I understand why marketing can look simpler from outside the function than it often feels from inside it. Most of what the rest of the business sees is the output: an advert, a website, an email, a landing page, a campaign, a brochure or a report. The finished item can look relatively straightforward, so when a marketer pushes back on what appears to be a small request, it can sometimes look as though they are making something unnecessarily difficult.
Questions such as “Why can we not just launch it?”, “Why does changing one field matter?” or “Why do you need the tracking in place before the campaign starts?” are perfectly reasonable questions to ask. There is one word that appears surprisingly often in those conversations, though, and that word is just. The person making the request can usually see the immediate change, while the person responsible for marketing may be looking at several systems, processes or decisions that depend on it.
Sometimes marketing genuinely is being overcomplicated. I have worked in the profession long enough to know that marketers are perfectly capable of creating unnecessary processes, becoming obsessed with details that make very little commercial difference or allowing technology and terminology to make straightforward decisions appear harder than they need to be. Complexity should never become an excuse for slow delivery, excessive process or avoiding accountability.
However, there are also plenty of occasions when the marketer is not being awkward at all. They are simply seeing something that is not immediately visible in the request. Modern marketing is no longer a collection of independent campaigns and creative outputs. It is a connected system that begins with the commercial objective, then moves through audience and segmentation, proposition and positioning, budget and channel planning, campaign development, customer experience, tracking, CRM, automation, sales handover, attribution and reporting before eventually influencing the next commercial decision.
Sitting across that entire system are customer data, technology, definitions, ownership, governance and increasingly artificial intelligence. A decision made at almost any point can alter what happens further down the chain, sometimes without the consequence becoming apparent until weeks or months later.
That is why I think marketing infrastructure has become one of the most important and least understood parts of a modern marketing function. Marketing infrastructure is the combination of people, processes, technology, customer data, definitions and governance that allows marketing activity to be delivered, connected and measured reliably. The campaign may be the part everybody sees, but the infrastructure underneath determines who receives it, what happens when they respond, how their information moves through the business, whether sales follows it up, how the result is measured and, increasingly, what an automated system decides to do next.
The complexity is therefore often not in the thing itself. It exists in the relationships around it. That distinction matters because good marketing leadership is not about preventing the business from making decisions. It is about understanding enough of the wider marketing system to recognise what else may change when those decisions are made.
The complexity is in the connections
The scale of the modern marketing technology environment gives some indication of how specialised the function has become. The 2026 Marketing Technology Landscape identified more than 15,000 commercial marketing technology products. Clearly, no sensible marketing department is using thousands of platforms, but the size of that ecosystem demonstrates how many specialist technologies now exist around activities that customers often experience as something incredibly simple.
An established business might have a website platform, CRM, email system, analytics platform, advertising accounts, tag management, call tracking, social tools, reporting software, consent management and customer service technology operating at the same time. Around those systems might sit an internal marketing team, several agencies, sales, external technology providers and legal or compliance requirements, all of which need to support the wider commercial objectives of the organisation.
The individual components are not necessarily difficult to understand on their own. The difficulty comes from the relationships between them and from the fact that information regularly moves from one system to another.
Consider something as simple as a website form. To the visitor, it might be five fields and a button. Behind those fields, however, the form could create or update a CRM record, identify the source of the enquiry, capture segmentation information, determine lead scoring, trigger an automated workflow, notify a salesperson, assign ownership, record a conversion, store a marketing preference and potentially return information to an advertising platform. The visible form is relatively simple, but the dependency chain around it can be extensive.
The same principle applies to an email. The recipient sees a subject line, some content and a call to action, while underneath that communication may sit customer segmentation, CRM data, marketing permissions, suppression rules, automation, domain authentication, sender reputation, conversion tracking and reporting. This is what makes modern marketing deceptively complicated: much of the infrastructure required to make the visible activity work properly is invisible to the people looking only at the finished output.
