# Marketing strategy: the guide for companies between 3 and 30 million euro

Source: https://unrivals.com/blog/marketing-strategy/
Site: UNRIVALS · Language: en · Updated: 2026-09-16

> A marketing strategy is the set of decisions that establishes which real advantage you build on, what you want to be known for, and in what order the budget gets spent. It is not the channel plan, and the difference shows up in acquisition cost twelve months later.

---
**TL;DR.** A marketing strategy is **the set of decisions** that establishes **which real advantage** a company builds on, what it wants to be known for in the mind of the market, and in what order the budget gets spent to get there. That definition differs from the common one, which describes strategy as _a long-term plan_ with objectives and channels.

The plan answers the question of **what we do next month**, while the strategy answers the question of **why anyone would choose this company** over the next one on the list.

The difference does not show up in the monthly report, because there the two look alike. It shows up **twelve months later**, in acquisition cost and in the margin the company manages to defend.

The conversation with founders of companies between three and thirty million euro starts almost identically every time. We have budget, we have people, we have campaigns that work, except **the result no longer grows at the rate the spending grows**. What marketing strategy do we need?

The answer they usually get is _a list of channels and a calendar_. That list is **precisely why** the result stopped growing.

## What is a marketing strategy?

Most definitions describe it as a long-term plan that helps a company understand its customers and increase sales. The definition is not wrong, yet it is broad enough that nobody can decide anything with it on a Monday morning. A forty-page document and a posting schedule both fit underneath it.

We work with a narrower definition, because the narrowness is what makes it operable. A marketing strategy is **the set of decisions** that establish three things, in this order.

Which real and verifiable advantage you build on, what exactly you want the market to remember you for, and through which proof you support the claim in front of someone who has no reason to believe you.

Everything that comes after those three decisions is **execution**. Channels, monthly budget, content calendar and campaigns are all consequences. They can change four times a year without the strategy moving, while a change at the level of the three decisions forces every execution underneath to be rebuilt.

This is where the most expensive confusion in the market comes from. Companies believe they are changing strategy when they change the agency, the channel or the wording in their ads. In reality they change execution and keep the same unclear position intact, which is why the result returns to the same ceiling after a few months.

## Why does a marketing plan not replace a strategy?

A marketing plan is an allocation document. It states how much we spend, on which channels, in which months and against which numerical objectives. A good plan is useful and necessary, except it assumes an already answered question that nobody actually answered.

That question is why someone would buy from you, phrased so that a competitor could not say the same sentence. If the company's answer is quality, competitive pricing and fast delivery, then the company **has no answer**.

It has the minimum definition of the category, which anyone can claim, and the plan built on top of it distributes budget behind an indifferent message.

The test we apply before anything else is called the _competitor test_ and it takes thirty seconds. Take the central sentence from the company's home page, place it next to the same sentence from the sites of the first three competitors, and check whether an outsider could tell which belongs to whom.

When the answer is no, the marketing budget pays for differentiation through repetition, which costs several times more than it would cost through a clear position.

This is where the loss I call the _Confusion Tax_ is produced, meaning the gap between the value a company genuinely creates and the value the market perceives. The tax never appears in a campaign report, because it gets paid in distributed form, inside the cost of every impression and inside every commercial conversation that opens on price.

## What are marketing strategies and what types exist?

Textbooks split marketing strategies by channel, meaning content strategy, performance, social media, email, events. That split describes instruments, and **an instrument cannot be a strategy**, because it carries no decision about position.

A more useful split for a company between three and thirty million euro is made by what the company is actually trying to change.

| Type of strategy | What it tries to change | When it makes sense |
|---|---|---|
| Positioning strategy | what the market believes about the company | the real advantage exists, yet nobody knows it |
| Category strategy | the criterion you get compared on | the category is crowded and comparison runs on price |
| Architecture strategy | the relationship between company and products | the portfolio grew unevenly, brands cannibalise each other |
| Commercial activation strategy | what reaches the sales pipeline | the position is clear, yet it produces no conversations |
| Performance strategy | the cost of an acquired customer | everything above works and gets scaled |

The order in the table is deliberate. It is the order in which the five get resolved, and most companies start from the last row because that is where results appear fastest. Those results exist and are real, except they stop at a ceiling no amount of optimisation will cross.

