Brad's Books and Organizations

Books

Books

Organizations

Organizations

Hi, I’m Brad Feld, a managing director at the Foundry Group who lives in Boulder, Colorado. I invest in software and Internet companies around the US, run marathons and read a lot.

« swipe left for tags/categories

swipe right to go back »

Be a DataHero!

Comments (22)

Be a DataHero!

I’ve been fighting with creating charts and data visualizations – well – forever. Anyone remember VisiPlot and VisiTrend? Harvard Graphics? Eventually Excel dominated for a while, but it was always sheer misery for me. Eventually I figured out how to make rows and columns of data look like a chart in my mind, and I just stopped making charts for myself.

Last year we made a seed investment in a company called DataHero. We loved the founders and their vision to make it trivial to turn rows and columns of data into charts. They’ve created a magical product that just works and includes the concept of Live Charts. You simply connect to whatever data source you want, go through their hero-like wizard to set up the visualization you want, and your charts automatically update.

The data I want to chart lives all over the place. In Excel spreadsheets in Dropbox. In my Google Drive. In apps like SurveyMonkey and Salesforce. DataHero connects to each service and just does the right thing. No more exporting and importing data, reformatting it, and tearing your hair out.

Try DataHero. Tell me what you think.

 

  • http://www.cornfedsystems.com/ Frank W. Miller

    Just an initial thought. It might be useful to know what kinds of output you can get from the tool on their website, e.g. pdf, gif, png, eps, etc. and where the output goes, local, dropbox, etc.?

    • Chris Neumann

      Hi Frank,

      Thanks for the suggestion! DataHero charts can be exported as .png files in a variety of sizes and styling options. The resulting image will be downloaded locally or to a connected drive. We’ll put this on our FAQ page to make it more visible.

  • http://www.datadyne.org Joel Selanikio

    Brad, at DataDyne (www.datadyne.org) we develop the Magpi mobile data collection system (www.magpi.com), which is used by many international organizations (e.g. UNICEF, World Bank, WHO, Red Cross) to collect critical health, agricultural, environmental, and other data (I talked about it at TED.com here: http://www.ted.com/talks/joel_selanikio_the_surprising_seeds_of_a_big_data_revolution_in_healthcare.html ). Would love to talk to the guys at DataHero about connecting Magpi to their system — but couldn’t find any contact address. Can you help?

    • Chris Neumann

      Hi Joel,

      I’m one of the cofounders of DataHero. Please send me an email at chris (at) datahero (dot) com and we can connect.

      I look forward to speaking with you,

      - Chris

  • Rob Icopos

    I love the concept, but in playing with it found the product underwhelming and even potentially dangerous. It makes it easy to draw the wrong conclusions from datasets as it often ends up highlighting things which there’s low statistical power to support or where there are confounders that are the real cause.

    I’m unconvinced more pretty charts and graphs are the answer. Why not build a tool that allows users to do real analysis ?

    • Chris Neumann

      Hi Rob,

      Our goal at DataHero is to empower the many people who have access to data and have questions they want to ask of that data, but don’t know how to do real data analysis. There are countless tools that exist for data analysts and other quantitatively-minded people (from SAS to Excel and everything in between), but all of them require that the user knows the process of analyzing data. What about everyone else?

      This is absolutely a challenging problem and we’re a long way from the finish line, but I firmly believe that as our society increases its focus on data, we need to empower not just mathematically-inclined people, but everyone to be able to get value from data.

      If you’re up for it, I’d love to speak with you and hear your feedback on aspects that you found particularly underwhelming and/or dangerous. Please email me at chris (at) datahero (dot) com if you’d be willing to talk.

      Thanks!

  • http://byJess.net/ Jess Bachman

    I think there is a lot of over-promising in the data visualization space (I’m in that space!), and the word “insights” can be very dangerous. I haven’t used the “Data Decoder” but I think we are a long way away from truely surfacing important and relevant insights, automatically on unfamiliar data. It’s currently a struggle to surface decent insights on familiar data like google analytics and facebook insights, let alone whatever is in that csv file in your box account.

    Human Data Interfaces is an area with lots room for improvement and I hope DataHero the best, though to be honest, I’ve already seen some cringe worthy examples of poor viz practices in their videos. It’s a hard problem and even the pros are not hitting the head of the nail in most cases.

    • Chris Neumann

      Hi Jess,

      I’d love to hear your feedback with regards to “cringe worthy examples of poor viz practices” in our videos. Our goal is to empower and enable people, which means being accurate not only in our product but also in our communication. Would you be willing to share your thoughts with me on the topic? I can be reached at chris (at) datahero (dot) com.

      Thanks,

      - Chris

  • NickN

    I have to disagree with the naysayers. I think this is a very promising tool. And it’s not just about pretty charts.

    In a previous startup I was involved with, we spent a lot of time looking at data exploration. One of the concepts we talked about a lot was “data wandering”. Good data analysis has traditionally relied on understanding the data you are working with. As the volume of data we have access to increases exponentially, the percentage we can innately understand gets ever smaller. That means you have less and less clue about more and more data.

    So either we all become more useless, or we adopt a new model, and that’s what data wandering is about. You build a tool that lets you take a walk through the data in different ways and see what you can discover.

    Now of course, a pre-requisite for successful data wandering is that the user (and eventually the tool) can have some sense of when the visualization is reliable (and statistically valid). In the near term, that means data wandering is most useful for people who already have a good idea of what they are doing.