A marketing campaign starts long before somebody creates the advert
Marketing infrastructure should not be thought of as a technical layer that sits underneath the real marketing. It begins much earlier than the platform, the advert or the campaign because it starts with the commercial objective the business is trying to achieve.
A business might decide that it needs to grow revenue, improve profitability, enter a new market, increase retention or generate more sales opportunities. That objective immediately creates a series of further decisions. Which customers matter most? How should they be segmented? What problem are we solving for them? What proposition will persuade them? What level of investment is commercially sensible? Which channels have a realistic role in delivering the outcome, and over what period should the business expect those channels to work?
Only after those questions have been considered do we reach the campaign, website, content or creative work that most people recognise as marketing. From there, the dependencies continue because the campaign creates a customer experience, the experience generates interactions and data, that information needs to be captured and attributed correctly, customer records move into the CRM, qualification rules determine what happens next, automation may trigger communications or route an opportunity, and sales eventually takes responsibility for progressing the commercial conversation.
The final outcome is then reported, analysed and used to determine what the business does next. That might mean changing the budget, adjusting the proposition, investing more heavily in one channel, changing a customer journey or deciding that the original commercial assumption was wrong. Marketing therefore looks less like a series of disconnected activities and more like a connected operating cycle in which the output from one stage becomes the input into another.
A simplified version of that cycle might move from commercial objective, through audience and segmentation, proposition and positioning, budget and channel planning, campaign and content, website and conversion experience, tracking, identity and consent, CRM and customer data, automation and nurture, sales handover and pipeline, attribution and reporting, before reaching optimisation and the next commercial decision.
Across that whole cycle sit privacy, permissions, security, definitions, documentation, ownership, compliance, change control and other governance requirements. A change to the customer journey can alter the data being collected, a change to customer data can affect segmentation, segmentation can alter automation, automation can change the sales handover, and a different definition of a qualified lead can alter both reported conversion rates and the information being sent back to advertising platforms.
This is one of the reasons I see a distinction between marketing delivery and marketing leadership. A senior marketer cannot judge every request solely by looking at the task immediately in front of them. They also need to understand enough of the wider system to recognise what that task might affect further down the line.
The marketing decision is only the beginning
Modern marketing is a connected system of strategy, customer experience, data, CRM, automation, technology, governance and commercial decisions. Select any stage to see how one decision can affect everything that follows.
Select any stage above. The diagram will highlight the connected systems and show the consequences that may need to be considered before what looks like a simple marketing change is made.
- Customer experience and conversion
- Data quality and reporting
- Automation and sales handover
- Future budget and optimisation decisions
A small change can improve one visible metric while creating problems further down the customer journey.
Why small marketing changes rarely stay small
Some of the clearest examples of marketing complexity come from requests that initially sound extremely straightforward. The issue is rarely whether the business can make the requested change. In most cases it can. The more useful question is what else needs to be considered before it does.
Can we launch now and fix the tracking afterwards?
Tracking is still sometimes treated as something that gets added once the creative work has been completed. The page is built, the adverts are prepared, the campaign launches and somebody then makes sure Analytics is recording everything correctly. That approach might once have resulted mainly in a weaker report, but the role of measurement within modern marketing technology has changed considerably.
Campaign tagging, referral information and conversion definitions affect how analytics platforms categorise activity and determine where traffic and commercial outcomes came from. Those decisions influence how the business subsequently understands performance and where it decides to invest.
More importantly, marketing platforms increasingly use conversion information to make decisions themselves. Automated advertising systems can optimise future activity using the conversion signals returned to them. If the business defines the wrong outcome as success, or if the conversion data being supplied is unreliable, the problem is no longer simply that a report is inaccurate. The platform itself may also begin optimising towards the wrong signal.
That makes measurement infrastructure part of the campaign rather than something that sits neatly behind it. Poor tracking can affect reporting, attribution, optimisation and eventually the next budget decision, which is why launching first and fixing the tracking later is not always the harmless shortcut it initially appears to be.
Can we just change the form or the lead stages?
