Contents
“Marketing” as a single word hides six or seven quite different jobs, each with its own outputs, its own timescale and its own way of being measured. Understanding the marketing work areas individually is what makes it possible to say which one your business is currently short of.
This article maps each area to the concrete artefacts it produces, the metric that genuinely reflects it, how long it takes to show results, and how to decide which area to invest in first.
Research and positioning
This is the area that decides what everything else will say. It produces documents rather than campaigns, which is why it is the easiest to skip and the most expensive to skip.
- Outputs: a segment definition, a positioning statement, a messaging hierarchy, a competitive map, and documented objections from real conversations.
- Metric: there is no clean direct metric. The honest proxy is whether downstream conversion improves after messaging changes — measured with a controlled comparison, not asserted.
- Timescale: weeks to produce, a quarter or more before the effect is visible.
- Failure mode: a positioning document nobody reads, because it was written for a funding round rather than for the team producing the work.
Content
Content is the area most often measured by volume, which is the one number that says least about whether it is working.
- Outputs: articles, video, documentation, comparison pages, case studies, sales enablement material.
- Metric: entrances and assisted conversions per piece — not word count, and not aggregate pageviews, which are usually driven by a handful of old items.
- Timescale: three to six months for organic compounding; immediate if the content is distributed rather than published and left.
- Failure mode: publishing on a cadence with no distribution plan, producing a large archive with a small audience.
Search
Search splits into two jobs that need different people: the technical work of making a site retrievable, and the editorial work of deciding what it should be retrievable for.
- Outputs: a keyword-to-page map, technical fixes, internal linking structure, structured data, and page-level optimisation.
- Metric: non-branded organic entrances and their conversion rate. Rankings are a leading indicator, not a result.
- Timescale: technical fixes can show within weeks; content-driven gains take two quarters or more.
- Failure mode: optimising for volume terms the business cannot credibly win, while the specific terms it could win go unwritten.
Paid acquisition
The fastest area to get a reading from, and the one that most reliably exposes problems that are not actually its own — a poor offer or a broken landing page shows up here first.
- Outputs: campaign structure, audience definitions, creative variants, landing pages, and a testing log.
- Metric: cost per qualified outcome, with the qualification defined by sales rather than by the ad platform.
- Timescale: days to weeks — the only area that gives a genuinely fast signal.
- Failure mode: optimising toward the platform’s conversion event instead of the business outcome, which produces cheap leads that never close.
Lifecycle and email
Consistently the highest-return area at small scale and consistently the last one staffed, because it works on an audience you already have rather than one you can point at in a report.
- Outputs: onboarding sequences, retention and reactivation flows, segmentation rules, and the broadcast calendar.
- Metric: revenue per recipient and repeat purchase or retention rate. Open rate has become close to meaningless as a decision input.
- Timescale: weeks for flows, one to two quarters for retention effects to be measurable.
- Failure mode: treating the list as a broadcast channel, which raises unsubscribes and suppresses the flows that actually earn.
Data and marketing operations
The area whose absence invalidates every other area’s numbers. It is unglamorous and it is the first thing to fix when different people bring different figures to the same meeting.
- Outputs: a tracking plan, event taxonomy, attribution model, reporting layer, and the tool stack that supports them.
- Metric: data completeness and the gap between platform-reported and internally verified numbers.
- Timescale: weeks to implement, permanent maintenance thereafter.
- Failure mode: more dashboards than decisions — reporting that describes the past without changing what anyone does next.
Deciding which area to invest in first
The right sequence follows the constraint, not the fashion. A useful diagnostic is to ask where the drop-off is largest and work on that area, even when it is the least interesting one.
- Nobody knows you exist: demand generation — paid for a fast reading, search and content for durability.
- People arrive and leave: positioning and landing experience, not more traffic. Buying more visitors to a page that does not convert scales the problem.
- They convert once and never return: lifecycle. Almost always the cheapest available gain.
- Nobody can agree what is working: operations first. Every other investment is unmeasurable until this is fixed.
Once you know which areas matter for your situation, the next question is who owns each of them — see our guide to marketing team structure.
How the areas depend on each other
The areas are separable as jobs but not as results. Investing in one while its upstream dependency is broken produces the most common disappointment in marketing: the work was done well and nothing changed.
- Everything depends on positioning. Content, ads and lifecycle all inherit their message from it. Unclear positioning does not stay contained — it degrades every output downstream.
- Search depends on content. Technical work makes pages retrievable; it cannot make pages that do not exist rank.
- Paid depends on the landing experience. Traffic sent to a page that does not convert is a way of paying to discover a problem you could have found without spending.
- Lifecycle depends on acquisition. There is nothing to retain until something arrives, which is why lifecycle is genuinely a later investment rather than merely a neglected one.
- All of them depend on operations. Without reliable measurement, every other area’s result is an opinion.
This is why “we tried content and it did not work” is usually an incomplete diagnosis. Content published without distribution, without positioning, or without tracking has three distinct ways of failing, and they call for three different fixes.
Which areas belong inside the company
Each area can in principle be bought in, but they are not equally safe to externalise. The distinction is whether the knowledge produced needs to accumulate inside the business.
- Keep positioning internal. An external partner can facilitate the work, but a company that cannot state its own position has outsourced its identity.
- Keep the measurement definition internal. Who defines success should not be the same party being evaluated against it.
- Production travels well. Writing, design and video are the most straightforward areas to buy, provided the brief comes from inside.
- Technical specialisms travel well. Site migrations, tracking implementation and complex ad operations are intermittent and deep — the exact profile that suits external work.
- Lifecycle is a mixed case. The flows can be built externally, but the customer understanding they encode should not leave with the supplier.
A workable rule: buy the areas where the deliverable is the value, and keep the ones where the learning is the value. The full cost comparison behind that decision is covered in our guide to marketing team cost.
Frequently asked questions about marketing work areas
Can one person cover all of these areas?
One person can cover all of them badly or two of them well. At small scale the correct answer is usually to run two areas deliberately and consciously neglect the rest, rather than to spread thinly across six and produce nothing measurable in any of them.
Which area is most affected by AI tooling?
Content production and paid creative, where output volume has become cheap. The scarce input has shifted correspondingly toward the areas AI does not resolve: knowing which segment to address, and judging whether a result is real. Volume is no longer the constraint in most teams.
How do you tell which area is currently failing?
Follow the funnel from first contact to repeat purchase and find the largest proportional drop. That step names the area. The common error is diagnosing from whichever area reports most frequently, which usually means paid — the one with the fastest data, not necessarily the one with the problem.
You can find additional frameworks for marketing measurement at Think with Google.
You can find more hands-on marketing guides at https://marketing365.vn.
Content from marketing365 is created for SMEs, online shop owners, and new marketers.
You can also read more articles on the same topic in the Analysis Corner category.
