Today, we’re proud to announce a new partner on our journey: Trinity Ventures. Trinity has led a $6.25M Series A investment in RJMetrics and Karan Mehandru has joined our Board of Directors. We’ve known Trinity and Karan for quite some time, and we couldn’t be more excited about adding their passion, vision, and expertise to the RJMetrics team.

You can read more about the financing on TechCrunch, Technically Philly, and PRWeb, so I’ll spare you the boilerplate here. Instead, I would like to take a minute to talk about what this news means for you, our customers.

Ever since our bootstrapped beginnings, customers have been the lifeblood of RJMetrics. This customer-centric philosophy is wired into our DNA, and it will continue to permeate our work as we put this new investment to use.

In just the past few months, we’ve rolled out a new dashboard user experience, a revamped chart builder, new ways to get your data into and out of our data warehouse, and a world-class customer success team.

This pace of innovation, which was bolstered by a small seed round last year, will now accelerate even further. Stay tuned for game-changing enhancements and, as always, please don’t be shy if you have feature suggestions.

Thank you so much for your continued support as we move into this exciting new chapter.

That’s because the new interface makes use of Google’s AngularJS. With Angular, we’re going to be able to roll out new features and iterate on existing ones faster than ever.

Here’s a snapshot of what you have access to in your account today.

Adding a chart

We’re working hard to consolidate these options, but everything related to adding a chart lives in the “Add Chart” dropdown.

Settings Navigation

The settings navigation has been split up into two groups: Data, and Account Settings. The Data group contains items that have to do with manipulating your data. The Account Settings group contains items that have do with managing your account.

Spotlight Search

Have an idea what your chart or metric is called and can’t remember where to find it? Type your keyword into the search box and we’ll take you right to it.

Upcoming Features

Here’s a quick look at what’s on the near-term horizon…

Chart Resizing

Break free from the confines of dashboard columns! You’ll be able to resize your charts on much more flexible grid to emphasize the most important charts and tables on the dashboard.

Chart Descriptions

You’ll be able to view a metrics definition from anywhere. Can’t remember what column your “Revenue” metric uses? Forget which filters are applied for “Users we count?” Now, they’ll be no more reason to lose your spot and jump to another page to answer this question. That information will live everywhere you’re using your metrics.

Plus, you’ll soon be able to enter your own custom description for a chart, a heavily requested feature.

Darkened Full-Screen Mode

Many of our customers display their dashboard on a large monitor. We’re working on a feature to automatically switch from a light theme to a dark theme for better visibility at large sizes.

We couldn’t be more excited for the pipeline of features to come. Stay tuned.

The Projects

• An inside look at what we eat in the company kitchen using RJMetrics dashboards to track snack and beverage metrics.
• An automated gong-ringer that will sound every time a new client signs up.
  

  Rohan and Ben test their project.

• A new RJMetrics data connector, deployed on our Data Import API. It sucks down data from our marketing automation platform, Pardot, and updates RJMetrics dashboards on an hourly basis.
• A dashboard that updates in real time.
• A wiki search function for our newly upgraded company wiki on Github
• A hack of our Keurig machine, which now refills on its own.
  

  Jake and Shaun are blown away by the endless coffee.

• A prototype of a pivot tables, bringing all the wonders of Excel pivot tables to your RJMetrics data.
• An A/B testing tool for sales conversations.
• A new UI which allows us to enable or disable features for different users

The Results

Congrats to Bob Moore, Connor McArthur, and Cathy Lennon who took home first place for their real-time dashboard project! Second place went to Matt Monihan, Nate Vecchiarelli, and Buck Ryan for their pivot tables prototype. And third place went to Shaun McAvinney and Jake Stein for their A/B testing tool.

Turns out hackathons are a lot of fun. Maybe quarterly isn’t enough…!

This is exciting for several reasons:

1. We now have a functional, close-to-release version of the Data Import API. This API will become the new standard for how data gets pushed to RJMetrics.
2. The API is dead simple to integrate with. I am not a “real” developer and I was able to build an incredibly useful connector in 1 day using Ruby and deploying on a completely free Heroku instance.
3. The dashboards that I use every day are now auto-updated without me having to lift a finger. This will save me several hours every week.

The Data Import API

We’ve kept our Data Import API a state secret to this point so as not to incite the mob—a lot of you have been waiting for this for a long time. But we’re close to launch, so I figured I’d let the cat out of the bag. Here’s the lowdown.