Imagine a business discovers that asking for company size on a website form is reducing completion rates. Removing the field might increase the conversion rate, and viewed only at the point of conversion that could appear to be an obvious improvement.
The wider system may tell a different story. Company size might determine lead scoring, route larger organisations to a particular salesperson, control which nurture sequence somebody enters, support customer segmentation in reporting or determine which enquiries are sufficiently valuable to be returned to an advertising platform as qualified opportunities. Removing the field could therefore improve one visible metric while weakening several downstream processes.
The same issue appears with CRM definitions. A business may use terms such as Lead, Marketing Qualified Lead, Sales Qualified Lead, Opportunity and Customer, and somebody may quite reasonably decide that the terminology should be simplified. There is nothing inherently wrong with changing those stages, but the important question is what the stages actually mean and what depends on them.
An automated workflow may run when somebody reaches a particular stage. Sales might receive an alert, a service level agreement may begin, a dashboard might calculate conversion between stages or an advertising platform might receive one of those events as an offline conversion. Changing the definition can therefore change the reported performance of the marketing function even if customer behaviour itself has remained exactly the same.
This is why I think definitions form part of marketing infrastructure. A qualified lead is not an objective thing that exists naturally inside a CRM. It is an operational definition agreed by the organisation. If marketing, sales and senior leadership are working with slightly different definitions of what success means, all three can produce accurate information and still reach very different conclusions about whether marketing is performing.
This connects closely with something I wrote recently about expectation and marketing performance. Performance becomes much harder to judge when the people involved are starting from different assumptions about what the business expected marketing to achieve.
Can we just email the database?
Few marketing requests sound simpler while containing quite as many hidden dependencies as asking somebody to email the database. Before the content is even written, somebody needs to understand where the customer information came from, who the recipients are, why their details were collected, what marketing permissions or objections have been recorded and whether appropriate preferences and suppression information are being maintained.
There is then the technical infrastructure required to make the communication arrive. Email authentication, domain reputation, sender reputation and configuration all affect deliverability, while major email providers increasingly expect organisations sending marketing communications at scale to meet particular authentication, reputation and unsubscribe requirements. A beautifully written email creates very little commercial value if weak infrastructure means it consistently lands in spam or never reaches the intended recipient.
The dependencies continue after the email has been delivered. A click might update a CRM record, change lead scoring, create a task for sales, alter a lifecycle stage, contribute to campaign attribution or move somebody into or out of another automated customer journey. What appears externally to be a relatively simple email can therefore sit at the intersection of content, customer data, permissions, CRM architecture, automation, infrastructure, sales and measurement.
The email itself is the visible output. The wider marketing infrastructure determines whether the right person should receive it, whether it reaches them, what happens when they respond and whether the business can reliably understand the commercial outcome afterwards.
Marketing data is constructed, not simply discovered
Another reason marketing can become difficult to understand is the assumption that marketing data exists independently and the marketer’s job is simply to report it. In reality, many of the numbers that eventually reach a dashboard are shaped by definitions and decisions made much earlier in the customer journey.
If somebody says there were 312 conversions last month, an experienced marketer should want to understand exactly what that means before deciding whether the number is good. Is a conversion a completed form, a qualified lead, an opportunity or a sale? Which system produced the number? What period does it cover? How has the result been attributed? Are we using exactly the same definition we used last month?
Different platforms can display different numbers without either necessarily being wrong. One system might be measuring sessions while another measures users. A CRM may record the date an opportunity became qualified, while an analytics platform records the date of the original website interaction. An advertising platform might attribute the eventual sale according to its own methodology and reporting window.
A marketing number therefore has a definition, scope, source and collection method behind it. Without understanding those things, the apparent precision of the number can be misleading.
This is why I think businesses should be cautious about assuming that better marketing reporting simply requires a better dashboard. A beautifully designed dashboard built on unclear definitions, unreliable customer data or inconsistent attribution simply makes the confusion easier to look at. It can be mathematically correct while still giving senior leadership the wrong commercial impression.
