Data Products

The Future Belongs to Purpose-Built Apps. We're Betting On It.

“Purpose-built apps”

“Low-code app development”

“hpaPaaS”

“Citizen Data Scientists”

“Data monetization”

Witness the cloud of new buzzwords floating in the air. Let me see if I can knit these concepts together to shed light on their meaning and implications for the future of analytics.

Collectively, these phrases are a reaction to the long-standing challenge of getting more data into more hands. “Democratization of data” can seem perpetually right around the corner (if you’re listening to vendor marketing) or a distant illusion (if you are in most organizations).

At Juice we have a picture that we call ‘The Downhill of Uselessness’. It shows how the usefulness of data seems to decline as you try to reach more users. On the far left, the most sophisticated data analysts and data scientists are happily extracting value from your data. But as you extend to the outer edges of your organization, data becomes distracting noise, TPS reports, and little-used business intelligence tools.

downhill-of-uselessness.png

Three barriers to democratizing data

The struggle of getting data to more people in more useful ways boils down to a few unsolved problems.

First, general purpose platforms and tools (data lakes, enterprise data warehouses, Tableau) can be a foundation, but they don’t deliver end-user solutions.

"Vendors and often analysts express the idea that you can master big data through one approach. They claim if you just use Hadoop or Splunk or SAP HANA or Pervasive Rush Analyzer, you can “solve” your big data problem. This is not the case.”

— Dan Woods, Why Purpose Built Applications Are the Key to Big Data Success

Second, reporting and dashboards deliver information, but often lack impact. In our experience, most data delivery mechanisms lack: 1) a point of view as to what is important; 2) an ability to link data insights to actions in a users’ workflow.

Third, the people who truly understand the problems that need to be solved don't have the technical capacity to craft re-usable solutions. We all have that elaborate spreadsheet that is indispensable to running your business and, frighteningly, only understood by a single person.

A better path forward

Finally, there is a realization that these problems aren’t going away. There needs to be better approach. It will come in two parts:

  1. Focus on creating targeted solutions (applications) that solve specific problems. Apps can integrate into how people work and the systems where actions occur. They attempt to let people solve a problem rather than simply highlighting a problem. And applications are better than general purpose tools because they can bake in complex business rules, context, and data structures that are unique to a given domain.

  2. Give greater impact and influence to the people best know the problems. It has always been unfair to ask technologists to create solutions for domains that they don’t deeply understand.

This direction aligns with Thomas Davenport’s view of Analytics 3.0 (from way back in 2013). He postulated that the next generation of analytics would be driven by purposeful data products designed by the teams who understand customers and business problems. (No offense, Tom, but we were griping about ivory tower analytics back in 2007.)

And so emerges a new model and new collection of buzzwords...

Purpose-Built Applications

Solutions that start with the problem and craft an impactful answer. Their success is measured by fixing a problem rather than in terabytes of data stored.

…built using a high-productivity Application Platform as a Service (hpaPaaS)

Cloud-based development environments requiring little coding ability (‘low-code’) — but requiring knowledge about the domain and the problem to be solved.

…to be used by Citizen Data Scientists (CDS).

the people who know the problems most intimately.

At Juice, we may have backed into this trend or cleverly anticipated it. Either way, now I can say that Juicebox is a low-code hpaPaaS designed for CDS to create purpose-built apps. Better yet, we are now fully buzzword compliant.

Is It Time to Jump-Start Your Data Offense?

jack-dempsey-i-was-a-pretty-good-fighter-but-quote-on-storemypic-7ceab.png

Legendary Alabama coach Bear Bryant believed in defense:

“Offense sells tickets, but defense wins championships.”

Legendary boxer Jack Dempsey saw virtue in offense:

"The best defense is a good offense.”

Legendary analytics guru Thomas Davenport takes a more neutral stance in his Harvard Business Review article What’s your Data Strategy?

"The key is to balance offense and defense.”

Davenport goes on to say:

“Data defense is about minimizing downside risk, including ensuring compliance with regulations, using analytics to detect and limit fraud, …and ensuring the integrity of data flowing through a company’s internal systems.

...Data offense focuses on supporting business objectives such as increasing revenue, profitability, and customer satisfaction.