    Datahero looks like a promising first step down that path. I also think it does a far better job than Excel at getting to an attractive chart quickly. The import tools and data source support alone make this worth having. In the near term, I would imagine that most users will be creating straightforward charts quickly, and it looks great for that.

  • Kathy Gallup Keating

    I’ve worked at the intersection of Artificial Intelligence and complex Data Analytics for years. It’s fairly easy to let technology make actionable insights from data that is relatively straightforward (e.g. revenue growth, hashtag trending, fitness health scores, etc).

    But pulling actionable insights from data that we don’t yet understand ourselves is extremely challenging. We need to be answering questions like:
    - “who is committing fraud within my organization”
    - “who are the terrorists (and can we catch them before they do something)”
    - “why can’t I lose weight when I exercise and eat right”
    - “what is preventing my process from being optimal”

    We (as humans) don’t even yet know what indicators or trends contribute answering these questions fully. Learning how to leverage technology to these types of questions is what is going to propel us forward.

    I love that companies like DataHero are making it super-easy to visualize the answers to simple questions; if we can get to these simple actionable insights faster maybe we can start focusing on the real questions that stump us and make our eyes roll back in our heads. Speaking of which, I better get back to that since no technology is going to do it for me. :-)

  • http://www.knackhq.com/ Brandon Griggs

    Love the concept and the tech is impressive. At http://knackhq.com we’re about converting data into useful apps that can be shared and published. Have they done any partnerships to date?

    • Chris Neumann

      Hi Brandon,

      Thanks for the kind words!

      We have a number of partnerships with companies who generate data as part of the services they offer and who want to provide their users with an easier way to analyze and visualize that data. Last week, for example, we announced a partnership with SurveyMonkey that enables anyone who creates an online survey to be able to instantly visualize their responses in DataHero (their only such integration).

      Please send me an email at chris (at) datahero (dot) com and we can connect and discuss whether there might be a fit with knackhq.

  • http://dreamsandlogic.vom LeadDreamer

    Bravo, Brad & Chris Neumann! The openness and willingness to engage with skeptics and believers alike is the true value brought to the community here!

  • ObjectMethodology.com

    “Eventually I figured out how to make rows and columns of data look like a chart in my mind, and I just stopped making charts for myself.”
    .
    Oh my, technology is pretty bad when a person just gives up and visuallizes in their mind.
    .
    Charts using a heterogenous data store, nice!

  • Dennis Chandler

    I like the concept and it looks very promising (much easier to use than Tableau in my opinion.) My only issue is I don’t believe I would use it. If you know what you want, then this is great for that, but I deal mostly with visual exploration, and DataHero (nor anyone else) does that well in my opinion. Plus I abhor pie charts, and it that seems to be a pretty common default with DataHero. There definitely needs to be a way for anyone to visualize anything at anytime, but you may be trying to make it too simple. The best machine learning algorithm will (okay may) not replace a person with domain knowledge looking at the data.

    • ObjectMethodology.com

      If you are willing to explain what you want I’d be glad to hear it. Visit my website and contact me. I don’t think I should put my email in a post. Brad probably won’t care but it might automatically get trapped as spam.

      • D Chandler

        Thanks for the offer ObjectMethodology.com, but I am fine with my current solutions. I currently use R and I teach others how to manipulate pivot tables in Excel (which I update on a monthly basis). My lament is that I have to teach people as there is nothing out there for people to just start using. Foe example right now I need to look at our June sales to see why gross margin dropped. I’m confident that it was product mix/quantity, but I need to ‘show’ that. I have all the data, but I need to generate 1-2 charts of temporary KPI’s on the subject so we can monitor over the next few months. I’ll need to create some new variables from the data and then visually map them. I can do this in R, but it would be nice if I could delegate this to someone else using a service that could walk them through the process. It’s not cut and paste and requires some knowledge of our product and our customers.

        Anyway, this is an opportunity for DataHero to get feedback on their product, so discussions should center around them.

        • ObjectMethodology.com

          That’s cool. I just wasn’t sure if I should invite what could maybe become an off-topic discussion to be handled here. So your problem is one of having a knowledgable minion?

          • D Chandler

            Don’t need a knowledge minion if the tools exist where they can take the information and easily derive insight from them. Isn’t that the point of DataHero and the rest of the players in this area; to allow anyone with any skill level to gain insight from their data, easily and quickly? I just don’t feel like basic charts accomplish this. You need the ability to flow from one visual to another based on questions and exploration.

          • ObjectMethogology.com

            Awesome! This little requirements interview was great. Thank you.

  • Taylor Flynn

    Reminds me a lot of datapine – they are doing drag&drop sql queries

  • Nikki Braziel

    I think the website is lovely — clean and modern — and I like the animations. I’d rather see a Tardis than an orange phonebooth, but I know what that means for licensing… :-)

    My “wish” going into this trial was to be able to upload a multi-tabbed excel sheet (my pro forma) and have DataHero pull out interesting graph and chart possibilities that I hadn’t thought of. Unfortunately, my data was formatted in a way that all the output charts were titled “by” and were illegible. Even after spending 20 minutes trying to make a single tab usable, I wasn’t able to output any useful charts.

    This highlights how important formatting the data with intention/clarity is to begin with. Multiple levels of row/column headers are really problematic, and blank rows used for spacing have to be cleaned up. One day, I’d love to find a smart chart tool that can identify chart possibilities even within a poorly formatted spreadsheet. In the interim, a tutorial on DataHero on how to massage your data before upload might be useful. Thanks, and good luck!

Build something great with me