As the leaves fall from the trees and the weather turns cold, many marketers find themselves with the often daunting task of putting together their annual year-end budget. Ask any seasoned professional about their methodology (aside from those spreadsheet jockeys in e-commerce) and they’ll admit the process is equal parts science, statistics, dart boards and witchcraft.

This post is for that group: those of us who don’t have the precise data to build a bullet-proof budget.

To be clear, the more a marketer can treat their budget like a math problem, the more accurate they’ll be. The budgeting process becomes immensely easier when we can determine the cost and conversion rates of our leads.

But what do we do when that data isn’t available, particularly for campaigns with indeterminate ROI and/or new product or service launches? Here are some thoughts.

Question History. Hard.

I suspect that every subsequent budget ever built begins with a simple copy and paste from the previous year. This allows whatever you did last year to automatically get through the first round of cuts and determines your capacity to implement new strategies and campaigns.

For this reason alone, it’s imperative to review everything old before you consider anything new. Force your previous campaigns to earn their way onto the budget. If the ROI was out of line with your key performance indicators, decide if it really should survive the cut.

RO- What?

There are items within our marketing budgets that seemingly have no discernible ROI. From the events we host to the holiday baskets we send, these are the types of projects that, unfortunately, can cost marketers some respect within the C-suite.

If that’s the case, what is the best way to justify these campaigns? I suggest tackling it from two angles.

Make Your Own Metrics

If the funding is coming out of the marketing budget, you need to treat it like marketing. In other words, these programs need to be all about putting people in and moving them through the funnel.

So if it’s a cocktail party, the coasters should be printed with an upcoming webinar link that can easily be tracked. If it’s a golf tournament, your sales team should know to track KPIs such as number of quality conversations held while out on the links. Afterwards, such conversations should always, always be captured in your CRM.

Ultimately, if you want to calculate ROI from these soft costs, you’d better find a way to measure them.

Basic Instincts

For times when metrics just aren’t available, marketers can examine such expenses through the lens of behavioral economics.

Consider the theory of costly signaling. Costly signaling suggests that, when people or animals invest more to stand out, it broadcasts a message of confidence and their ability to thrive. In business, the costly signaling theory has been applied to such campaigns like the Goodyear blimp. Though it’s impossible to measure tire sales from the blimp, Goodyear clearly demonstrates they have the brand power to do something big.

On a simpler scale, marketers can also consider the implications of mood bias. Countless studies have seen brand and ad recall lift when audiences are in a better mood. That’s why a logo on a banner at an event where the attendees really enjoy themselves is more valuable that publicity at a similar sized event where the attendees are not having a good time.

It probably won’t surprise you that some researchers have actively looked at measuring mood bias with digital marketing.

If what I’ve written above sounds more like advice on how to justify such expenses more than an endorsement to do them, you’re right.

In the war for attention, most of your firing power should be directed to campaigns where ROI is measurable, especially if your company is looking to drive sales more than maintain them. But the academic research on behavioral economics offers a decent framework to measure some of the fuzzier elements of marketing.

New Product and Service Launches

Even if you have robust KPIs established for existing lines, new products, services, audiences and/ or price points guarantee that you’ll have more than the usual number of unknowns.

When determining a budget for less established products, you need to understand the risk tolerance of your organization. After that, it’s important to understand the role of optimism bias, which simply states that, in the majority of the cases, our results will not live up to our expectations.

So if you think it’s going to take X budget to earn Y as a result, you should either add a percentage to X (if the sales budget dictates that Y is non-negotiable), or you should expect fewer results than with X alone.

And truth be told, marketers should even apply optimism bias to campaigns where we do have established metrics. Again, this bias suggests that our results will not live up to our expectations. So if you find yourself asking for more budget each year, you might not be greedy but rather, you might just be smarter than you think.