…The challenge for CDOs and the rest of the C-suite is to establish the appropriate trade-offs between defense and offense and to ensure the best balance in support of the company’s overall strategy.”

Balance is fine. But at Juice, we’re all about building data products. That’s an offensive data strategy (we’re with you Jack Dempsey, June Jones, Mike Leach, and Mike D’Antoni).

In practice, most organizations start from a defensive crouch. The relevant question is: when is it important that you shift to a more offensive data strategy?

Davenport shares a few indicators that suggest more data offense is warranted. For example, offensive strategies are often employed at organizations that operating in largely unregulated industry where customer analytics can differentiate. He also sees opportunity for offensive data strategies at that those organizations with decentralized IT environments and where “Multiple Versions of the Truth” are encouraged.

His HBR article even provides an evaluation tool to determine whether your organization has shifted its balance toward offense or defense, giving you a snapshot of where you’ve (organically) evolved. It doesn’t tell you where you should be.

When we think about the dozens of companies we’ve worked with who are launching data products, some common patterns emerge in terms of the characteristics of those organizations. Here are four categories where an offensive data strategy provides like a good fit:

Government, non-profit or public-service organizations

These organizations aren’t necessarily in the “competitive” markets that Davenport describes. Nevertheless, they are sitting on tons of valuable data that can shape conversations and influence the decisions of their constituents. We’ve worked with Chambers of Commerce, Universities, and State Departments of Education that are taking on offensive data strategies.

Data science startups

There are hundreds of start-ups who are building their businesses on offensive data strategies. These companies have mechanisms for collecting data across an industry and are adding value through predictive algorithms, identifying patterns, and ultimately helping their customers make smarter decisions. We’ve working with a couple healthcare start-ups who have proprietary methods for predicting performance of healthcare providers. This is deeply valuable information for health systems and employers, and a purely offensive strategy.

Consultants

We’ve seen a couple different offensive data strategies by consulting firms. First, if they are delivering a project with an analytical deliverable, why not make the deliverable a recurring data solution? Another approach by the most innovative consultants is to view data collection and data products as an opportunity to proactively identify problems for clients. An annual survey of customer brand awareness can be turned into an incisive discussion starter, spurring clients to pursue the next project.

Companies with dominant market shares

If you are a market leader, you may be collecting enough data from your customers to be able to provide benchmarking solutions. In some cases, this offensive strategy is core to the original purpose of the business (e.g. US News & World Report’s surveying of colleges). In other cases, the opportunity to create new data products can be a result of “data exhaust”.

If you find yourself wondering how your data might be turning into a revenue-generating or customer-differentiating solution, you should download our ebook Data Is the Bacon of Business: Lessons on Launching Data Products.

Is Your Data Product Ready for Launch?

Looking to transform your data into a valuable, customer-facing data product?

From concept to design and launch, we've worked with dozens of companies to create successful data products. Our checklist provides seven evaluation criteria to see if your data product has what it takes to succeed.

Does your data product...

  1. Solve a distinct problem?

  2. Meet users where they work?

  3. Guide users to insights and actions?

  4. Make users feel safe and in control?

  5. Bring credibility to your data?

  6. Have the ability to operationalize the solution?

  7. Support customers for success?

Download the PDF here.

Data Monetization Workshop 2018: Key Themes & Takeaways

“Data Monetization is a hot topic because it has two words that everyone loves. We all love data, and who doesn’t want to monetize something?”

These were the words that kicked off the 2018 Data Monetization Workshop to a roomful of attendees and industry experts who had gathered to discuss the question that followed this observation: what does Data Monetization actually mean?

zach presenting.jpg

 

This question was discussed at length over the course of the half-day event and was the impetus for speaker topics related to using data for social good, how to account for data on a balance sheet, how AI will affect the future of Data Monetization, and more. Here are some of the most important themes and takeaways from the discussions of the day.

What Is Data Monetization?

  • Data monetization is about data value, not data dollars. It’s not about selling customer lists, but about deriving value.

  • Data Monetization encompasses business intelligence and takes a much broader perspective on what can be done with data. Analyzing what options exist outside the enterprise, what products and services can be created using data, and trying to get data into the hands of decision-makers are all components of Data Monetization.

Data for Good

  • Most organizations aren’t trying to sell your personal data; they’re focused on using information to improve city performance, prevent mass shootings, and rescue people from sex trafficking.

