By Julie Perry, Boardable
At a lunch with fellow tech marketers recently, we all collectively groaned at the mention of those three dreaded words: Customer Acquisition Cost (CAC). One person bemoaned the fact that their non-marketer CEO didn’t understand how much information is lost when they only attribute a conversion to the last marketing touch a customer had before converting. Another was grateful for leadership who “gets it.” Others at the table are still at a loss for what their company should be doing to most accurately attribute the money spent on different marketing channels across the customer journey, otherwise known as the CAC per marketing channel.
The most common formula for determining Customer Acquisition Costs over a period of time is:
CAC = Total marketing + sales costs/Total number of customers acquired
More simply, it’s the amount your company spends to convince a customer to make a purchase. And when looking at the full picture of how prospects moved between sales and marketing channels before converting, analyzing CAC per marketing channel based on “first touch” attribution (how the customer first came in contact with your brand) vs. “last touch” attribution (the last marketing channel used before converting) will produce very different results. So which attribution model is better?
First, let’s review a few things about the process of exploring the best CAC analysis and conversion attribution model for your business. It is critical that you take a hard look at your customer culture and how that impacts the amount of time it takes a customer to go through your sales and marketing funnel. At Boardable, we are a SaaS solution for nonprofits, structured as a monthly or annual subscription that starts with a two-week free trial period. We work with nonprofit boards of directors, which means spending decisions are typically made at a meeting once a month. Strict budgets may be set for a year at a time. As a result, the journey from first touch to board consideration to purchase can be months long.
The longer your sales cycle is, the more misleading it is to attribute that last-touch marketing channel in your CAC analysis. Who cares if the prospect clicked on a Facebook ad right before submitting credit card information if that person was already prepared to buy your product after reading a blog post discovered by way of a search engine, downloading an eBook from that page, and then reading five emails in a drip campaign? Do we really think it was that Facebook ad that sealed the deal after six months of automated marketing touches?
That said, using a “last-touch” attribution model can be helpful for seeing what material causes a prospect to finally commit to buying. If you have a certain marketing email that has a high conversion rate, that might be your cue to take a closer look at the wording and timing of it to see if there is something unique about it that is prompting more conversions. Try to see if you can create a cohort that has a similar history with your marketing campaigns and look at where they convert. Otherwise, whatever preceded the last touch could be influencing the conversion.
Looking at the first touch of customers who converted is probably the best way to see where you should go fishing. Obviously, if you notice that out of your first-touch leads from a LinkedIn ad campaign, a very small percentage end up converting to customers, perhaps that isn’t your best audience or your best platform. A word of caution: make sure you’re accounting for how you follow up with those leads, as well as over what time period. Again, the same warning applies from the last-touch analysis—ideally you would have a clean A/B test of leads from various channels getting very similar follow-up, in order to see where your highest converting and lowest cost first-touch leads come from.
Another great tool to glance at from time to time is the Multi-Channel Funnels Model Comparison Tool in Google Analytics. While the tool is limited in its ability to accurately track across channels and devices, the ability to compare a first-touch vs. a last-touch vs. a linear-touch attribution model within the tool is often very insightful. As Google states, “in the Linear attribution model, each touchpoint in the conversion path shares equal credit for the sale.” Distributing credit across various marketing channels is, in my opinion, one of the best ways to most accurately assess marketing spend and decide whether or not a channel or campaign was “worth it” in the acquisition costs to win a customer. (Note that Google also offers several other multi-channel attribution models in the tool, such as Last Google Ads Click, Last Non-Direct Click, and another cool perspective, a Time Decay attribution model that gives touchpoints closest in time to the sale or conversion the most credit. Check it out in Google Analytics under the Conversions > Model Comparison Tool tab.)
You may be noticing a trend here: There is no “silver bullet” answer for CAC per marketing channel analysis. The honest answer is that we are constantly trying to balance the data of first touch with last touch (not to mention everything in between!). Admittedly, this is incredibly complex and labor-intensive to track accurately. At Boardable, we use Hubspot and Google Analytics, combined with some data analysis tools we’ve built in-house, to get the clearest picture possible of the full marketing journey a customer takes. It isn’t a perfect picture, and it is important to remember that your goal is just to get are getting as close to the truth as you can. There will always be a bit of imperfect data, and your business environment is always changing. In short, you are never done determining your CAC per marketing channel model. It is a moving target that you will need to refocus on from time to time.
Fortunately, with better analytics tools, we can automate some of these calculations. One technique that has been a game changer at Boardable is taking Google’s linear-attribution model one step further by using machine learning to weigh our marketing methods. Our machine learning specialist has set up algorithms that can look at the entire marketing touch-point history of our customers and weigh their experiences with us against the likelihood of a conversion. For example, if a large percentage of our customer conversions came from the first touch of a PPC campaign on Google, even if the last touch were via organic search, the PPC campaign would get a higher attribution score. In this way, we are able to assess in aggregate what behaviors and marketing touches are likely to contribute to a conversion. This is incredibly useful data for budgeting. However, it doesn’t answer questions like what specific content resonates, what features of the software sealed the deal, or what alternatives might have worked just as well. Again, it isn’t a perfect picture, but it gets us a lot closer to “one source of truth” on CAC.
That’s all well and good for the marketing department, but that really just gets the lead from “stranger” to free trial. The sale isn’t made yet. How do we study the cost of all the steps a SaaS takes to convert a trial to a customer? For us, that may involve additional helpful content, a human sales touch, emails, instructional videos, and so on. This is another area where we have started to lean on machine learning. As a product-led growth company, we are very interested in understanding how potential customers behave during their free trial. With our machine learning specialist, we have established a “product qualified lead” (PQL) score. We are able to see what behaviors during the free trial period are likely to lead to a conversion. Did the lead log in during the first 24 hours of the trial? Did he or she create a meeting? Has the account been dormant for more than two days? All of these behaviors tell us important information about a lead’s likelihood of converting.
Using the PQL score, we can determine what steps are needed for various activity levels. Maybe a lead with a low PQL score needs a check-in from the sales team. Or it could turn out that people who have planned a meeting but not added other users would benefit from a tutorial video on adding new users. We can then factor the costs of these interventions into our CAC determination. After all, the marketing costs from lead to trial are generally the longest and most expensive portion of the journey, but they aren’t the entire journey.
As our customer behavior data from both marketing channels and our product itself improves, the CAC measurements will no doubt shift. Don’t make the mistake of thinking that you will be able to nail down the perfect CAC per marketing channel depiction and just stick with it forever. You’ll make yourself crazy, because you will never get there! Educate your stakeholders on the importance of a full picture, set up the best data gathering tools you can, and be patient. The perfect customer acquisition cost model may not exist, but your marketing and sales results will benefit from your effort.
About The Author
As Vice President of Marketing at Boardable, Julie is passionate about helping nonprofits tap into new technology in order to better serve their missions and constituents. With 17 years of online experience under her belt, Julie has previously served as the VP of Digital Marketing at two digital agencies and as the Director of Marketing at a mobile app startup.
Boardable is an online board management portal that centralizes communication, document storage, meeting planning, and everything else that goes into running a board of directors. Founded in 2017 by nonprofit leaders and founders, Boardable has a mission to improve board engagement for nonprofits. Boardable is based in Indianapolis, Indiana.