What is missing from the table matters as well. No social media strategy appears, no TikTok strategy and no artificial intelligence strategy, because those are channel decisions.

They enter the discussion only after the company knows what it wants to be known for, otherwise the new channel amplifies a message the market will not retain anyway. We covered the artificial intelligence case separately, in [what a marketing strategy that uses AI looks like](/blog/marketing-ai).

## Why does the marketing budget keep hitting the same ceiling?

The first pillar of our methodology holds that the marketing budget is **for development**, meaning for strategy and architecture, not for endlessly optimising the same channel. The right strategy lowers real acquisition cost, while tuning an ad moves percentages around the same average.

I state the consequence plainly. From the moment the budget goes into optimising the same channels, it buys the **upkeep of the ceiling** the company already has.

Market figures show that pattern at scale. Marketing budgets stayed effectively flat at **7.8% of company revenue** in 2026, against 7.7% in 2025, according to the [Gartner CMO Spend Survey](https://www.gartner.com/en/newsroom/press-releases/2026-05-11-gartner-2026-cmo-spend-survey-finds-cmos-allocate-15-point-3-percent-of-marketing-budgets-to-ai-but-only-30-percent-are-ready-to-scale-ai-capabilities). A year earlier, [the same source](https://www.gartner.com/en/newsroom/press-releases/2025-05-12-gartner-2025-cmo-spend-survey-reveals-marketing-budgets-have-flatlined-at-seven-percent-of-overall-company-revenue) showed paid media consuming **30.6%** of the marketing budget, which equals 2.4% of company revenue.

Read together, the two figures describe a trap. The share of revenue is no longer growing, close to a third of it leaves for paid channels, and the only lever that seems to remain within reach is optimising those same channels. The company performs better and better on the same foundation, which produces smaller and smaller gains.

The 2026 report adds an observation that confirms the second pillar. Marketing leaders already allocate **15.3%** of budget to artificial intelligence, yet only **30%** of them declare themselves ready to scale those capabilities (Gartner, 2026). Money moves to a new instrument ahead of the infrastructure that would make it produce anything.

The model we use to explain the ceiling has four stages. The first stage is optimising what already exists and it does deliver a real gain, though a limited one.

The actual jump comes from the later stages, which touch brand and strategy. A company that keeps returning to the comfortable channel stays in the first stage permanently, while the competitor who moves into the second takes over the comparison criterion.

## How do you recognise an infrastructure problem?

The second pillar says something uncomfortable for anyone selling marketing services. The client rarely has a marketing problem. The client has _something else_. The client lacks infrastructure, meaning systems, channels, positioning and brand foundation, and marketing built on a poor base burns money with remarkable efficiency.

The operable rule that follows is short. **Diagnose before advertising.** Are there channels? Are there salespeople who know what they sell? Is there a clear brand? If the answer to any of those is no, then that is the work, in that order.

European data shows how uneven the foundation still is. In 2025, **63.57%** of EU enterprises with at least ten employees used at least one social media platform, [according to Eurostat figures](https://business-review.eu/business/romania-among-eu-countries-with-lowest-business-use-of-social-media-eurostat-data-297289).

The split by company size is the interesting part, since adoption sits at 60.59% among small enterprises, rises to **76.2%** among medium-sized ones and reaches 89.09% among large companies.

Those numbers get read wrongly if the conclusion drawn from them is that everyone should start posting. They show something else.

Medium-sized companies sit in the middle of a catch-up phase, and that is exactly the moment when the temptation to buy channels ahead of position becomes strongest. The company opens four platforms at once, carries the same indifferent message across all of them, and mistakes activity for strategy.

That moment also explains why the three to thirty million euro band matters. A company below three million does not yet have the volume that justifies foundation work.