  • Nobody owns data. Companies and organizations have rights to data, but in order for progress to be made data must be shared and communicated.

The Dark Side of Data

  • While data offers many beneficial opportunities, there also exists a dark side of data. What complications does something like what happened with Cambridge Analytica have on future opportunities for Data Monetization?

  • Using certain data is not always a question of “Is this legal?” but rather “Is this ethical?” Sometimes data is available but not right to use, which can feel like a restraint at times but leads to being an organization being perceived as trustworthy. It is important to have a solid core philosophy on what data you do and don’t use before it becomes necessary to bring in lawyers and PR teams.

Education, Train, Explain - Data Literacy

  • Poor data literacy is seen across the board. If you don’t read the fine print, you can sign your data rights away. Many problems with the use of personal data are often due to mismatched expectations.

  • People don’t always understand how valuable data is and what an asset they hold. You have to teach people to think in technicolor. Some companies try to exclude information, but more information changes the landscape and provides more context.

  • Creating data products with different derivations is one way to communicate data to different roles (e.g., an analyst versus a CEO). You have to meet people where they are.

  • Being transparent with a product roadmap is a great way to demonstrate to people that data products will look different as time goes on. Users can know what features they can expect and when.

Doing Things Differently and Looking to the Future

  • There are emerging technologies that can help make processes easier. Right now you just have to ask yourself, “How can I do things a little bit differently today?”

Doug Laney Is One Cool Dude

  • Doug Laney was kind enough to join us remotely from his vacation to answer audience questions about his book Infonomics -- of which every audience member got a free copy!

Special thanks to all of the speakers, to MapR for sponsoring the post-workshop networking reception, and to everyone that attended! If you have questions or comments about the Data Monetization Workshop, feel free to reach out to info@juiceanalytics.com.

Related Reading:

4 Steps to getting started with data products

Over the years, we’ve had the pleasure to work with many great individuals and companies and through our work have gained the ability to sympathize with their experiences of what we like to call “going from 0 to 100."

No, we’re not endorsing excessive speeding in your car. We’re talking about going from having nothing but hopes and dreams about delivering engaging analytics (0) to having an interactive data story that your users don’t want to put down (100).

Because we’ve focused our efforts on taking clients from 0 to 100, commonalities or trends for best practices in the data and design experience (read: everything between 1 and 99) have become increasingly clear. Use these four tips to make your introduction to data products a better, more frictionless experience.

1. Know your audience

  • What do the end users you have in mind for the product look like? What questions will users ask of the data? What actions will they take with the answers to these questions? These are all things you should know before beginning to work on data products.
  • Be specific about for whom you are creating a data product. If you try to provide insights for too many types of business roles you run the risk of making it too broad for any role to gather insights from the data.

2. Gather the right data

When putting together the data to be used in your product, it’s important to discern the difference between “more data” and “more records."

  • More data: It’s not always in your best interest to gather the most “data” possible. By doing this, you run the risk of gathering data that you may not use and wasting money in the process.

  • More Records: Gathering “more records” (read: rows of data) is a better strategy as you prepare for your data product. Doing so can alleviate the effects of outliers and unearth trends in the data.

3. If you’re new to the data, begin with an MVP (Minimum Viable Product) and let your users determine what features should be included

Building out all the bells and whistles you think you might need at the beginning the data product’s life can be expensive. Starting with an MVP that is put in the hands of actual end users will help determine what data is actually needed and what design aspects are best for your purposes.

  • Helps with data: Starting with an MVP helps determine the shape and caveats that exist within your data, and allows your users to make decisions about what data is most important to them.

  • Helps with design: By starting with an MVP, all of the questions that you and your users have for the data are answered by the design. Additional features can then be added from that point on in a more cost-effective manner.

4. Be open-minded about visualizations

  • We won’t get into data visualization principles in this section because that warrants a totally separate article, but a simple point here: just because you saw similar data in a pie chart once doesn’t mean that is the only (or best) way to visualize your data.

  • Because your users are the ultimate consumers of the data, let them be the judges of what visualizations will be most effective for them.

Easy peasy, right? We think so, but maybe that’s only because we’ve helped so many customers get from 0 to 100. If you're still not sure what your next steps should be, we’re here to help. Learn more about our 0 to 100 process by checking out the document below.

