How To Decide What To Change In Your Marketing

Knowing What Not To Change Is A Decision Too

Marketing optimization uses performance, context, and experience to determine whether an initiative needs more time, a thoughtful adjustment, additional investment, or a different direction.

Optimization Begins With An Answer, Not An Assumption

By the time a marketing initiative reaches Optimization, the business should know considerably more than whether performance went up or down. Planning established what mattered and what the business hoped to accomplish. Execution documented what was actually put into market and why it deserved resources. Reporting connected the resulting performance to timing, investment, activity, and business conditions so the organization could develop a reasonable understanding of what happened and why.

That understanding is what makes optimization possible. Without it, optimization can easily become a habit of reacting to whatever changed most recently. A campaign underperforms for a month, so the creative is replaced. Cost per lead increases, so the budget is reduced. Website traffic improves, so more resources are immediately committed to search. Engagement declines, so the content strategy changes. Each response creates activity, but activity alone doesn’t mean the business is making the right decision.

Optimization asks something more disciplined: does the evidence actually justify a change? Sometimes the answer is yes. Other times the business needs more information, more time, or enough restraint to allow a strategy to continue working. Knowing the difference is what turns optimization from constant adjustment into informed decision making.

Change What The Evidence Supports

Not Every Result Requires A Response

Marketing performance rarely moves in a straight line. Search demand changes throughout the year. Advertising costs fluctuate. Competitors enter and leave markets. Creative performance naturally rises and falls. Customers take longer to make some purchasing decisions than others, and campaigns often require enough time and volume before meaningful patterns emerge. A single reporting period can reveal something worth watching without providing enough evidence to justify changing the strategy.

Consider a campaign that generated fewer leads this month than last month. Reporting may show that investment remained consistent, conversion rates changed very little, and the decline closely followed lower overall search demand. Changing the campaign because lead volume fell could introduce a new variable without addressing the actual reason performance changed. Maintaining the current approach for another reporting period may provide more useful information than immediately rebuilding something that appears to be functioning as intended.

The opposite can also happen. Performance may appear stable while the underlying evidence suggests a problem is developing. Lead volume remains consistent, but qualified opportunities have steadily declined. Traffic continues growing, but increasingly comes from audiences the business doesn’t serve. Cost per acquisition remains acceptable only because a previously successful campaign is compensating for another initiative that has deteriorated over several months. Optimization requires looking beyond the direction of the headline metric and determining whether the broader evidence supports staying the course.

Doing nothing isn’t the absence of optimization when the evidence supports that decision. It is a deliberate choice to maintain the current approach until new information provides a stronger reason to act.

Choose The Size Of The Change Based On What You Know

Once the evidence supports action, the next decision is determining how much should change. Marketing teams can create unnecessary uncertainty when they respond to one problem by changing several variables at the same time. New creative launches alongside a different audience, budget, offer, landing page, and channel strategy. Performance may improve afterward, but the business has made it considerably harder to understand which adjustment actually contributed to the result.

A more useful optimization decision matches the response to the strength and specificity of the evidence:

  • Maintain: Continue the current approach when performance remains aligned with expectations and no meaningful problem has been identified.
  • Observe: Give the initiative more time or collect additional information when a pattern appears interesting but the evidence remains limited.
  • Refine: Adjust a specific variable when reporting provides a reasonable indication of where the problem or opportunity exists.
  • Expand: Increase investment, audience reach, production, or channel activity when performance has demonstrated enough consistency to justify additional resources.
  • Reduce: Pull back investment when an initiative still provides value but no longer deserves its current share of resources.
  • Stop: End an initiative when it has received a reasonable opportunity to perform and the evidence consistently shows that resources would be better used elsewhere.

These aren’t rigid categories, nor does every initiative move through them in order. Their value is in forcing the decision to become more specific. “We need to optimize the campaign” doesn’t explain what the business has learned or what should happen next. “Lead quality remains strong, but volume has plateaued after three months, so we’ll expand the audience while keeping the offer and creative unchanged” provides a decision that can be executed, measured, and evaluated later.

Every Change Should Create The Next Opportunity To Learn

The strongest optimization decisions don’t simply improve current performance. They make future performance easier to understand. If reporting indicates that an offer may be limiting conversions, changing only the offer gives the business an opportunity to evaluate that conclusion. If one audience consistently produces stronger qualified opportunities, shifting additional investment toward that audience tests whether the pattern continues at greater scale. If performance remains healthy and the decision is to maintain the strategy, another reporting period provides additional evidence about its consistency.

This is where Optimization connects the entire marketing operating process back to Planning. The business began by making decisions about its need, audience, position, resources, and definition of progress. Execution turned those decisions into activity. Reporting transformed the resulting performance into understanding. Optimization now uses that understanding to decide what should be carried forward and what should be reconsidered the next time the business plans.

The result isn’t simply a better campaign. Over time, the business develops a stronger body of evidence about how its marketing actually works. It learns which audiences respond, which messages resonate, which channels deserve investment, which resources are required, how long initiatives typically need to develop, and which performance patterns are meaningful enough to influence future decisions. Those lessons make the next planning cycle more informed than the one that came before it.

Marketing optimization therefore doesn’t require a change at the end of every reporting period. It requires a decision supported by what the business has learned. Sometimes that decision is to change one variable. Sometimes it’s to invest more aggressively in something that has earned greater support. Sometimes it’s to stop spending resources on an initiative that has had a fair opportunity to work.

And sometimes, after reviewing the objective, execution, performance, and circumstances surrounding the results, the most informed decision is to leave something exactly as it is.

The observations and examples shared here are based on real-world experience across industries, but results will vary based on business model, market conditions, and execution. The Method is a structured framework designed to bring clarity to planning, execution, reporting, and optimization, not a one-size-fits-all solution.