It’s Never Too Soon to Prepare Your 2014 Holiday Marketing Strategy

using your holiday data-01

The stress of the holiday season is behind you and it’s time to put your feet up and take a breather, right? Hardly. Successful ecommerce companies know that the next holiday season isn’t far away. Testing Adwords campaigns, preparing website updates, planning email campaigns, and procuring products takes an enormous amount of time. So grab your numbers from 2013 and get ready for a data-driven 2014.

Really? Do I have to?

Could you have been more prepared in 2013? We thought so. So get out that data. You probably have a whole lot of it and it comes with the added benefit that it was within a short time period – no Google algorithm changes or economic conditions to muddy your analysis.

Continue reading

Jelly Data: An Outside-Looking-In Analysis

Biz Stone’s new startup Jelly was launched with great fanfare one week ago today. Jelly is as ambitious as you would expect from a co-founder of Twitter: it aims to transform all smartphone-wielding humans into a single collective neurological entity.

Or, failing that, it’s a pretty fun question-and-answer app.

While I’m not qualified to assess Jelly’s ability to transform us into a benevolent overmind, I did manage to get my hands on quite a bit of data about how people are using the app. Since Jelly’s launch, I’ve been collecting data on all of the questions and answers that have been posted. If you’ve been curious what’s been going on, read on.


Continue reading

What’s a Holiday Shopper Worth?

Don't Let Them Walk Away

A holiday shopper is not your average customer. The rest of the year, you might be selling high-end kitchen supplies to self-taught chefs. Come November, your new customer could be a take-out addict – hard pressed to describe the difference between a braise and brine. Making the assumption that a holiday shopper will become a year-round customer sets ecommerce retailers up to overspend, underspend, or spend in the wrong places to acquire these shoppers. Before jumping in to capturing new holiday shoppers, make sure you first understand just what a holiday shopper is worth.

According to the NRF 36% of holiday shoppers decide where to shop based on discounts and sales. A separate study by Deloitte found that 71% of holiday shoppers expect free shipping and 47% expect free returns. Couple this with the fact that that advertising becomes more expensive and inboxes become more crowded, you are looking at substantially higher acquisition costs around the holidays.

Continue reading

E-Commerce Churn Rate

Some online businesses, like SaaS companies, spend a lot of time thinking about their churn rate. (This is typically the percentage of customers who unsubscribe from their service in a given month.) For subscription businesses, a low churn rate can be the difference between life and death.

Companies with non-subscription business models, however, typically don’t think about customer behavior in terms of churn. E-commerce players, for example, think about long-term customer relationships in terms of conversion rates and repeat purchase rates. There is rarely a clear distinction between a “current customer” and a “churned customer” in these analyses.

That is, until recently. Every month we’re seeing more and more e-commerce leaders adding “churn rate” into their RJMetrics dashboards.

How can an e-commerce business measure churn if customers never “unsubscribe?” There are a number of approaches that make sense, and typically they depend on how your company plans to act on the data.

One of the most popular methods is to set a cutoff date after which, if a customer has not made a purchase, they are considered to have “churned.” The choice of this date can be arbitrary, or it can be influenced by things like time between orders, repeat purchase rate, and cohort analysis of historical data.

Once you have defined populations of “active” and “churned” customers, you can pursue a number of new analyses and strategies. Here are some tactics we’ve seen:

  • Segment things like referral sources and product categories by percentage of customers churned. This can tell you more about where your most loyal customers come from and what they tend to buy, which can inform marketing and merchandising decisions. This is probably most valuable as part of a cohort analysis.
  • Identify populations of customers who are “about to churn” and send them special promotions or offers to encourage a purchase. While these groupings are definitionally as arbitrary as your churn threshold, instituting this practice on a regular basis will ensure that you are consistently reaching out to new populations of at-risk customers (since customers from previous batches who did not purchase will have churned and those who did purchase will no longer be at-risk).
  • Monitor the health of your business by tracking churn rate over time. Changes in the percentage of customers who move from “active” to “churned” in a given month can be an directional indicator of changes in customer loyalty or behavior.

Giving lapsed customers a definitive “churn” event changes the slow, unpredictable fade-away of an e-commerce customer into a sharp step function that can be monitored, quantified, and acted against. It can simplify goals and clarify vision.

For more on churn rate. check out To give this a try on your own data, you can try RJMetrics free for 30 days.


Lifetime Revenue Cohorts

There are a lot of different ways to look at your data in RJMetrics, and we know that interpretation and understanding are just as important as calculation and visualization. So, I’m writing a series of blog posts where I will do a deep dive into some of our analyses and visualizations.

The first in the series is the lifetime revenue cohort analysis.

What does lifetime revenue cohort analysis mean?

This chart shows the cumulative spending per user for a period of time after they are acquired. Cohorts of users are split up by their acquisition month.

For example, the orange line above shows the average for users who were acquired in November 2011. The first data point means that in their first month, users who were acquired in November spent an average of about $200. The second data point means that by the end of their second month, these users had spent an average of about $240. Their average spending in month two was approximately $40 (240 – 200).