6 Cool Companies Who Are Rethinking How We Work

It can be a challenging climb to reshape how people think about solving problems. We encounter this challenge daily as we work to build the best solution for communicating data the world has ever seen. We operate in an arena where good-enough solutions — Excel, PowerPoint, and other visual analytics tools — have left people with deeply-rooted habits and a blasé acceptance of the status quo. That’s not good enough for us, and it isn’t good enough for these six companies that are rethinking how business tools should work:

Slack

slack.png

Slack is the current king-of-the-hill for shaking up the status quo. Sure, we had email, file sharing, and messaging apps before Slack, but we didn’t have single, elegant tool for team collaboration.

What’s cool about it?
Slack made integrations easy from the start. We use everything from ChatOps with our development team to HeyTaco for everyday appreciation of our colleagues. Slack's approach to 'channels' found the right balance for open communication by topic.

Flourish

flourish.png

I only recently stumbled across the excellent visualizations available through Flourish. There are many, many tools for putting data visualizations on a screen; few vendors are so obviously passionate about their craft. 

What’s cool about it?
Flourish is more than another charting library — they are making world-class visualizations accessible. I was particularly impressed by the clever use of animation in those visualizations. At Juice, we appreciate that new users won’t always be able to read a visualization without some guidance. Animation can help draw a user’s attention to the most important information right off the bat.

Kialo

kialo.png

Kialo is “a debate platform powered by reason.” It cuts through the noise of social and online media by removing the worst parts of debating online (trolls, fake statistics, unrelated cat gifs) while strengthening the best.

What’s cool about it?
Kialo creates a structured dialogue with visualization, voting, and commenting. Whether discussing politics or the merits of a new project, Kialo has focused on an overlooked need: a place other than the comments section to examine arguments and consider new viewpoints. 

Typeform

typeform.png

Typeform is "the versatile data collection tool for professionals." It's a thoroughly modern survey-building solution that I’ve enjoyed using for over a year.

What’s cool about it?
Typeform's survey-authoring interface is remarkably intuitive. Adding questions, structuring logical flows, and navigating your survey is silky smooth. Similarly, the end-user experience is beautifully designed with selectors and animations that make it (almost) fun to fill out a survey.

Beautiful AI

beautifulai.png

From the founder of SlideRocket, Beautiful AI is a next-generation solution for creating web-based presentations. They say all you have to do is "think of an idea, choose a template, and get to work."

What’s cool about it?
Beautiful AI has taken a giant leap past a tool like Google Slides. It comes with a collection of smart slide layout templates. Better yet, these slide layouts automatically update as you add more content. The tool also comes with an easy-to-use integration with third-party image libraries so you can incorporate pictures into your presentation.

Toucan Toco

toucantoco.png

A data storytelling solution to build data apps for your business. We may be a bit biased, but that sounds awesome.

What’s cool about it?
While I don’t have hands-on experience with this solution, I love their message. Like Juice, they see the need to:

  • Communicate data to non-analysts with guided narratives ("Address the remaining 99% of your employees”)

  • Create targeted applications that solve specific business problems (“A business need = an app”)

  • Include simple, clear data visualizations ("The comfort of using consumer apps, finally in a business setting”)

Honorable mentions

  • Quid: Quid puts the world’s information at your fingertips, drawing connections between big ideas.

  • Skuid: Accelerate deployment of personalized applications that let your business people drive innovation, without the wait.

  • Trifacta: Trifacta enables anyone to more efficiently explore and prepare the diverse data.

  • Datawrapper: Datawrapper makes it easy to create beautiful charts.

Thinking about changing the way you work? Check out our app trial process. Download the info sheet below to learn more.

You Don't Need a Slide Factory

You might be surprised to learn that one of our most popular blog posts of all time is Automated PowerPoint Generation, or Making a “Slide Factory.” Even though this post was published almost nine years ago, month after month we continue to see it rise to the top of our most visited pages. 