The RJMetrics Data Import API is an incredibly straightforward way to import data into your RJMetrics account. In fact, that’s all it does. Send it data, and it saves that data to a table in your account. Just make sure your data includes an ID field and is formatted as JSON and you’re good to go. Once you push data, it will become available for you to use in metrics and dashboards.

This means you can now integrate anything with RJMetrics. Customer support platform? Check. Inventory management platform? Check. CRM? Check. Custom application developed by Fred from your IT department? Check. Every aspect of your business can now be seamlessly integrated into your cloud business intelligence solution.

Building Your Connector

The logic of a connector is incredibly easy:

• Identify a data source
• Select all the rows
• Push them to the Data Import API

The wonderful thing about this logic is that you don’t have to track what rows are new, old, updated, etc. Just push all the rows across. Our warehouse will match records by ID and perform inserts and updates as appropriate. Much less for you to worry about.

If you have many millions of rows in a given table, you may want to include logic to only push certain of those rows. The dataset that I was using was small enough that this wasn’t a concern.

Copy My Connector

If you develop in Ruby, you’re more than welcome to use my code as a starting point for developing your own connector. I wrote a (very) simple API wrapper that is bundled as a gem for easy inclusion. Include it in your gemfile by adding:

gem 'ruby_rjmetrics', :git => 'https://github.com/jthandy/ruby_rjmetrics.git'

Don’t feel like you have to use the gem, though. If you want to see how the API call is made, it’s really just 1 line:

The connector itself is a simple Rails app with no persistence layer, deployed on Heroku. There is a single controller for each type of data that’s responsible for getting data. The controllers are then called as rake tasks from the command line, allowing them to be easily integrated into cron (or the Heroku scheduler).

Again, feel free to steal all of the code, or use it as a jumping-off point. It’s quick and dirty, but it does the trick. Get it here.

If I can do it, you can too

I’m a marketer. I write some code every once in a while, but our developers here would never let me within a mile of the RJMetrics code base. So, when I say I wrote a connector, that means that you probably can too.

We’re now accepting applicants to participate in the closed beta of the Data Import API. If you’d like to participate, just contact support.

In this blog post, I’m going to share some critical segments that we often recommend to our customers. I’ll also go into detail on what questions these segments can help you answer.

User Segments

User segments help you understand who your users are and how they behave.

• Age / Birth Year: How old are your users? How old are your most active users? It usually makes sense to bucket the values into ranges for more effective analysis.
• Gender: Do men and women engage with your website differently?
• Address: Where do your users come from? Should you focus your marketing efforts on a particular region? Have your recent advertising campaigns performed as expected in your target regions?
• Customer acquisition source: Do you know what marketing channel your users come from? Did they click on an ad or find you via search? Segmenting your data by user acquisition source is the first step in optimizing your new customer acquisition. Step two is to spend more money in what’s working and kill what’s not.
• Registration device: Did users register via your mobile app or your website? iOS or Android? Is your mobile user base big enough to allocate more resources to develop your mobile product?
• Referred by: Who are your top influencers? How many users were directly referred by others?
• Industry: If you’re a B2B business, in which industries do your users work? Which trade organizations are worth joining?
• Survey responses: If you perform customer surveys, use the responses as segments for a deeper level of profiling. You can ask questions that complement what you already know about your users or confirm your guesses.
• First order amount and product category: Is there a correlation between a user’s first order and future purchasing patterns?

Orders / Events Segments

Order and event segments help with analyzing user behavior and engagement over time.

• Billing / Shipping Address: Where do most of your orders come from? Is there a difference between billing and shipping addresses?
• Status: How many of your orders failed to complete? What is the ratio of pending orders in the past 7 days?
• Customer acquisition source: Beyond tracking user acquisition data at a user level, you can also track the it on an order or event level. A user that registered via one source may very well continue to access your site via other sources.
• Device: Are the number of mobile orders increasing? How much of your revenue is currently generated via mobile purchases?
• Fulfillment Center: Which one of your fulfillment centers is generating the most revenue? If you’re analyzing the difference between order time and shipping time, which fulfillment center is most responsive?
• Delivery Carrier: Which is the most popular carrier? Which carrier has the least number of returned items?
• Discount / Coupon Codes: Are your promotions actually generating extra business? How many extra items did your customers buy in addition to the item on sale? How do coupons affect your average order value? What’s your average margin on discounted vs. non-discounted items?
• Satisfaction / Rating: How satisfied are your customers with their orders? Are your customers likely to refer business to you?

Product Segments

Product segments help you make merchandising decisions.