A company above thirty million usually has internal structures and an agency budget already. Between those two thresholds sit the companies that built real value in their operation and never turned it into perceived advantage.

## Which layers does a marketing strategy sit on?

For a strategy to stop being a document, it has to sit on layers that can be worked separately. We use four, plus an orchestration layer. The full system, including how the layers connect, is described in [marketing architecture](/blog/marketing-architecture).

**L1, performance marketing.** The question of this layer is how we bring qualified traffic now. The indicators are cost per acquired customer, cost per lead and conversion rate. This is where most agencies work and where most companies ask for help.

**L2, revenue and the commercial system.** The question is how attention turns into pipeline and then into revenue. The indicators are pipeline value, sales cycle length and win rate. This layer holds the double funnel, in which marketing hands the salesperson a conversation that is already prepared.

**L3, the AI brain.** The question is how the methodology becomes a repeatable system instead of knowledge that lives in a few people's heads. This layer produces no revenue on its own, though it is what keeps the other three from starting over on every project. We wrote separately about [what an AI Brain means for a founder](/blog/ai-brain-for-founders/).

**L4, positioning and category.** The question is which category the company owns in the mind of the market. The indicators are cognitive, meaning preference, trust and branded search. This is the densest layer of the method and also the one most often skipped.

The relationship between layers produces the conclusion that matters. You build bottom up, because that is how a company matures, and you derive the message top down, because positioning decides the offer and the offer decides the advertisement.

When that derivation is missing, the ad invents a message of its own, the salesperson invents another, and the market receives three different companies under one name.

The commercial key to the whole discussion sits here too. Most of the market operates only at L1 and L2, because that is where leads and campaigns get requested today. A company that resolves L4 makes its L1 and L2 investment compound instead of evaporate, and the difference shows in cost per customer twelve months later.

One thing deserves to be stated explicitly about L3. An AI brain built on top of unclear positioning learns and repeats the confusion faster, more consistently and in more places. The order holds there as well.

## What does a strategy look like when it gets run on a real company?

Therezia Prodcom, a dairy producer in Pănet, Mureș county, Romania, shows what the difference looks like in figures. Everything above can be checked against a company whose numbers anyone can pull from public registers.

The financial statement filed with the Romanian tax authority for 2025 shows revenue of 84.4 million lei, profit of 9.72 million and 125 employees, which works out to a net margin of **11.51%**.

In euro terms that revenue sits in the middle of the three to thirty million band this article addresses, so the company is a typical reader rather than an outlier. Anyone can re-read the figures [directly in the public ANAF register](https://webservicesp.anaf.ro/bilant?an=2025&cui=6525450), under tax ID 6525450. The category average, calculated across fifty comparable companies, sits at **5.2%**.

The company therefore earns more than twice as well as its sector average. The situation matches the second pillar exactly, since the advantage exists, can be verified in the operation and is known to nobody outside.

Our audit found the company appearing _twice_ in AI engine answers to category questions, while a processor from the same county appears **one hundred and twenty-eight times**.

The real advantage, once dug out, turned out to be the very thing that looked like an operational handicap, meaning the milk collection system spread across dozens of villages.

Out of that came the permanent tagline attached to the name, “60 VILLAGES. ONE TASTE.”, and the map of the collection network became verifiable proof instead of decorative illustration. The full analysis sits in [the Therezia audit](https://unrivals.ro/deck/therezia).

The case also demonstrates the rule we apply to proof. **Proof gets dug out** of what the company already has, and it has to be verifiable by the client without us. Anyone can re-read the **11.51%** margin in the public filing, without depending on what the company told us in a meeting.

A second example shows how the stake gets calculated before anything is promised. At Doctor SKiN, the site converts at **2.72%**, while the first realistic threshold for their category is **8%**. The distance between those two rates is worth **776,402 lei per year**, roughly double the profit the company made in 2025.