Good marketing reporting begins much earlier. The organisation needs agreed definitions, sensible data structures, consistent measurement and a clear understanding of which system should be trusted for which purpose. Once those foundations are in place, dashboards become much more useful because the business can concentrate on what the information means rather than debating where the number came from.
Automation means the consequences travel further
Marketing automation is powerful precisely because it allows an organisation to define a process once and repeat it consistently. That creates enormous efficiencies, but it also means that a poor decision can be repeated just as reliably as a good one.
A manual mistake might affect one customer. An incorrect automation rule could affect thousands. A badly designed CRM property could place the wrong customers into a nurture journey, a poor segmentation rule could send inappropriate communications, and an incorrect lifecycle stage might trigger the wrong sales handover or exclude somebody from a process they should have entered.
The question before automating something should therefore go beyond whether the technology is capable of doing it. The more important question is whether the underlying decision, definition and data are reliable enough to make automatically.
This is one of the reasons marketing operations and infrastructure matter so much. Automation makes the relationship between customer data, definitions, workflows and customer experience considerably tighter. If the information entering the system is unreliable, the fact that the workflow executes perfectly does not make the outcome correct. It simply allows the underlying error to operate faster and at a much greater scale.
AI makes marketing infrastructure more important, not less
Artificial intelligence extends the same principle because more marketing systems are becoming capable of making or recommending decisions rather than simply presenting information to a person. Much of the public discussion about AI in marketing still concentrates on content generation, which is understandable because writing, imagery and creative tools are the most visible applications. The more significant development, in my view, is the growing use of AI and machine learning inside decision making systems.
These technologies can increasingly influence which bid is made, which audience is prioritised, which customer receives a particular experience, which lead is scored more highly, which communication is selected and which action the system recommends next. That makes the quality of the information and processes underneath those decisions increasingly important.
We are beginning to introduce AI into marketing functions that, in many businesses, have not yet been properly joined together without AI. If CRM information is unreliable, giving another system permission to act on that information does not fix the underlying problem. If lifecycle definitions are inconsistent, automation does not resolve the disagreement, and if attribution is poorly understood, adding machine learning does not suddenly create a reliable commercial truth.
AI therefore does not remove the dependency on good customer data, sensible definitions, clear ownership and appropriate governance. It gives the existing marketing infrastructure more ability to act. Where that infrastructure is strong, there are significant opportunities to improve efficiency, personalisation, analysis and decision making. Where it is weak, the same technology can allow poor information and bad assumptions to travel much further and much faster.
For me, the sensible starting point is therefore not simply asking where AI can be added to marketing. It is understanding whether the information, process and governance underneath the proposed use are good enough to automate in the first place.
Governance is part of marketing infrastructure
Governance is another area that can appear separate from marketing until you consider how customer information actually moves through a modern marketing function. Data can pass between websites, CRM systems, email platforms, analytics, advertising tools, sales processes, automation and AI applications, which means marketing frequently sits close to questions involving privacy, permissions, information security, retention, access and appropriate use.
Those considerations cannot simply be added at the end of the process. If a website form collects customer information, somebody needs to understand why that information is being collected and how it will subsequently be used. If information is passed into an AI system, the organisation should understand what is being shared, why that use is appropriate, which provider receives it, what output is generated and whether the resulting decision could affect a customer.
The same principle applies to email permissions, tracking technology, customer profiling, data retention and access to marketing systems. The technical implementation may belong to different specialists, but somebody needs to recognise that those dependencies exist before the decision is made.
I do not believe the CMO should become the lawyer, Data Protection Officer, security architect, CRM administrator, analytics engineer and AI specialist rolled into one extraordinarily tired person. Senior marketing leadership is not about personally being the deepest technical expert in every discipline.
