Why influencer marketing?

CM by Lukinski

Influencer marketing is today one of the most effective ways to reach target audiences directly. With creative strategies and data analysis, influencers help convey brand messages authentically and personally. Whether for lifestyle products, fashion or even high-end real estate — influencers build trust and bring reach.

This is exactly where CM Creator comes in: with tailor-made campaigns that set trends instead of following them. So if you feel like diving into the world of influencer marketing, CM Creator offers the perfect opportunity.

Job opening: Influencer Marketing Specialist (full-time)

In this role you develop and manage influencer campaigns for well-known clients. From selecting the right creators to performance analysis, you’re involved in every step. Your workplace is our office in Cologne, where a young, creative team is waiting for you.

• Develop creative strategies
• Plan and implement campaigns
• Collaborate with influencers
• Analyse data and optimise results

Sounds exciting? Then apply now and get started!

• Apply: Influencer Marketing

Internship in influencer marketing – career starter

Want to gain your first experience in influencer marketing? As an intern, you’ll support our team in planning and implementing campaigns. In doing so, you’ll learn everything you need for a career in this field.

• Support in developing strategy
• Research and outreach to influencers
• Organising and evaluating campaigns
• First insights into data analysis

An internship at CM Creator lays the foundation for your career in digital marketing.

• Apply: Influencer Marketing

What connects us

Whether at CM Creator or here at Lukinski — the key to success lies in the combination of creativity and strategy. While at Lukinski we focus on real estate marketing, CM Creator shows how these principles translate to influencer marketing. Both worlds benefit from innovation, precision and a focus on genuine connections.

Apply now

Whether a full-time position or an internship — at CM Creator you can expect a creative environment where you can grow. Apply today and become part of a team that is redefining influencer marketing.

We look forward to your application!

What does an influencer agency do?

You can find our influencer agency and more cases here: CM Creator Cases!

Outdoor and event campaigns:

Indoor campaigns:

Fashion! All the way to virtual influencers:

Platforms like Model Agency One bring brands and creators together for high-impact campaigns.

AfA stands for Absetzung für Abnutzung (depreciation for wear and tear). Landlords can depreciate the acquisition cost of a building for tax purposes over many years. The building loses value through use — this loss in value reduces taxable income. Only the building is depreciated, not the land.

Depreciation rates at a glance

Calculating depreciation: step by step

Purchase price: €500,000 | Year built 1985 | Location: mid-sized town

Determining the land share

The German tax office recognises three methods:

• Standard land value method: the most precise method – ratio of (standard land value × plot size in m²) to the total purchase price.
• Federal Ministry of Finance working aid: a free Excel tool from the tax administration (search term: “Arbeitshilfe Kaufpreisaufteilung”).
• Flat rate of 20–30 %: depending on location – not always accepted by the tax office.

Declining-balance depreciation for new builds

Since 2023, investors in new builds (completed from 1 January 2023) can choose between 3 % straight-line and 5 % declining-balance depreciation:

• 5 % of the respective remaining book value → higher depreciation at first, then decreasing.
• A one-off switch from declining-balance to straight-line is possible – not the other way round.
• Worthwhile if you have high taxable income in the early years.

Renovations and depreciation

• Maintenance expenses (windows, bathroom): immediately deductible as income-related expenses.
• Production costs (extension, loft conversion): increase the depreciation base → depreciated over 50 years.
• 15 % threshold: renovations exceeding 15 % of the building value within the first 3 years after purchase count as production costs (§ 6 (1) No. 1a EStG).

A property holding GmbH is a limited liability company used exclusively or primarily to acquire, manage, and rent out real estate. Instead of holding property privately, the GmbH buys and rents out the properties – offering substantial tax advantages once a certain portfolio size is reached.

Property GmbH vs. Private Ownership: The Tax Comparison

The decisive difference lies in how rental income and sale proceeds are taxed:

When Does a Property GmbH Pay Off?