That figure **measures the problem**, the way a diagnosis measures an illness. It tells the founder what their current position costs them every year, and the decision from there belongs to them. The eight percent threshold stays a reference point for the calculation, _never_ a promise of a result.

## The nine steps, in the order they get done

Layers tell you where you work. The steps tell you in what order, and the order is the part the market gets wrong most often.

1. **Economic advantage.** You start from what cannot be copied and can be verified in the operation. The anti-pattern is mistaking advantage for a quality the whole category claims.
2. **Customer value.** You translate the advantage into what it protects for the buyer rather than what it produces for you. A cable manufacturer actually sells the thing that is not allowed to fail.
3. **Distinct attribute.** **One attribute**, decided by the _competitor test_. Seven parallel formulations do not cover margin, while one standard does.
4. **Proof.** It gets dug out of the company and has to be verifiable without you. **Positioning establishes what the proof must demonstrate**, and without a central idea the company presents thirty capabilities and forces the client to work out for himself why they matter.
5. **Memory.** One sign, one permanent tagline, repetition. The name stays in place and receives an addition, which is why [a rebranding never starts from zero](https://unrivals.ro/blog/ce-pastrezi-ce-schimbi-rebranding/).
6. **Decision.** The attribute has to change the comparison criterion at the moment of choice, including where choice now begins, meaning search engines and AI models.
7. **Commercial.** The position gets tested on one product and one channel, ninety days, and only then generalised. Launching across the whole range at once makes measurement impossible.
8. **Category.** You occupy the free space instead of claiming the superlative. The category above is usually crowded, while the one below has no owner yet.
9. **Economics.** You show the order of magnitude of the stake, and the promise of a result stays out of the conversation.

The order explains why visual identity appears **only at step five**. If the attribute does not exist at steps one through three and is not demonstrated at step four, then the visual sign has nothing to carry.

This is the error the market makes when it starts a rebranding from the logo. The document that holds all nine decisions together is called a [brand platform](/blog/brand-platform).

## How do you know whether you need strategy or execution?

The diagnosis takes four questions and lasts about as long as it takes to drink a coffee.

1. Can you state in two sentences **why someone would choose your company**, phrased so a competitor could not say the same thing?
2. Do you have **proof** for that claim, which the client can verify without you?
3. Do your salespeople say **the same thing** the website says?
4. Has your cost per acquired customer **fallen** over the last twelve months, at the same volume?

**Three negative answers out of four** mean the problem sits at L4, and additional budget poured into campaigns will return to the same ceiling. A single negative answer means the foundation exists and the work is execution.

You can run this check yourself, without us, in the [two-minute brand diagnostic](/#diagnostic), which returns a score from zero to one hundred on the coherence between what you are and what you communicate.

## Frequently asked questions

### What is a marketing strategy, briefly?

It is the set of decisions that establish which real advantage the company builds on, what it wants the market to remember it for, and through which proof it supports the claim. Channels, budget and calendar are consequences of those decisions rather than parts of them.

### What is the difference between marketing strategy and marketing plan?

The plan allocates resources across channels and months, while the strategy decides the position the company speaks from. A plan can be rewritten four times a year without the strategy moving, though a change of position forces every plan underneath it to be rebuilt.

### How long until a marketing strategy shows results?

It depends on what you measure, and we work with two windows taken from our own engagement cycle rather than from a market average.

Step seven of the method tests the position on one product and one channel for **ninety days**, because that is how long it takes to see whether the message changes anything in conversion. Cognitive effects, meaning preference and branded search, get read across the standard **twelve-month** cycle, because market memory is built through repetition.

### What marketing budget is normal for a company this size?

The international average reported by Gartner for 2026 sits at 7.8% of revenue. The percentage matters less than how it gets split, because a company that puts the entire budget into paid channels without clear positioning pays for differentiation through repetition, which costs more.

### Can we build the strategy internally, without a consultant?

It is possible, with one condition. The competitor test has to be applied by someone outside the company, because people on the inside know the context missing from the sentence and read it as differentiating even when it is not. The rest of the method can be run internally if someone has the time and the mandate.