Why marketing needs different approaches to measurement

Some marketing processes are predictable and repeatable. Others operate in constantly changing conditions, where outcomes are harder to forecast. Each requires a different approach to measurement.

Stable marketing work lends itself to precise metrics, but applying the same standards to unpredictable activities can lead to misleading conclusions, unrealistic expectations, and poor decisions.

Whether a function remains stable largely depends on the degree of variety and certainty in the work. Work with low variety and high certainty has reliable, consistent inputs and outputs, with known sources of variance. Information about the work is easy to obtain and unambiguous. 

Why stable marketing processes are easier to measure

Constraints often drive stability. Most accounting processes are stable because GAAP rules dictate them. External constraints, such as contracts or regulations, and internal constraints, such as brand guidelines, serve to keep processes orderly.

Examples of stable marketing processes include:

• Brand asset usage: Logo placement, color codes, font specs, sizing rules
• Email list hygiene: Bounce handling, unsubscribe processing, suppression list management
• Creative approval checklists: Legal/compliance disclaimers, required disclosures, trademarks
• Social media posting cadence: Platform-specific formatting and timing rules
• Event/webinar registration confirmation sequences
• A/B test statistical significance thresholds
• Budget allocation reporting: Standard monthly/quarterly rollups

Stable work can be measured precisely. Managers love stable work. Knowing the rules makes life easier. Stable processes work like a factory. Even when work is complicated, you can mostly control it and predict the future accurately. You can create efficiencies and develop best practices, including new ones that incorporate AI. You can optimize. Factory-like work emphasizes deterministic KPIs that help you see precisely what works.

For decades, business schools have taught students about managing stable business processes. Managers are good at it. Many executives don’t realize other kinds of processes exist. They think enough intelligence and technology could make all work orderly. But not everything in the world is stable.

Marketing also has a wild side

You can’t control many things in life. Raising a child, the weather, economies, emotions, organizations, stock markets, sports, customer markets, geopolitics, and traffic are examples. The U.S. Army War College, readying soldiers for unpredictable battles, called these conditions VUCA (which stands for volatile, uncertain, complex, ambiguous). Instead of mechanistic factories, these environments work like interconnected ecosystems.

Elements that destabilize a system and kick it into VUCA mode include continually changing external conditions, novel situations, competing objectives, and the feedback created by interactions between independent agents, like customers, competitors, partners, influencers, and governments.

Examples of VUCA business processes include:

• Marketing and sales campaigns.
• Customer journey management.
• B2B deal management.
• Revenue forecasting.
• New product development and innovation.
• Strategic planning.
• Customer service for unusual problems.
• M&A integration.

Marketing, along with sales and customer support, performs at the company’s edge, meshing daily with the VUCA world outside. Customer journeys are more like children’s scribbles than orderly funnels. Inputs and outputs don’t easily correlate. Customers’ changing behavior creates high engagement variance in most industries. 

VUCA marketing work can’t be measured precisely. Complex areas of life and business aren’t completely chaotic. They exhibit patterns and thus are semi-predictable and semi-manageable. Marketing automation is a great first step in producing data to help understand shadowy customer journeys, but people grew too optimistic that marketing will finally be controllable and predictable. Instead, the data gives us more evidence that it isn’t. 

When ranges and probabilities outperform precise numbers

For marketing processes on the wild side, where performance is inherently variable, false precision carries high risk. Using measurement practices that work in stable environments leads companies down paths that are, at best, frustrating and at worst, catastrophic.

Your single North Star metric is quietly wrecking your decisions.

A single-number metric, commonly used for targeting performance under stable conditions, is useless for decision-making in complex environments. 

It’s easy to see why companies favor single-number metrics. A single number carries an aura of authority and objectivity, even if it’s wrong. Saying, “This initiative will increase sales by 6.7%” feels more useful than saying, “We estimate there’s roughly a 60%-70% chance this initiative will increase sales.” The latter is more honest because the precision of a single data point is unwarranted, but it’s psychologically uncomfortable. 

A single metric is a static snapshot that tells you what happened at a point in time but there’s no guarantee that snapshot will hold if the context changes. If customer engagement were a stable environment, then marketers could forecast linearly from a single data point. For example, you might conclude that if you got 10 leads in the first week of a campaign, then a four-week campaign should produce 40 leads. Maybe. More likely, results will vary, leading managers to applaud higher numbers and condemn lower.

Why ranges work better

View variance in a complex business process as critical information, not failure. In the real world, complex environments produce a range of results with natural inconsistencies. Ranges reflect real-world uncertainty. Some days you’re up. Other days you’re down — as with weather, sports, or the stock market. Changes in the range’s trend line aid understanding about the probability of future direction and therefore help make better decisions. 

Consider one B2B company that conducted most of its business development at events. They decided to try inviting a customer segment they’d never worked with. The first event targeting this segment drew the highest attendance the company had ever recorded.

The local sales team burst with enthusiasm, imploring the marketing team to invest their entire field marketing budget into this segment. However, the analytical marketing team understood that even this high achievement was within the expected probability confidence level for their events. They continued testing in this new market, but the high results were never replicated. This market eventually sank to lower-than-average results. 

You still need good, detailed data to feed models and support diagnostics. But in a variable environment, like most of marketing, emphasize broader, multi-level metrics, such as:

• Leading indicators of demand (search behavior patterns, category signals).
• Narrative and qualitative insight (customer interviews, community behavior).
• Relationship depth and account health.
• Brand equity and trust.

Rebuilding trust between executives and marketers

Much of the frustration and mistrust between the C-suite and marketing leaders stems from the inappropriate application of stable business management and measurement practices to inherently VUCA marketing. 

Humans dislike uncertainty. Methods for managing and measuring in complex environments are newer and less mature in business (although other arenas like the military, first responders, and sports have extensive expertise).

Today’s managers may excel at managing stable environments and desire precise, deterministic metrics. However, applying the wrong tools just because you know and prefer them is like the proverbial situation where, to someone with a hammer, everything looks like a nail. Some business processes are just not nails.

Using the appropriate measurement techniques for the business context demonstrates confidence and skill. They protect against unrealistic expectations, help prevent trust erosion, and deliver better results.

The post Why marketing needs different approaches to measurement appeared first on MarTech.