Posted June 1, 2012 by Veronika Sonsev
Jake Stein, Co-Founder of RJMetrics gives the SCA Accelerator insight onto the world of data & how it can be used to both optimize your customer’s experience & become more effective at acquiring new customers.
RJMetrics splits data into three categories to the software:
Category 1: Revenue
Category 2: Life-cycle analysis, life time value. Ex. cohort analysis, time between orders
Category 3: Segmentation. Any demographic/geographic factors or segmenting groups in any kind of pattern
The Social Aspect of Data
What does social have to do with data?
- Acquisition of new customers
- Retention of existing customers
- Referrals to friends & followers – get your customers to be your marketing force
Important to understand what the impact of social initiatives and how to compares to other places where you dedicate resources. Collect data and apply metrics in a diverse environment including social points in order to prioritize and how to understand which things work and which doesn’t.
How to Judge a Metric
A lot of people assume that getting more data is always better – thats not always true. There are an infinite number of ways to slice your data but have to allocate what your budget is and what is most effective.
ACTIONABLE vs Vanity Metrics – if the data makes you take some sort of action then its worthwhile, if it doesn’t then its probably only there to make you feel better. Knowing that you grew 20% over the last year is great but if you want to use that in order to acquire more customers its not very relevant.
What are you trying to ACCOMPLISH? – focus on the end goal. Ex. the total dollar volume for running business is not relevant for acquiring customers but extremely relevant if you’re trying to raise money .
Can you get the DATA? Understand the different level required to get useful data.
Why Should You Care?
Better Acquisition Decisions – the biggest improvement with customers. RJMetrics’ customers (more than 100 ecommerce sites) see 5x difference among best and worst channels. Customers are able to get five times as much revenue by shifting your spending from one advertising channel to another.
Case study: Thrillist/Jackthreads
Men’s apparel flash sale site. Very actively acquiring customers through diverse channels, both social and paid. Optimizing CPA and tracking where people came from when they signed up but had no insight on what their customer LTV was. Spending more of their budget on channels that had lowest CPA. With RJMetrics, they shifted focus on acquiring and more on retaining customers they already had.
Thrillist’s Question: What acquisition channel generates the most profitable LTV?
With Cohort analysis focusing on ‘user acquisition source’ Facebook vs Groupon they found:
- Facebook segment customers spent $100-150 in their first month but 6-12 months in customers were generating $200 per person.
- Groupon segment in their first month are generating $50-75 and 6-12 months in they still struggled to break the $100 mark.
We found some of the most expensive sources were most profitable so Thrillist/Jackthreads shifted budget and it has since been successful.
Case Study #2 – Totsy
Totsy used a Cohort analysis grouping by the dollar amount of customer’s first purchase. They had a lot of success with acquiring more users by offering a discount off their first purchase.
Totsy’s Question: What is the greatest LTV return we can get from offering a discount?
Case Study #3 – Fab.com
Fab.com has a model similar to flash sites with a very active community. A lot of their customer acquisition is based on referrals from their existing customers – users get a $10 off their order. Every user would be given a special link that would identify that they had referred a friends to the site.
Fab.com’s Question: how do you quantify your influencers? They had a lot of people referred from many different social networks and many level within those referrals, ex A referred B & B then referred C using the same special link. Fab.com if users they acquired form new sources were better or worse than the users they acquired from their old sources.
Fab.com wanted to look at the data from the per-user basis as well as through cohorts. RJMetrics pulled data based on the dollar value a customer spent, how many people they referred, as well as how much the people they referred spent as well. Using that data Fab.com was able to analyze which the complete value of a customer based on these factors, and could identify which customers to give special discounts to.
Tactis – Targeting Lapsed Customers
REMARKETING - retargeting you existing customers is a lot cheaper than acquiring new customer.
SEGMENTATION - Don’t send the same message to all of your customers all the time to avoid customer fatigue and unsubscribed. Tailor your offering so its more interesting, giving different offers to specific segments
REACTIVATION - send a targeted offer for your last customers to keep them engaged. Run tests to optimize performance.
How to segment your customer base?
- Those who haven’t bought anything in the past 90 days
- People who have been referred by a friend
Its not very important the first segment you do – keep it simple. Split customers by multiple segments and keep testing.