The different lines represent different cohorts of users. The green represents the users that were acquired in December, and the blue is users that were acquired in October.

Why is this important?

This kind of cohort analysis can be useful for several different purposes, but the most immediate benefit is often better customer acquisition decisions.

Many companies limit their marketing spend to channels that yield profitability on a customer’s first purchase. These companies will pay to acquire customers through a given channel as long as that their average first purchase yields more gross margin than it costs to acquire them. The problem with this approach is that it often results in an underinvestment in growth. If your competitors are marketing based on a deeper understanding of buying behavior, they will outgrow you.

The lifetime revenue cohort analysis helps you to understand the consequences of expanding your customer acquisition spending, and it provides an easy way to convey this to the rest of your team. If future customers behave like existing customers, then acquiring customers for a higher CPA will result in a predictable payback period. Depending on the cash position of the business, you can define what payback period you are comfortable with, find the relevant spot on the chart, and spend accordingly.

Additionally, you can use this analysis to see if you are getting better at onboarding, engaging, and generating revenue from the users you acquire.  For example, this cohort analysis is a great way to see if a free shipping promotion for new users resulted in repeat buyers or one time purchasers that never come back.

How will this vary for different business models?

For most businesses, the lifetime revenue cohort analysis chart will show a large amount of spending in the initial period and then increase more slowly over time.

That initial spike is due to the fact that customers are more likely to make their first purchase soon after they are acquired than at any other time. In cases where the acquisition event itself is a purchase, 100% of customers make a purchase in their first period. In cases where registration can happen before purchases, this effect is less drastic. As an example, Groupon would likely have a much lower initial jump than Amazon, because many of the people who sign up for Groupon don’t make a purchase right away.

Unless there are a high number of refunds, this chart will slope up and to the right after the initial jump. The rate of growth tends to decrease over time because customers are usually most active when they first sign up. This causes the average to drop because the number of people in the cohort stays constant regardless of how many come back to buy more.

In subscription businesses, the slope will decay less aggressively than in businesses where people make one-off purchases. Occasionally, a subscription business will actually have a slope that increases over time. It is rare to see this, but it is a great signal for the business when it happens. This does not mean that there are zero churning customers, but rather that upgrades for customers that stay more than make up for the customers that leave.

How is this calculated?

There are two simple inputs to this calculation: how many members are in the cohort (which never changes), and how much revenue those members generated in the given period.

To determine the members in the cohort, we count the number of users who were acquired in the period in question. An acquisition can be a first purchase, account creation, newsletter sign up, or some other event.

The revenue calculation is a bit more complicated.  We want to sum revenue for orders that were placed by members of this cohort and took place within a fixed time period from their acquisition date (ie the first three months).

Finally, we divide the revenue by the number of members in the cohort for each time period in the chart and add this value cumulatively over time.

What are the variations of this chart?

There are many different kinds of useful cohort analyses.  The most common variation is filtering by user acquisition source. For example, you might want to look at this chart for customers who came from organic search, paid search, or an affiliate program.

This will help you understand if the customers from one acquisition source are more loyal or valuable than another. Thrillist’s subsidiary JackThreads used this analysis to understand that one of their most expensive acquisition sources was actually its most profitable. After learning this, Thrillist shifted their marketing budget to the more expensive acquisition source and accelerated their growth.

Another way to look at the data is with an incremental, rather than cumulative, data perspective.  This shows the incremental amount that an average user spends in each month after they are acquired.  This is useful for forecasting the amount of repeat purchases you will get from existing users.

We can look at this with other things besides revenue as well.  Some examples include margin as well as non financial metrics like invites, votes, or messages.


Lifetime spending cohorts are a powerful way of looking at at your customers’ buying behavior.  Stay tuned for more information on how to use and interpret your metrics.  

Business Intelligence vs. Business Analytics

When I was first hired at RJMetrics, I was a recent college grad, and the concepts of “business intelligence” and “business analytics” were still a bit new to me. I started reading up on the these fields and quickly learned that even practitioners who have been working with business data for decades can’t agree on concrete definitions.

As I waded through countless websites, forums and flame wars looking for the answers, I decided to help other people like me by condensing my findings into one convenient space. If you’re wondering about “ba vs bi” or “business analytics vs business intelligence,” hopefully this post saves you some time.

BA vs BI

When you’re talking about data and what it can do for your company, a lot of terms get thrown around. Business analytics (BA) and business intelligence (BI) are two terms heavily used, but rarely given the same definition by any two sources. Some take the stance that they’re interchangeable, and others staunchly defend their position as to the meaning of each, and what would fall under those respective umbrellas. Ultimately, I came to realize that these are two complementary concepts that have grown with, and out of, each other.

On a basic level, BI is the ability to take information resources and convert them into knowledge that is helpful in decision making. The traditional method of doing this involves cataloging and examining data from past decisions and actions, and using this as a way of setting metrics benchmarks for the future. In method, BA is an offshoot of BI. BA focuses on using data to net new insights, whereas traditional BI used a consistent, repeating set of metrics to steer future business strategies based on this historical data. If BI is the way to catalog the past, then BA could be called the way to deal with the present and predict the future.

