Actuals, Assumptions, & Improvements

The Missing Piece to Marketing Models

Here’s a quick one for you today…

Turns out, I’ve been thinking about marketing models all wrong. 

Well, not all wrong, but I was clearly missing a key piece to the puzzle. 

Actuals & Assumptions

For most of my career, I’ve viewed marketing models as having two core components:

1. Actuals: The components of a marketing model built off of actual, in-market client data, such as current cost pers (e.g., cost per lead or cost per application), historic conversion rates, etc.

2. Assumptions: Best-guess insertions on the part of marketing where and when actual data doesn’t exist and/or anticipated performance improvements, where we want to plan on performance above/beyond what we’re seeing from actuals. 

Actuals, Assumptions, and Improvements

Recently, while working with the team at Kanahoma, it became apparent that our “Assumptions” category was trying to do too much, as it was both representing best guesses for missing historical data, as well as baked-in anticipated improvements to performance. 

As a result, it only made sense to split assumptions and add a third category: “Improvements.” 

This simple change meant that we now had three key categories of data to inform all marketing models: 

1. Actuals: The components of a marketing model built off of actual, in-market client data, such as current cost pers, historic conversion rates, etc.

2. Assumptions: Best-guess insertions on the part of marketing where and when actual data doesn’t exist.

3. Improvements: Anticipated performance improvements, where we want to plan on performance above/beyond what we’re seeing from actuals. 

Why Does This Matter?

A marketing plan is only as strong as the confidence the team has in it. 

All too often, marketing models will be developed by a centralized budgeting process that excludes many of the key stakeholders actually responsible for fulfilling the plan. 

One of the key causes of this problem is that folks asked to own the delivery of such a plan (e.g., marketing, enrollment, etc.) often don’t even have insight into what information informed the plan, leaving many in the dark on everything except the aspirational goals they’re suddenly saddled with. 

So How Do We Fix It?

The best thing organizations can do is utilize a transparent and progressive approach to building marketing models that works through the three data points linearly:

• Model 1 (Actuals): Build an initial marketing model with your existing historical data. Missing data? Just leave it blank and mark it in red as MISSING.

• Model 2 (Actuals + Assumptions): Next, building off of Model 1, develop a second iteration of the model where you insert assumptions for all MISSING data points. These should represent your best guess of what the actual data likely is.

• Model 3 (Actuals, Assumptions, & Improvements): Chances are Model 2 won’t get you the kind of growth the institution is looking for, so we now move on to Model 3. In Model 3, we duplicate Model 2 and now insert anticipated performance improvements, such as decreased acquisition costs, increased conversion rates, etc. 

Improving the Process with “Improvements”

There’s obviously a lot more to marketing models, such as building them top down, bottoms up, or middle out, but the key lesson I want to share for today is…

When you’re building a marketing model, it’s critical that all involved understand where the data comes from and what it represents. 

Is this historical information? Great, color code it accordingly as “Actuals.”

Are these assumptions in lieu of actual historical data? Color code it separately and note them as “Assumptions” so everyone knows. 

Or are these anticipated improvements to historical performance? Ok, then mark it accordingly as “Improvements” and make sure everyone knows both what performance is expected to improve, as well as why. 

Conclusion

The next time your organization sets out to build a new budget or marketing/enrollment plan, I’d encourage you to slow down and ask yourself: When it comes to the data you’re intending to use…

Is It Actuals, Assumptions, or Improvements? 

---

About the Author
Seth is the founder and CEO of Kanahoma, a San Diego-based performance marketing agency on a mission to build a better agency for organizations building a better world.

You can learn more about who we are and what we do at www.Kanahoma.com.