I was reading an interesting article on PandoDaily that inspired some tangents as I’ve been giving these metrics (Cost Per Acquisition, Customer Lifetime Value, etc.) a ton of thought lately.
Instead of posting a really long comment on the thread (I actually did that already but opted to delete it now that I have this), I thought I’d make use of my new blog and post my response to the article’s author, Jordan Elkind.
After all, that’s the whole point of this, right? Right. (Don’t you love answering your own questions?)
Here’s a link to the original article: “Why you’re not spending enough to acquire customers”
And here’s my response…
This is solid info for entrepreneurs who are not so acquainted with these concepts (and maybe even some who are but lack clarity). Thanks for putting this together and sharing your thoughts.
I do, however, think you’re missing some key analysis that is crucial to decision-making as you attempt to establish the actual CLV (Customer Lifetime Value), optimal CPA (Cost Per Acquisition) in relation to CLV, and what kind of ROI (Return on Investment) to expect from different targeting campaigns based off of these conclusions. I realize essays upon essays and books upon books can (and are) written on the subject, but I thought I’d chime in as I’ve been giving this a lot of thought on my own lately…
1) How do you determine what a “lifetime” is for your specific demographic and product?
This answer doesn’t seem to be determined by one simple equation and from my understanding, varies greatly on the product, market, target customer, etc.
To illustrate, let’s think about a scenario that could definitely take place in today’s market: a vertical ecommerce company selling baby clothing. This baby clothing company recently launched a line of products targeting from 0-3 months, 3-6 months, 6-9, 9-12, 12-15, 15-18, capped out at 18 months. If you’re on your toes, then you might immediately conclude that the CLV should be calculated on an 18 month scale.
Sure that might be the case in a perfect world, but that doesn’t necessarily compute here. Consider the fact that it’s commonplace to have another baby 1-3 years after your 1st child; if you liked the clothing company the first time, you’d probably buy clothes from them again with your next child, right? Likewise, if you’re anything like my mother or sister, you’ll probably be recommending to all your friends that they should also purchase their child’s clothing from this amazing brand (assuming the experience warrants). The brand might even have a referral marketing incentive (which makes their CPA a little harder to calculate, but more on that in point #2). Again, these are simply examples of minute factors that come into play when you delve deeper into any specific business.
Given all of this, how should one go about determining what the realistic lifetime is for this type of customer when you have to calculate in the unknown factors of potential future children, the value that they may bring from word of mouth referrals, and the age of their child when they 1st purchase from you?
This brings me to my next point, which is more about questioning the Initial CPA as being the only metric that should be highlighted in terms of weighing CPA to CLV, and whether these metrics are reliable as the only (or main) data we should be looking at here…
2) By all calculations then, it would make the most sense to spend the entire customer acquisition budget on acquiring customers who are in the 0-3 month range, as you would have an ability to sell them a product every 3 months and the earlier you acquire, the more chances you have to sell them product. A customer that you acquire at 0-3 month stage might be worth 6x more than the customer who has a 15-18 month child. But take into account that you’d have to spend a ton of money on making sure that the customer from 0-3 month is not only converted to a customer, but also retained across that entire period, and you have an entirely different story.
So then I ask, naturally, “What about the CPRCOL?”.
What about the CPRCOL you might say? Yes, I went overboard with the acronym. But you guys, I’m seriously… what about the Cost Per Retaining Customer Over Lifetime?
As stated, you may acquire the customer initially by spending that $40 (the example of CPA provided in the article) and you may justify this by saying that your CLV is $50, so you achieved your goal of 25% ROI.
But, in order to realize the full potential of CLV, you need to retain that specific customer beyond their 1st purchase, not only acquire them initially. Lots and lots of money is spent on retention marketing, specifically on campaigns such as referral credit, discounts, email marketing, funnel/retention analysis, the list goes on. Long story short, there is a huge market out there for retention marketing and it ain’t nuthin’ to scoff at.
In my opinion, the cost to retain that customer over the predetermined lifetime period is therefore arguably as important as the initial CPA and should be calculated into the overall cost that goes into not only acquiring, but keeping the user. So in theory, the ratio that should be optimized is the CPA + CPRCOL :: CLV, not simply the CPA :: CLV.
Of course, this is just touching the surface. When you get into calculations such as Churn, Retention, etc. you’ll be able to generate #s that are much closer to reality than simplified CPA/CLV.
Oddly enough (but not surprisingly), Wikipedia has compiled an interesting resource on the topic that goes into more detail (e.g. calculate retention as 1 – churn): Customer Lifetime Value.
Went a bit long, but hope that added something!