Whenever someone reaches out to us asking if we have a ‘Slide Factory’ solution, we tell them two things:

  1. Sorry, we do not.

  2. You don’t actually need a slide factory.

In fact, the need for automated presentation delivery is the genesis of our data storytelling solution, Juicebox. We are intimately familiar with the need to deliver data to customers, co-workers, stakeholders, etc., in a consistent, structured manner that communicates a message while providing each person with the data that is most relevant to him or her. Instead of attaching a 50-slide PowerPoint deck, Juicebox does that same job with an interactive reporting application. Users benefit from a guided analytical story, ability to capture insights, and features to collaborate with others.

Not only do your users benefit, but you no longer have to deal with report production and ad hoc headaches! Juicebox was designed with the data consumer in mind, meaning that the need to spoon feed your audience data and information through long-drawn-out PowerPoint slide decks is no more. 

Screen Shot 2018-01-03 at 5.50.06 PM.png

In recent months, we have made it more affordable and simpler than ever to get started with Juicebox. Through our Guided Design Process, customers are seeing what their data looks like in a Juicebox application within days not months. We give you four weeks and ten user accounts to test Juicebox with your data, you have plenty of time to get user feedback and build a business case for using Juicebox. Pricing starts at only $6 per user (with a 50 user minimum). With tiered discounts for more than 500 users, Juicebox is a competitive option for any budget!

If you would like to test drive your data in Juicebox, fill out our Get Started form and we will be in touch ASAP.

Check out some of our Juicebox apps in action:

The Rise of Analytical Apps — Are We Seeing the Last Days of Dashboards and Reports?

dinosaur_comic.png

66,038,000 years ago, a massive asteroid smashed into the earth in what is now Mexico's Yucatan Peninsula. After this massive collision, it took only 33,000 years before the dinosaurs were entirely extinct — a blink of an eye in terms of the history of the earth.

This asteroid is considered to be the "final blow" after a series of ecosystem changes (other asteroids, volcanos, etc.) created a fragile environment for the poor dinosaurs. The climate changed, the dinosaurs died out, and the mammals took over.

Incumbent solutions for delivering data —dashboard and reporting tools— are facing their own "fragile environment." The big asteroid may not have hit yet, but it is only a matter of time. Here's why.

Exhibit A:

A thoughtful answer from an experienced Tableau user to the question “Why do people still use Tableau?”

We need to consider why (and when) people might stop using Tableau. My opinion is that Tableau has failed to realise two important things about their software and that if another company can solve this problem then Tableau could really lose out:

1. Companies need to create applications, not just reports

Yes, Tableau is interactive but you cannot use Tableau to make applications that write back to a database. It has maps, yes.. But you cannot use Tableau as the basis for an app like you might with MapBox (which has multiple SDKs for different platforms) or Leaflet.js for instance. Tableau is not designed for this, so if you need apps and not reports then it is not for you. You need a developer (or dev team) instead. This is a big gap in the product that other companies are also failing to see.

2. Tableau’s software does not directly generate revenue for (the majority) of their users

For a company to run several copies of Tableau desktop costs several thousand pounds. This is without the additional costs of Tableau Server or end-user licenses that you will need if you want your customers to use your hosted visualisations and dashboards. Any business that chooses to use Tableau to deliver interactive reports to its customers would need to consider passing some of that cost (or all of it) onto its end users. But when we’re talking about interactive reports, not applications, it is hard to justify data reporting as a stand-alone or additional cost.

That’s a real user wondering whether the paradigm of visual analytics tools for analysts, dashboards for executives, and reports delivered to customers and stakeholders is going to hold up for much longer.

Exhibit B:

Analytics vendors and market analysts are using language that leans more toward delivering "apps." 

 Alteryx

Alteryx

 PwC analytical app marketplace

PwC analytical app marketplace

 Infor

Infor

 Gartner's IT Glossary

Gartner's IT Glossary

 IBM Cognos

IBM Cognos

Is “app” more than a rebranding of a decade of data visualization tools? We think so. Here’s why we see analytical apps are on the way to taking over the BI world:

1. Apps have a purpose. A report or dashboard may carry a title, but it is less common that they have a clear and specific purpose. A well-conceived analytical app knows the problem it is trying to solve and what data is necessary to solve it. In this way they are similar to the apps on your phone — they solve a problem the same way a mapping app shows you how to get to the Chuck E. Cheese and a weather app lets you know if you need to bring an umbrella.

