Why Signal Design Is Now a CMO-Level Decision

Your campaigns may be telling the algorithm the wrong thing, and that mistake compounds over time.

Most CMOs will recognise the problem.

A campaign appears to be performing well in the monthly report. Conversion volume is up. Cost per lead looks efficient. The channel appears to be doing its job.

But the commercial reality tells a different story.

Sales are not improving. Pipeline quality is poor. Revenue is not moving in line with the media spend. The marketing dashboard says one thing, but the business says another.

When this happens, the instinct is usually to question the budget, the channel mix or the campaign structure. Those things matter, but they are often not the root cause.

The more important question is this:

What has the platform actually been taught to optimise for?

That is where signal design becomes critical.

Signal design is the process of deciding which data points, behaviours and business outcomes are fed back into advertising platforms so they understand what success really looks like. In an automated media environment, this is no longer a technical detail. It is a commercial decision.

And increasingly, it needs to sit at CMO level.

What the algorithm is actually learning from

Modern ad platforms do not understand your business by default.

They do not know which customers are profitable. They do not know which leads are worth pursuing. They do not know which enquiries waste the sales team’s time. They only know what they are told.

If an account is optimised towards form submissions, the platform will try to find more people who submit forms.

If an account is optimised towards low-cost leads, the platform will try to find more low-cost leads.

If an account is optimised towards purchases, but not margin or customer lifetime value, the platform will try to find more purchases regardless of whether those purchases are actually good for the business.

That is the central issue.

The algorithm is not necessarily making a bad decision. It is doing exactly what it has been instructed to do.

The problem is that the instruction is often incomplete.

When good-looking campaigns create poor business outcomes

Take a professional services firm using contact form submissions as its primary conversion event.

On paper, the campaign looks strong. Leads are coming in. Cost per lead is improving. The account is learning. The platform is doing its job.

But six months later, the sales team is dealing with a different reality.

Many of the enquiries are from businesses that are too small, have no real buying intent or are completely wrong for the service. The platform has become highly efficient at finding people who fill in forms, but not people who become valuable clients.

That distinction matters.

A form submission is not the same as a qualified opportunity. A qualified opportunity is not the same as closed revenue. Closed revenue is not the same as profitable long-term growth.

If those differences are not reflected in the data being fed back into the platform, the algorithm has no reason to care.

Why platform defaults usually create the wrong incentives

Ad platforms are designed to make campaign setup easier. That does not mean the default setup is right for your business.

The default path usually favours what is easiest to measure. Clicks. Leads. Conversions. Purchases. Revenue.

Those metrics are useful, but they are not always the metrics that should shape optimisation.

For many businesses, the real value sits deeper in the customer journey:

• Which leads become sales-qualified?
• Which enquiries convert into profitable customers?
• Which products drive margin, not just revenue?
• Which customers stay, repeat or expand?
• Which early behaviours predict long-term value?

These are commercial questions, not just media questions.

That is why signal design cannot be treated as a platform setup task. It requires the business to define what quality actually means.

What better signal design actually involves

Better signal design starts with a commercial question:

What does a valuable customer look like?

From there, the business needs to identify the behaviours, events and data points that indicate genuine value.

For a B2B company, that might mean feeding sales-qualified leads or closed-won revenue back into Google Ads instead of raw form fills.

For an ecommerce brand, it might mean weighting purchases by margin, stock position, product category or customer lifetime value rather than revenue alone.

For a software company, it might mean identifying the actions that predict future conversion.

Imagine a SaaS business that discovers customers who attend a live demo within seven days of enquiry are far more likely to convert. In that case, the form fill is not the strongest signal. The demo attendance is.

The platform does not need to understand why demo attendance matters. It simply needs a better signal to learn from.

That is the point.

Signal design turns commercial knowledge into optimisation data.

Why this cannot be delegated too far down

Signal design depends on information that rarely lives in one place.

Some of it sits in the ad account. Some of it sits in the CRM. Some of it sits in finance data. Some of it sits with sales leaders who know which enquiries are genuinely worth pursuing.

That is why the decision cannot sit solely with a media buyer or account manager.

They can implement the structure. They can manage the platform. They can advise on what data needs to flow where.

But they cannot define customer value on behalf of the business.

That requires senior input.

A CMO does not need to personally configure every conversion action, CRM field or offline conversion import. But they do need to own the strategic definition of success.

Because if that definition is wrong, the platform will optimise efficiently in the wrong direction.

The compounding problem of weak signals

Weak signals do not just create isolated inefficiency. They compound.

Every auction the platform enters, every bid it places and every audience pattern it identifies is shaped by the data it receives.

If the signal is shallow, the learning becomes shallow.

If the signal points towards low-quality leads, the platform finds more of them.

If the signal ignores profitability, the platform has no reason to prioritise it.

Over time, the gap widens. The dashboard may continue to show progress, but the business impact becomes increasingly detached from the reported performance.

That is why this problem is so dangerous. It does not always look like failure at first.

It can look like growth.

More leads. More conversions. More activity.

But if the underlying signal is wrong, the account is not scaling performance. It is scaling waste.

Signal design is now part of marketing strategy

Automation has changed the role of marketing leadership.

The CMO’s job is no longer just to decide where budget goes. It is to ensure the systems spending that budget are learning from the right information.

That means connecting media strategy to sales outcomes, CRM data, customer quality, margin and long-term value.

It also means asking harder questions of campaign performance:

• Are we optimising for what is easy to measure, or what the business actually values?
• Do our conversion signals reflect customer quality?
• Are we feeding the platform enough commercial context?
• Can we distinguish between volume, value and profit?
• Are we teaching the algorithm to find customers, or just conversions?

These questions sit upstream of campaign optimisation.

They are not tweaks. They are strategic decisions.

The businesses that get this right will pull ahead

The businesses that win with automated advertising will not necessarily be the ones with the biggest budgets or the most complex campaign structures.

They will be the ones that give the algorithm better instructions.

That is what signal design does.

It gives automated platforms a clearer definition of success. It aligns optimisation with commercial outcomes. It helps marketing teams move beyond volume metrics and towards the customers that actually drive growth.

In a manual media environment, campaign settings were a major source of control.

In an automated media environment, signal quality is control.

And that makes signal design a CMO-level decision.