Please forgive the extremely loose nature of this blog post. The following thoughts are not even close to being complete or organized – much like this whole website, this is a work-in-progress and a dumping ground for ideas that emanate from JCM’s marketing approach. In case someone stumbles onto this, I (Jeff Travilla) am trying to develop a sort-of grand theory of digital marketing… like an actual math equation or theorem or stepwise process – something more tangible and concrete than the intuitive process I’ve been using for the past few years. I figure there are enough common threads across the way I approach these things, and there’s enough consistent, repetitive success… there must be something here worth breaking apart. Maybe it isn’t a “grand theory” and maybe these processes break down as technology shifts, but I think these ideas are fundamental enough to survive tech changes. Anyway, if you stumble onto this and want to help me develop these ideas into a full-fledged mathematical theory or something like that, I’d love to work with you on turning this into a book or a piece of software… something. Or loop you into this site as a collaborator, partner, paid consultant, whatever… So here’s the stuff that is sputtering around in my head right now…
The JCM Equation
These puzzle pieces all play together, but they are somewhat dependent and flow generally in the following order:
DATA
ROI is the only KPI
Connect all sources of marketing data where possible.
Set up conversion goals on website and trackable ancillary properties. Conversion goals must generate revenue directly (eCommerce) or indirectly (lead generation).
Revenue from lead gen must be properly attributed to customer touch points – possible methods include CRM integration, sending contact forms to Google Sheets, verbally asking customers how they learning about you and entering into CRM manually (least effective)… at minimum, we need to know how much revenue to credit back to digital marketing activity. At best, we’d like to know what digital channels to credit, but this can get tricky when customers have multiple interactions with a website and enter via different means (e.g. first clicked an ad, then a remarketing ad, then bookmarked and came back direct, then googled for an organic result – which channel should get credit?)
Regression test all potential metrics –
- First use Revenue as dependent variable and test all metrics. Conversion Rate should almost certainly be the most influential, and if so, we want to break it out of the model and see what behavioral variables drive Conversion Rate
- If any other metrics are statistically significant drivers of Revenue, they become instant KIMs (key influential metrics)
- Regression test Conversion Rate as the dependent variable and retest all metrics. This is an attempt to determine what behaviors and/or channels lead to better conversion rates, IF Conversion Rate is the primary driver of revenue (I’ve never seen it not be)
- Statistically significant drivers of Conversion Rates also become KIMs
EMPATHY
An empathic understanding of your customers allows you to understand the reasons behind the influential metrics and derive meaningful strategy. Methods for gaining this understanding include: heatmapping, watching recorded user sessions, focus groups, surveys, talking to the sales department about customer conversations. Openness and vulnerability are key to truly hearing and absorbing this feedback. Take these empathy insights over to your KIMs and make strategic guesses about what changes to channel or assets would push KIMs forward.
CREATION
Using the data/empathy-derived strategy to change something or make something.
TESTING
Iterative changes either via A/B testing (most scientific) or done slowly and methodically enough to intuit impact on KIMs and ROI. If KIMs improve but ROI does not, wait for potential time lag of KIM’s effect on ROI. If there is still a disparity that is longer than the average sales cycle, KIMs need to be re-evaluated
BELL CURVE
Pushing each KIM forward often involves additional cost. This additional cost should be tracked against the KIM it is meant to influence to find the top of its individual bell curve. Ad spend vs revenue generated directly from ad clicks is a simple example. Once ads start working, you can add more and more ad budget to generate more and more revenue… until you’ve saturated and then the more you spend produces smaller revenue returns.
ROI on marketing expenses (incl internal salaries) should also be charted to determine the overall bell curve. The top of the curve is the most efficient use of resources. Somewhere over the top of the curve to the right is the current paradigm’s growth break-even point where spending more money on marketing actually causes you to lose money.
To find growth-break-even, non-marketing expenses are now a factor because you have to determine how far over the curve you can push marketing expenses into their less effective territory before other variable and fixed expenses catch up to the slower addition of new revenue. This might seem complicated, but consider a simple example:
- 100 people are in a room and all of them are in need of a toothbrush.
- You start advertising to them via their cell phones, and the ones with the most desperate need buy a toothbrush immediately. It only cost you one add click for these really stinky-breathed target customers.
- More people in the room eventually see a second ad and then decide “OK, I think I’ll buy one now.” You generated additional revenue, but your ads took two clicks this time – less efficient.
- By the third ad click, everyone in the room except for one person has bought a toothbrush from you, and you have been turning a net profit, even though your advertising was getting more expensive per customer as you saturated the captured audience with ads.
- Well what about this last guy? He’s a stinky-breathed hold out, and he won’t buy a toothbrush from you until you show him 10 ads. Well, the problem is, the cost of 10 ads is higher than gross profit on the toothbrush (revenue minus COGS). So this guy isn’t worth it.
- That last guy represents your growth-break-even. You can no longer turn a profit by increasing marketing spending. In fact, you start losing net profit because you have reached maximum performance, or “growth break-even.”
SCALE
Consistently maxed out performance beyond the peak of the bell curve requires new input to the model in the form of a scaling mechanism. This can be vertical or horizontal, but new opportunity is required
Determining the top of the bell curve for each metric. Using empathic understanding of customer needs to make guesses about why the top of the curve was reached and how to push the whole curve taller