2. Apps make data exploration easy. I’ve spent a decade railing against poorly designed dashboards that put the burden on users to find where to start, how to traverse the data, and what actions to take. Good analytics apps willingly carry that burden. Whether we call it “data storytelling,” narrative flow, or quality user experience design, the app should deliver a useful path through the data to make smart decisions.

3. Apps are collaborative. Most business decisions are made as a group. If that weren’t the case, you’d have a lot fewer meetings on your calendar. Why should data-driven decisions be any different? Historically, reports and dashboards treat data delivery as a broadcast medium — a one-way flow of information to a broad audience. But that’s just the start: the recipients need to explore, understand, and find and share insights. They should bring their own context to a discussion, then decisions should be made. Our belief is that data analysis should be more social than solitary. It is at the heart of the “discussions" feature built into our data storytelling platform, Juicebox.

4. Apps lead to action. "What would you do if you knew that information?” That’s the question we ask again and again in working with companies that want to make data useful. Understanding the connection between data and action creates a higher expectation of your data. Analytical apps connect the dots from data to exploration to insight to action.

5. Apps are personalized and role-specific. The attitude of "one size fits all" is typically applied when creating a dashboard or report, and then it is up to individuals to find their own meaning. Analytical apps strive to deliver the right information for each person. How? By utilizing permissions for a user to only see certain data, automatically saving views of the data, and presenting content relevant to the user’s role.

The mammals took over because conditions changed, and the outdated species — with its size and sharp teeth — couldn’t adapt. Expectations are changing the analytics world. Consumers of data want an experience like they enjoy on their mobile devices. They don’t have the attention to pour over a bulky, unfocused spreadsheet, and they expect the ability collaborate with their remote peers. The climate has changed, and so too must our approach to delivering data.

If you’re still churning out reports, we can help you do better. Or if you’ve constructed a one-page dashboard, we can show you a different approach. Drop us a line at info@juiceanalytics.com or send us a message using the form below.

Why We Prototype

At Juice, we’ve spent the last year relentlessly pushing to make it easier to build world-class interactive analytical applications, or "data stories.” This was an important change for us. In the past, like a design agency, we would create carefully-crafted user interface mock-ups with detailed descriptions of functionality and interactions. Anything we couldn’t show in a static picture we would describe in words. Now we can do something massively more effective: we can build a live, interactive prototype in the time it takes us to draw all those pictures.

Here are the most important reasons we felt it was necessary to be able to prototype with ease:

1. Non-designers don't speak the language of mock-ups

With a decade of experience designing analytical interfaces, we became adept at making the mental leap between a static mock-up and the live application it would become. Static mock-ups imply — but don’t show — interaction points. They suggest what the data may look like, but don’t try to accurately show the data. They highlight dynamic content, but can’t show it change.

Take the following visualization mock-up as an example. Can you tell:

  • How the orange button will change as you interact with the visualization?
  • What happens when you roll over the points?
  • Why the title indicates “4 categories”?
  • The image implies a lot of functionality to an experienced information design audience. That doesn’t help everyone else.
why we prototype image one.png

2. Uncover data difficulties early

Your data isn’t always what you think it is. It certainly isn’t as clean or complete as you might hope. By prototyping with real data, you discover some of the issues in your data that run counter to your assumptions. You may also find trends or patterns that reshape what information you want to show.

Recently we built an application for a client that delivers an assessment checklist. We expected that we’d be able to look at the average scores to see how well students were performing. But in reality, students didn’t need to submit their scores until they were complete (100%). As a result, all the scores were perfect. And perfectly lacking in insight.

Here are just a few of the common things we run into when we prototype with real data:

  • Missing values where data should be
  • Multiple date fields, sometimes with confusing meanings
  • Averages that need to be weighted
  • Unexpected behaviors captured in the data that create unexpected data results

3. Validate hypotheses about the story you want to tell

Designs are based on a lot of assumptions about users. How will users interact with the data? What data is important to them? What views will be most impactful?

Prototypes give us the opportunity to test these hypotheses.  We utilize a user experience tool called FullStory to see in detail how users interact with their data story. We can see where they get confused and where they focus their attention. We also ask pointed questions to ensure our assumptions are playing out as we expected.