Hosted business intelligence solutions like RJMetrics offer a combination of BI and BA by providing a data warehousing and reporting solution alongside a flexible interface for ad-hoc analysis and data discovery that can point you toward smarter decisions.

The Evolution of Business Intelligence vs Business Analytics

In the past, BI has been used to talk about the people, processes and applications used to access and extrapolate meaning from data, for the sake of improving decisions and understanding the effectiveness of targeted decisions. But this is where BI as a baseline failed; something that runs entirely off of static, historic data severely limits a user’s ability to make predictive decisions and forecast for the future market. When an emergent situation arises on a Friday afternoon, the user doesn’t greatly benefit from looking at metrics collected prior to the introduction of that situation.

The rapid growth and demand for BA comes from this failing, and is in a way the evolved form of BI solutions. In a business world whose speed is ever-increasing, the user needs to be able to interact with information at the speed of business, not looking back over his or her shoulder at what happened in the past. BI setups alone do not support the occurrence of users asking and answering questions in the face of marketplace events as they happen. A company that is data-driven sees their data as a resource, and uses it to hedge out competition. The more current the data the user has, the better jump he or she has on the competitor, who may or may not have become a threat in a time so recent that traditional BI data reporting wouldn’t even take them into consideration.

Many companies are commonly implementing advanced analytics on top of their data warehouses, to bridge the gap between BI and current day needs. Perhaps this is the origin of the confusion between terms, as organizations pick and choose from different combinations of services and have no real understanding of what to call these mashups.

Equally relevant is the fact that more and more people are being asked to interpret data in roles that are not strictly analytical. Product managers, marketers and researchers are moving towards data as a way to formulate strategies, and traditional BI platforms make it difficult to push data into real-time situations and what-if scenarios.  With the importance of data-driven decisions increasingly becoming a realization for less tech-savvy branches of company teams, the need for more user-friendly and faster producing platforms also grows. Moreover, delivering the data that supports these decisions to a broader company team demands a more visual form of modeling tool, to improve understanding across all departments. Charts and graphs showing BA findings are quicker and more impacting than written out statistics and excel sheets full of data.

Data interpretation and the manipulation method of choice change as the market demands. While having a set of established methods is important to the effectiveness of a company’s strategy, it’s understanding the need for flexibility in the face of these changes that can be a company’s most valuable asset.

If you’re looking for a robust business intelligence tool that empowers the use of business analytics within your organization, sign up today for a free trial of RJMetrics.

Cohort Analysis Example

Almost every company we work with is interested in running cohort analysis on their data. This comes as no surprise to us because cohort analysis is an extremely powerful tool with many potential applications.

Many companies understand that cohort analysis can be a valuable tool, but they ask an important, fundamental question:

How can I use cohort analysis to improve my business today?

Here, we step through an example of how a fictional company uses cohort analysis to make smarter business decisions. This example was compiled based on observations of how real companies are using the cohort analysis functionality in our online dashboards.

A Cohort Analysis Example

Let’s use our Vandelay Industries demo data set. There are a few important things to know about this business.

  • Ecommerce site
  • Actively spending on acquiring new users
  • Some users buy immediately upon signing up, others only after a while, and others not at all
  • Cost per acquisition for paid search is $80
  • All-in acquisition cost (including discounts and revenue splits) is $60 for a group buying site

Based on this information alone, it makes sense to double down on group buying, because it has a significantly lower cost. The missing information is what these users do after they are acquired. That’s where the cohort analysis comes in.

Below, we have two weekly cohort analysis charts. One is for customers acquired through paid search, and the other is for a group buying site. The average incremental revenue from customers acquired through group buying sites is significantly smaller than other customers. Once the data is presented in this way, it’s easy to see that our strategy of focusing on group buying rather than paid search is shortsighted.

Three months in, users acquired from paid search generate $140 and group buying generate $75, almost 50% less.

  • Paid search 3 month net revenue per user is $140 – $80 = $60
  • Group buying 3 month net revenue per user is $75 – $60 = $15

We could look at the data for different cohort time periods, or in different segments, but the underlying message is the same. Without the benefit of cohort analysis, the folks at Vandelay would be acquiring drastically less profitable customers.

For a real life example of how one of our customers uses this analysis to make smarter customer acquisition decisions, see the Jackthreads case study.

You can also test drive the Vandelay Industries demo to run this analysis yourself, or start your free trial of RJMetrics to run cohort anlayses on your business’s data.

Web Traffic Analytics Versus Database Analytics

People often ask about the difference between database analytics services like RJMetrics and web traffic analytics services like Google Analytics.

The difference is actually quite significant, so much so that RJMetrics and Google Analytics are actually quite complimentary products.  You can learn more by downloading the white paper below.

Click Here to Download the White Paper

You can also access your Google Analytics data from within our product by signing up for our free 30 Day Trial.