• Merchant / Brand: Is one specific brand selling faster than the rest? Which brands are under-performing?
• Type / Category: Do different user segments enjoy different types of products? Which product categories generate the most repeat business?
• Discount / Coupon Codes: Are promotions hurting sales of non-discounted products? How do coupons affect the perceived value of your products?
• Social activity: Is there a correlation between the buzz generated on social media and the quantity sold for a product?
• Size / Variant: What is the ratio of inventory that you need of each variant? Which variants can be sold at discount rates?

If you’re interested in merchandising, check out a blog post where I explore how to use product segments to drive repeat business.

Establish Customer Profiles

Segmentation experts may want to move beyond one-dimensional slices and begin establishing real customer profiles. For example, people between ages of 13 and 24 that registered via a mobile device put in a group “Young & Mobile”. How does this group’s behavior compare to the rest of your user base?

This type of analysis is what marketers at Fortune 1000 companies do all day. Prior to the advent of cloud-based business intelligence platforms like RJMetrics, it was largely out of reach for the rest of us. Fortunately, that’s no longer the case.

If you’re an existing customer and want to improve your segmentation, just contact our support team. If you’re not a customer but want to learn how to slice and dice your data like an Iron Chef, we’d be happy to tell you more.

We’re really excited about getting the Data Auditor into your hands. From the very earliest days of RJMetrics, we’ve had questions from customers that would go something like this: “How did you get $xxx,xxx for your revenue number? We show $yyy,yyy in our database. Are you sure you’re pulling the right numbers?”

This is, of course, a very reasonable question. There’s not a lot of point in having a business intelligence tool if it’s not giving you accurate numbers. But what we found was that the problem was (almost) never one of accuracy, but rather one of data definitions. Different users would query their data differently, and this would cause confusion when they tried to validate what they saw in RJMetrics. After investigating for a while, we’d ask the customer whether or not they’d remembered to exclude sales tax from their revenue calculations. There would be an “Aha!” moment, and everyone would go home happy.

The problem? This process took a long time, both for us and for you. We’d prefer that you have a dead simple way to verify the accuracy of a given metric and to cross-reference it against your database.

And now you do.

The Data Auditor gives you everything you need to verify the accuracy of your data for every metric in your dashboard. Just choose a metric and a date to sample on, and we’ll show you the relevant rows from our warehouse as well as a query to run on your transactional database to replicate the results. If you’d like a full run-down, check out the help site.

Happy auditing!

The new chart editor is now live for every customer and trial user, and we’ve put together a quick walkthrough on our help site. Take a look and tell us what you think!

We loaded data from Twitter’s API into RJMetrics and here’s what we found:

• Overall, video creation is still an extremely underdeveloped market.  Only about 4% of highly active users shared a video through Vine or a top competitor during Vine’s first month on the market.  In that same period, 98% of the same group shared at least one photo through a leading photo sharing service.
• In its first month, Vine steadily gained market penetration to 2.8% of Twitter’s highly active users, blowing past competitors Viddy and Socialcam, which were used by 0.5% and 0.2% of the same population, respectively.
• Twitter’s built-in tools for photo and video sharing are dominating the competition.  Vine.co and pic.twitter.com are the most popular tools in their respective categories by a wide margin.

Vine Adoption

Vine showed impressively stable adoption growth over the course of its first month.  We were expecting to see a spike in adoption around the time of the announcement followed by a leveling-off period, but instead the percent of new users each day has remained consistent.  This is a good sign for future growth because the rate of adoption does not appear to be slowing as time goes by.

Vine vs. Viddy vs. SocialCam

In Vine’s first month, the percent of highly active users who used Vine was meaningfully higher than the percent who used competitors Viddy and SocialCam.

We were concerned that this might not be an apples-to-apples comparison since many users may have just been “trying out” Vine in this month.  As a check, we looked at the average number of times each of these tools was used during the month.  As it turned out, repeat usage of Vine was actually more likely than the other apps.

Video vs Photo

While Vine’s performance is impressive relative to its competitors, it’s still a tiny player in the universe of media sharing on Twitter.  We looked at the percentage of users that linked to various media sharing services and found that photos represent the vast majority of the links sent out by highly active users.

As you can see in the chart above, native Twitter photo hosting (pic.twitter.com) is the dominant player, followed by Instagram and then a number of less prominent competitors.

When you remove Twitter and Instagram, you can see just how small a player Vine is when it comes to sharing media.

About The Data

We decided to sample from Twitter’s most active users to find early-adopter activity.  Twitter’s API was used to identify and download the twitter streams of about 2,500 randomly-selected “highly active” users, each of whom had tweeted at least 100 times so far in 2013.