The responsibility is to understand enough of the whole marketing system to recognise where another dependency exists and when specialist input is required. I do not need to personally configure every email authentication record to know that changing email infrastructure can affect deliverability. I do not need to be an analytics engineer to recognise that changing a conversion definition can alter the information feeding an optimisation system, and I do not need to be a technical SEO specialist to understand that restructuring hundreds of established URLs requires proper migration planning.
Senior marketing leadership is therefore partly about knowing where the dependencies are. It means having sufficient breadth to recognise when an apparently local marketing decision could become a wider commercial, technical or governance problem and bringing in the right expertise before that consequence appears.
Good marketers should reduce complexity, not celebrate it
There is an important counterargument to everything I have written because marketing people can absolutely hide behind complexity. Processes can become unnecessarily bureaucratic, teams can insist that everything needs to be perfect before anything launches, and new technology can be introduced because somebody likes a particular platform rather than because the business genuinely needs it. Documentation, meetings and governance can accumulate until the marketing operation becomes slower without becoming materially better.
None of that represents good marketing operations. Complexity should not be celebrated simply because the modern marketing environment contains genuine dependencies. It should be understood, managed and removed wherever it adds no meaningful value.
In fact, good marketing infrastructure should make the function easier to operate. Definitions are agreed, customer data is cleaner, campaign naming follows sensible conventions, teams know which system owns which information, automation is documented, reporting has a clear source of truth, marketing permissions are understood, important integrations are monitored and agencies know where their responsibilities begin and end.
The objective is to deal properly with the complexity that genuinely matters so that the wider business does not have to experience it every time something needs to happen. Good senior marketers should therefore remove unnecessary platforms, reduce duplicated processes, challenge bloated reporting and avoid designing a technical solution where a straightforward operational change would work perfectly well.
The goal is not to create more process. It is to create fewer surprises and reduce the chances of discovering several months later that an apparently quick fix altered something nobody thought to check.
Before somebody says “just”
One useful discipline is to ask a small number of questions before making a material change to the marketing system. The questions themselves are not especially complicated, but together they create a useful pause between identifying the immediate request and understanding the wider consequences.
What are we actually trying to change? This should begin with the commercial or customer outcome rather than assuming that the activity being requested is automatically the right solution. A request to rebuild a landing page, change a CRM field or introduce a new platform is not an objective in itself, so the first job is to understand what problem the change is intended to solve.
What data does this affect, and what depends on that data? A seemingly minor change could alter CRM fields, tracking, customer identity, permissions, reporting or historical information. Those same data points might then feed automation, segmentation, sales routing, advertising integrations and customer communications, which means the impact can travel considerably further than the original change.
Does this alter what one of our marketing metrics means? If the definition of a lead, conversion, customer or opportunity changes, historical comparisons may no longer be genuinely comparable. The resulting movement in the dashboard may reflect the new definition rather than any meaningful change in customer behaviour.
Does anything outside the immediate marketing activity depend on it? Sales, customer service, finance, search engines, advertising platforms, agencies and other technology may all rely on the same information or process. The fact that the request originated in marketing does not mean the consequence stays within marketing.
Is there a governance implication, and how will we know if something breaks? Privacy, permissions, security, compliance, brand standards and AI governance may all require consideration depending on the change. Somebody should also own the monitoring afterwards and understand what the organisation will do if an unexpected consequence appears.
The value of these questions comes from asking them before the change rather than after the problem. They do not need to create a lengthy approval process, and most will be answered quickly where the marketing infrastructure is already well understood.
Why good marketers sometimes push back on simple requests
I think this is where some of the tension between marketing teams and senior leadership originates. The person making a request understandably sees the immediate decision, while the experienced marketer may be mentally working through the systems, data and commercial consequences that sit behind it.
That can make the marketer appear cautious, defensive or unnecessarily difficult, particularly when they do a poor job of explaining what they are concerned about. There is therefore a responsibility on marketers too, because simply saying that something cannot be done is rarely a useful leadership response.