Screen Shot 2017-10-11 at 11.43.53 AM.png

4. Gather user feedback to sand-off the rough edges

User feedback isn’t only helpful for the big things. It can help you understand whether you’re on track the small, but important, details. A great data application needs to communicate the meaning of the content, including everything from the metrics to the labels to the descriptive notes. A few things to look for:

  • Do users understand the meaning of the metrics accurately?
  • Do the descriptions and labels convey the right meaning?
  • Is the styling — color, contrast — work for users or is it distracting?
Screen Shot 2017-10-11 at 11.45.10 AM.png

5. Build buy-in and a bandwagon

Making the transition from standard reporting to an interactive data application can be a big step for some organizations. For example, it can be scary to imagine giving your customers the ability to explore data by themselves. What will they find?

Taking this big step sometimes requires baby steps. Prototyping is an easy baby step. If you can create a real, working version of a solution to put in front of senior leadership, it will go a long way towards helping them get on board. Now people don’t need to envision what is possible, they can see it.

Screen Shot 2017-10-11 at 11.46.18 AM.png

Interested in building a prototype with your data? Get started by sending us a message!

Q&A with Treasure Data: Everything You Ever Wanted to Know about Data Viz and Juice

This post originally appeared on the Treasure Data blog. 

Tell us at the story behind Juice Analytics. What’s your mission?

My brother and I started Juice Analytics over a decade ago. From the beginning, our mission has been to help people communicate more effectively with data. We saw the same problem then that still exists today: organizations can’t bridge the “last mile” of data. They have valuable data at their fingertips but struggle to package and present that data in ways that everyday decision makers can act on it. Even with the emergence of visual analytics tools, data still remains the domain of a small group of specialized analysts leaving a lot of untapped value.

Our company has worked with dozens of companies, from media (Cablevision, U.S. News & World Report) to healthcare (Aetna, United Healthcare), to help them build analytical tools that make it easy and intuitive to explore data. We published a popular book in 2014 titled Data Fluency: Empowering Your Organization with Effective Data Communication (Wiley) with a framework and guidance to enable better data communication. To bring our best practices and technology to a broad audience, we built a SaaS platform called Juicebox that enables any organization with data to create an interactive and visual data storytelling application.

Why is data visualization so important to an organization’s ability to understand its data?

Data visualization is one of the most useful tools in bridging the gap between an organization’s valuable data and the minds of decision makers. For most people, it is difficult to extract insights or find patterns from raw data. When we tap into the power of visuals to help us recognize patterns, data becomes more accessible to a broader audience.

For many of the organizations we work with, data visualization has the added value of uncovering issues with the data. Once you start visualizing trends and outliers, the weaknesses or mistaken assumptions about your data come to the surface.

What is data storytelling? How can it be useful to marketing professionals?

The term data storytelling has become increasingly popular over the last few years. We know that data is important to reflect reality — but absorbing data, even in the form of dashboards or data visualizations, can still feel like eating your vegetables. We all recognize the power of storytelling to engage an audience and help them remember important messages. People who focus on communicating data — like our team at Juice — feel that there is an opportunity to use some of the elements of storytelling to carry the message. Stories have a narrative flow and cohesiveness that distinguishes them from most data presentations.

However, data storytelling is different from standard storytelling in some important ways. For one thing, in a data story the reader is encouraged to discover insights that matter to them. One analogy I like to use is a “guided safari.” Data storytelling should take the audience to the views of data where new insights are likely to occur, but it is up to the audience to “take a picture” of what is more relevant to them.

In our experience, data storytelling is particularly valuable to marketing professionals. For internal audiences, data storytelling techniques can help you explain the impact of your marketing efforts to your stakeholders. For customer or prospects, data stories can be used to lend credibility to your marketing messages and enable deeper insights of your product.

What are essential tools for data storytelling?

The tools for data storytelling fall into a couple of categories: human skills and technology solutions.

The most critical skill you can have for data storytelling is empathy for your audience. You want to know where they are coming from, what they care about, how data can influence their decisions, and what actions they would take based on the right data. Knowing your audience allows you to shape a story that emphasizes the most important data and leads them to conclusions that will help them. Data storytellers must remember that an audience has a scarcity of attention and a need for the most relevant information.