The result was 2.3 million tweets that were sent between January 24th and February 24th.  320,000 of these tweets contained links, which we followed through any link shorteners to find their final destinations.

The data was then loaded into RJMetrics, where we generated this analysis with just a few clicks.

Conclusions

Twitter’s efforts to add native photo and video sharing into its service are proving fruitful.  These tools have quickly become the most popular options for end users, causing a major impact on the market for 3rd party apps.

Vine appears to be establishing itself as the de facto tool for short video creation and sharing.  However, the significance of this move will only be felt as its market matures.  Today, Vine is a service only used by a small minority of even the most highly active users.

The best part of our job at RJMetrics is the chance to help new entrepreneurs succeed. That’s why we’re so excited to announce our partnership with 21212, Brazil’s premier digital startup accelerator. Through this partnership, every company in 21212 will have access to the same level of business intelligence that the industry’s leading companies use to power their decisions.

From the 21212 blog:

After completing our first two acceleration programs, and now in our third, we have come to realize the importance of data analysis in the entrepreneurial process. Starting a business is all about asking, and answering, a growing list of questions. What do my customers want from me? How much are they willing to pay? Where can I find them? Answering these questions successfully determines whether a fledgling venture will succeed or fail. And the only way to come up with answers is to dig into your data.

We couldn’t have said it better ourselves. Good luck to everyone in 21212′s third class, and we’re looking forward to helping you grow.

Here are four of my favorite examples (from our team and others) of lean startups in action.

1. Faking A Move

To me, being lean is all about minimizing the ratio of resources consumed to insights gained. At RJMetrics, we did this to great effect when we were hiring our first employees. At that time, our company was based in Camden, NJ, and new applicants were few and far between. We suspected that our location was the reason for the talent shortage, but we had no way to prove that without making a really expensive bet and moving the business to nearby Philadelphia.

Just then, my co-founder Jake had a great idea: let’s just say we moved. We published an identical job posting but changed our address to a location in Downtown Philadelphia. The applicants started pouring in. We simultaneously started interviewing candidates and looking for new office space. We made our move literally one day before our first new employee started work.

Our Awesome Philly Office

I love this story for two reasons. First, had job applicants not increased we would have saved ourselves the trouble of moving with effectively zero downside. We then could have searched for more fundamental reasons that we weren’t getting applicants. Second, it saved us tons of time. By getting the answer to our question up-front, we were able to move and recruit in parallel rather than in sequence. I have no doubt that this accelerated our growth trajectory by months.

2. Actually Talking to Users

In the age of cookies, cheap storage, and abundant APIs, many companies have developed a strong bias against actually talking to their users.  Instead, they attempt to infer intent and sentiment from user actions.

While data can do a great job of telling us what is happening, it can often fall short on why it is happening.  Relying solely on data can also mask client frustrations.  Instead, from time to time companies should rely on a more time-honored tactic: asking what customers think.

Customer interviews are inexpensive, fast, and remove a lot of the interpretation risk associated with data-only strategies.  No matter how many numbers we crunch, we always learn something new when we ask our customers about their most and least favorite parts of our product.

3. Naming a Book

Tim Ferriss has a huge bag of tricks, but there is one that always sticks with me. When Ferriss was writing his book “The Four Hour Work Week,” he and his publishers were considering a number of potential titles. Rather than go with his gut, Ferriss devised a simple and brilliant strategy: buy Google Ads for each of the different potential titles, have them display when people were searching for content related to the book, and see what got people to click the most.

Users who clicked the ads would land on a blank page or “under construction” page.  All Tim needed to know was what moved people to click.

This inexpensive experiment allowed Ferriss to create an ad-hoc focus group consisting of thousands of ad viewers and learn which titles piqued their interest the most. Based on sales of The Four Hour Work Week, it’s quite clear that he made the right choice.   Ferriss shares this story in the video below.

4. Interactive Mock-Ups

Philadelphia entrepreneur Chris Cera recently turned me onto a really neat tool he uses called Axure. With Axure, you can to mock up interactive user interfaces without any backend coding. In other words, as long as testers follow pre-determined steps, they can get the impression that they’re interacting with a fully-built product.

Anyone who has done user experience work can appreciate why this is a brilliant and valuable tactic. An interactive mock UI/UX can answer huge up-front questions about how and when potential users would derive value from a product (or experience frustrations). With these questions answered, the development of the backend can be much more focused and deliberate, and there is far less risk associated with the final product’s release.