If changing a CRM stage affects several workflows and alters the comparability of historical reporting, explain that. If a campaign can launch on Friday but the qualified lead tracking will not be available until Tuesday, explain what the difference means and allow the business to make an informed decision. If restructuring the website creates an SEO migration requirement, explain what needs to happen and what the risk looks like if it does not.
The useful response is therefore rarely a flat refusal. It is to explain that the business can make the change, identify what else the decision affects, describe the trade-off and recommend the most sensible way of managing it. That turns apparent marketing resistance into commercial judgement.
For me, good marketing leadership rarely means simply saying no. It means understanding the cost, dependencies and consequences of saying yes.
Marketing infrastructure is a leadership issue
CRM, analytics, automation, website architecture and customer data can easily be dismissed as operational marketing details, but increasingly they determine whether the marketing strategy can work at all. A growth strategy that depends on sophisticated customer segmentation requires reliable customer data. Sales and marketing alignment requires common definitions and a sensible handover process. Performance marketing requires trustworthy conversion information, while automation and AI require underlying rules and data that the business is comfortable allowing systems to act upon.
The infrastructure is therefore not separate from the marketing strategy. It is part of what allows the strategy to operate consistently, scale successfully and be measured properly.
This is also why I see marketing infrastructure as a leadership issue rather than simply a technical one. A senior marketer does not need to personally configure every CRM workflow, analytics event, email authentication record or integration, but they do need to understand how the commercial objective connects to the audience, proposition, investment, customer journey, data, technology, sales process and eventual measurement of performance.
The value is in seeing enough of that system to recognise when a decision in one area creates a dependency somewhere else, understanding the commercial significance of that dependency and involving the right specialist before the consequence becomes a problem.
It is also why adding more activity is not always the answer when marketing performance disappoints. If customer data is fragmented, reporting is unreliable, sales and marketing definitions are inconsistent and several agencies are operating independently, another campaign simply travels through the same fragmented system. The business can become significantly busier without fixing the reason performance was disappointing in the first place.
Sometimes the organisation does need more marketing. At other times, it needs the marketing function it already has to work better.
That is one of the reasons my Marketing Function Reviews look across leadership, strategy, people, agencies, systems, customer data, reporting and performance rather than examining individual campaigns in isolation. The visible marketing problem and the underlying cause are not always the same thing.
The same principle sits behind my Fractional CMO services. My role is not to personally operate every platform or replace every specialist. It is to understand how the different parts of the marketing function fit together, establish clearer responsibility and make sure decisions in one area support rather than inadvertently undermine what the wider business is trying to achieve.
Marketing should not be complicated for the sake of being complicated
Modern marketing contains more dependencies than are visible from its outputs, but that does not mean businesses should accept unnecessary complexity or that marketers should become immune to challenge. Good marketing leadership should simplify the function wherever possible through fewer unnecessary platforms, clearer definitions, cleaner customer data, better processes, more useful reporting, sensible governance and stronger ownership.
Simplifying a connected system, however, starts with understanding how it is connected. The marketer who asks what appears to be an awkward question may be thinking about something that is not visible in the immediate request, whether that is a CRM workflow, a reporting definition, an SEO dependency, an automation rule, a customer permission, an attribution issue or a conversion signal being fed back into an algorithm.
That does not mean the marketer is automatically right. It does mean that the dependency deserves to be understood and that a good marketer should be capable of explaining it clearly enough for the wider business to make an informed decision.
When marketing is underperforming, the instinct is often to add something: another campaign, another agency, another platform or another person. Sometimes that is exactly what the business needs, but if the infrastructure underneath marketing is fragmented, additional activity simply travels through the same fragmented system.
The more useful starting point may be to understand how the marketing function actually works, where its dependencies sit and what is preventing it from performing properly. Only then can the business make a sensible decision about whether it needs more marketing, different marketing or better marketing infrastructure.
That, for me, is one of the central responsibilities of senior marketing leadership. It is not simply to understand the decision immediately in front of the business, but to understand enough of the wider system to anticipate what is likely to happen next.