The GmbH structure generally pays off once the following conditions are met:

• High personal tax rate: Anyone paying 42–45% personal income tax saves around 12–15% on rental income through a GmbH.
• Long-term buildup: You plan to reinvest profits rather than withdraw them immediately (retention advantage).
• No sale planned: Since property sales within a GmbH are always taxable (no 10-year exemption), the GmbH is ideal for buy-and-hold strategies.
• Portfolio of 3–5+ units: Only then do the tax savings exceed the cost of a tax advisor and annual financial statements.

Types of Property GmbHs

Depending on the purpose, different structures exist:

• Classic rental GmbH: Holds and rents out properties, profits remain within the GmbH.
• Commercial property GmbH: Active property trading (more than 3 sales in 5 years = commercial property trading).
• GmbH & Co. KG: Combination of a GmbH (general partner) and a KG (limited partners) – popular for family structures, enables special depreciation.
• Holding structure: A parent company (holding GmbH) holds shares in operating subsidiary GmbHs – enables tax-free dividends within the group (Section 8b KStG).

Setting Up a Property GmbH: Step by Step

Transferring Existing Property Into the GmbH: What Does It Cost?

Transferring already privately held property into a GmbH is expensive:

• Real estate transfer tax: 3.5–6.5% of the market value — due even when transferring to your own GmbH.
• Speculation tax: If the transfer happens within 10 years of purchase, income tax is due on the value gain.
• Notary and land registry costs: Approx. 1–1.5% of the purchase price.

Conclusion: Transferring existing property into a GmbH rarely pays off. It is usually better to buy new properties directly through the GmbH.

Disadvantages and Pitfalls

• No 10-year exemption: Profits from property sales within a GmbH are always taxable — even after 10 years.
• Commercial property trading: More than 3 sales in 5 years triggers trade tax liability, including on value gains.
• Double taxation on withdrawal: Salary or dividends from the GmbH are taxed again privately (total burden ~50%).
• Complexity: Mandatory annual financial statements, bookkeeping under the German Commercial Code (HGB), a tax advisor is essential.
• Financing hurdles: Banks often require personal guarantees for GmbHs — partially undermining the liability privilege.

Property GmbH: Who Is It Right For?

The GmbH structure is suitable for:

• Investors with a high income tax rate (42–45%) who rent out long-term.
• Entrepreneurs building real estate as part of their wealth strategy who want to reinvest profits.
• Families who want to structure and pass on real estate wealth via a GmbH & Co. KG.
• Investors with a portfolio of 5+ units, where tax savings exceed ongoing GmbH costs.

For beginners with 1–2 apartments and a medium tax rate, the complexity of a GmbH usually isn’t worth it.

Why real estate data and DSGVO belong together

Real estate data may seem harmless at first glance: square meter figures, construction years, energy values, purchase prices. Once you look one level deeper, the picture changes. A rental agreement contains names, dates of birth, sometimes bank account details and employment verification. A statement of additional costs reveals consumption profiles of individual households. A financing request at the bank includes income and asset ratios. A due diligence check before a real estate sale bundles land register extracts, rental agreements, maintenance histories, and sometimes even health-related information, for example in care or nursing properties. All of this constitutes personal data within the meaning of Article 4 No. 1 GDPR, as soon as they can be attributed to a natural person.

The three data categories that collide in the real estate industry

In practice, three data categories overlap in almost every process: first, official and semi-official market data such as standard land values or rent indices, which are not personally identifiable on their own, but can quickly provide insights into owners or tenants in the context of a specific property. Second, contractual data from rental, purchase, and financing agreements, which are clearly personally identifiable. Third, internal calculation data from investors and developers, which usually do not contain personal data, but possess a comparable level of protection as a trade secret. A professional AI solution used in the real estate industry must be able to handle all three categories cleanly – technically as well as legally.

