data processes at a faster pace than we’ve seen previously.

It’s also exciting to see is the advancing integration of data into everyday business operations. Gartner spent much of the summit focusing on how businesses can transform into agile powerhouses by leveraging their data, particularly though the role of the Chief Data Officer.

Now that I’ve had time to absorb everything I saw and heard at the Summit, Here are my three thoughts for how you can take advantage of your data to become a lean, agile, high P (& low L) operation:

1. Data has gravity – keep it where it is!
Here’s a paraphrase of a question I was asked frequently during the Summit:

“We just transitioned our data to [insert expensive and high-performing data warehousing solution here], do we have to move our data to work with you?”

During an organization’s foray into modern analytics, technical analysts inevitably take stock of the data they currently use and figure out how to get everyone governed access. Sound familiar to you?

If you’ve jumped into or completed a transition to modern data organization and away from spreadsheets and fragmented files to a data warehouse, that exercise isn’t in vain. Moving your data to a centralized repository where business users can access data and share the same data definitions is critical to making decisions fast.

The good news is that if you’re happy with your existing data warehouse solution, AnswerRocket can connect directly. You can keep your data in place. If you’re not ready to build your own data warehouse, we can help create one with you.

2. Unleash the bots!
Also heard at the Gartner Summit was an increasing reliance on machine learning algorithms and artificial intelligence to unearth valuable insights. During the Innovative BI in Action session, AnswerRocket presented a revolutionary move toward the next step of analytics, incorporating advanced search-based data discovery for descriptive analysis with auto-discovery analysis for prescriptive results. Behind the jargon, we combine advanced statistical analyses and a customized and unique understanding of your business to harness the power of advanced computing.

You aim in the right direction, we target and fire, returning the most relevant results. We see a future where business users utilize the computation power at their fingertips to multiplex their searches broadly and deeply within their data. With natural language processing (NLP) and natural language generation (NLG) working together, we help users ask questions and understand answers, becoming the oft-used term, citizen data scientist.

The crowd feedback to this presentation was especially gratifying. 61% of the attendees rated advanced search-based data discovery as high impact or transformational in the marketplace. And of the five presentations made, attendees rated our capabilities as tied for the most relevant to their business needs today.

3. Keep your data scientists focused on what’s important.
Business users have the tools they need to access answers to many of their data questions, without the intervention of a BI analyst or data scientist. While it may seem contradictory, now is when data scientists are needed more than ever.

With self-service analytics tools like AnswerRocket available to users today, the routine questions commonly answered by cut and paste SQL are now answered in seconds. This leaves the interesting and challenging business analytics to the data experts.

And by the way, these data experts can use AnswerRocket, too. We’ve got some new ways of visualizing data, like Sankey diagrams and waterfall charts. Our augmented analytics known as insights are a great way for discovering trends that might not be immediately evident. Data experts and business users alike can take advantage of built-in machine learning algorithms to apply advanced analysis to their questions.

Conclusion
This year’s Gartner Data & Analytics Summit was an amazing opportunity to see the rapid changes occurring in the market. If you’re looking for more information from Gartner, check out their 2017 Cool Vendors in Analytics Report.

If you weren’t able to attend and would like to see how AnswerRocket can help business users get quick answers to their questions, request a demo today.

The post 2017 Gartner Data & Analytics Summit: What Was Most Relevant? first appeared on AnswerRocket.