At Juice, we’ve thought a lot about the capabilities that make data storytelling most effective — after all, we’ve created a technology solution that lets people build interactive data stories. Here are six features that we consider most crucial:

  1. Human-friendly visualizations. Your audience should be able to understand your data presentation the first time they see it.
  2. Combine text and visuals. There are lots of tools for creating graphs and charts. But data stories are a combination of data visuals flowing together with thoughtful prose and carefully-constructed explanations.
  3. Narrative flow. The text and visuals should carry your audience from a starting point (often the big picture of a situation) to the insights or outcomes that will influence decisions.
  4. Connected stories. In many cases, it takes more than one data story to paint the whole picture. Think of presenting your data as a, “Choose Your Own Adventure” book, in which the audience can pick a path at the end of each section to follow their interests.
  5. Saving your place. The bigger and more flexible a data story becomes, the more important it is to let the audience save the point they’ve arrived at in their exploration journey.
  6. Sharing and collaboration. Data stories are often a social exercise with many people in an organization trying to find the source of a problem and decide what they should do about it. Therefore, it is critical to let people share their insights, discuss what they’ve found, and decide on actions together.

Where do you see organizations struggling the most with managing and understanding the data they collect? What should they be doing differently?

A common problem is that organizations don’t truly understand the data they are collecting. Ideally, data is truth— it should allow us to capture and save the reality of historical events, such as customer interactions and transactions. However, more often than not, what the data is capturing isn’t exactly what people imagine. We find it useful when we can get a data expert in the same room as the business folks who will be using the data. A deep dive discussion about the meaning of individual data fields will often reveal mistaken assumptions or gaps in understanding. Working together to build a data dictionary can be invaluable as you continue to use data.

Data exploration is an iterative process. Answering one question will raise a few more. In this way, organizations will eventually identify where they lack understanding of their data. The faster you can iterate on analyzing and presenting data, the sooner you will resolve the issues.

Is all data visualization created equal? What do organizations need to know about finding the right type of visualization to help better understand their story?

Not all data visualization is created equal. There are visualization approaches — charts and graphs — that could be a good fit for your data and message and there are poor data visualization choices that will obscure your data. One mistake that we see is an ambivalence toward finding the right chart for the job. You may have seen dashboards that default to show data as a bar chart, but also give users the ability to pick a variety of other charts types. Why not choose the best chart to convey your data and unburden users from making any more decisions?

There are also well executed and poorly executed data visualizations. Good data visualization emphasizes the data over unnecessary styling, clearly labels the content and directs attention to the most important parts of the data.

From where you sit, how should organizations approach their data management – from collection to storing to analyzing?

We start from the end, then work our way backward. One of the biggest mistakes we see is organizations trying to collect and consolidate all the data they may possibly need in one place. These types of data warehouse projects quickly spin out of control with endless requirements and increasing complexity. It doesn’t have to be that way. Instead, we’d encourage people to start with three simple questions:

  1. What important action do we take today that could be better informed by data? Only include high impact actions where you have the data to answer the question.
  2. How would we present that data to the people who make take those actions? Most of the time it isn’t a data analyst who is going to be acting on the data on a day-to-day basis. Consider the simplest possible view of the data that would enable the end users.
  3. What data is necessary to deliver that view? Now you’ve narrowed down to just the critical data that is going to make an impact.

Once you’ve answered these questions for one specific action, you can go back and do it again for another.

What trends or innovations in Big Data are you following today?

Here are a few of the areas that are interesting to us:

  • Data narratives. Companies like Narrative Science are turning data into textual summaries. Like us, they are looking for ways to transform complex data into a form that is readable to humans.
  • The intersection of enterprise collaboration (e.g. Slack), data communication (e.g. Juice), and business workflows (e.g. Salesforce). Our goal isn’t just to help visualize data more effectively. We want people to act on that data. To do so, data visualization needs to connect to places where people are having conversations and into systems where people make business decisions.
  • Specialized analytical tools. The pendulum appears to be swinging away from do-it-all business intelligence platforms and toward best-of-class, modular solutions. Companies like Looker, Alteryx and Juice aren’t trying to be everything to everyone — but rather serve a specific portion of the data analysis value chain. We’ve found more and more companies that are looking for the best tools for the job, but require mobility of the data between these tools.

Do you have a question about data viz, data storytelling, or Juice that we didn't answer? Send us a message at info@juiceanalytics.com or fill out the form below.