The legal basis for this is provided by the General Data Protection Regulation itself, supplemented by the Federal Data Protection Act as a national complement. Those who want to familiarize themselves in detail with the basic principles – lawfulness, purpose limitation, data minimization, storage limitation – will find the complete regulation text under dsgvo-gesetz.de and the German supplement in the BDSG 2018 under gesetze-im-internet.de. For the real estate industry, it is particularly relevant that processing of personal data by third parties – such as an AI provider – may only take place on the basis of a clear legal basis and, as a rule, through a processing agreement under Article 28 of the GDPR. Exactly at this point, many market participants encounter the real problem.

Special categories of data: when living and health coincide

The legal situation becomes even more complex as soon as special categories of personal data under Article 9 of the GDPR are involved. This is a common occurrence in the real estate industry, more frequently than many market participants assume: when it comes to care homes and supported living, health data is included in occupancy plans and care agreements. When renting barrier-free apartments, degrees of severe disability are sometimes documented. In creditworthiness checks as part of a rental or financing process, scoring data is generated, which can be classified under profiling as defined in Article 22 of the GDPR, as soon as they automatically influence a contractual decision. Anyone using an AI solution in such scenarios must first ask themselves whether and to what extent particularly protected data may even be entered into the system – regardless of how powerful the model is.

No theoretical risk: Fines and reputational damage

Violations of the GDPR are not a mere formality with no consequences. Fines can reach up to 20 million euros or four percent of the global annual turnover, whichever is higher. For a medium-sized property management company or a regional construction developer, however, the fine itself is usually not the greatest risk, but rather the loss of trust: An owners’ association that learns that tenant data has ended up uncontrolled in a US cloud will typically change its management. An institutional investor who is asked during the due diligence review of a fund about the data protection concept of the software tools used and cannot provide a clear answer will likely lose the mandate.

The Compliance Dilemma: US Cloud AI and Professional Secrecy

Popular AI chatbots process requests on servers outside the EU by default, mostly in the USA. For a private user who wants to generate a recipe idea, this is unproblematic. For a tax advisor, a notary, a law firm specializing in real estate, or a bank engaged in financing, the situation is different. These professions are subject to a legal duty of confidentiality – for lawyers and notaries, this is even punishable under § 203 of the German Criminal Code. Whoever enters client data into a tool whose server location, training use, and deletion concept are unclear, is operating in a legal gray area that could, in the worst case, lead to professional legal consequences.

Why “the model doesn’t learn from my data” is not a sufficient answer

Many providers advertise that user inputs will not be used in the training of future model versions. This may be technically accurate, but it does not answer the actual data protection question. What matters is not only whether data is used for training, but also where it is processed and stored, who has technical access within the provider company, what deletion periods apply, whether a transfer to a third country under Article V of the GDPR takes place, and whether an effective data processing agreement including standard contractual clauses is in place. According to the “Schrems II” ruling by the European Court of Justice, it is also necessary to check whether US authorities could access the data under surveillance laws such as FISA 702 – regardless of what the provider contractually guarantees. For professional groups with confidentiality obligations, this is a crucial difference that a single marketing statement from a cloud provider cannot resolve.

The industry’s response: Data processing that remains in Germany

That’s exactly why professional users are increasingly turning to specialized AI solutions that comply with the GDPR and operate data processing on German servers. The difference is simply explained at its core: A request from a property manager for a rent increase under § 558 BGB or an investor’s question about land value development in a specific city district does not leave the European or German legal framework with such a solution. This does not replace your own duty of care – a data processing agreement and a look at the register of processing activities remain mandatory – but significantly reduces the structural risks associated with the use of non-European cloud AI.

Technical and organizational measures as a selection criterion

For property management companies and developers looking to integrate a AI solution into their ongoing operations, a glance at the provider’s marketing claims is not sufficient. What matters is whether technical and organizational measures (TOMs) in line with Article 32 of the GDPR are documented: encryption during transmission and storage, access restrictions at the employee level, logging of access and clear data deletion concepts. Anyone who includes an AI solution in their record of processing activities (Article 30 of the GDPR) – which is mandatory for most commercial users – should be able to name the server location, data processors and deletion periods without needing to ask the provider. If this is not possible, that alone is already a warning signal, regardless of how convincing the tool’s performance may be.

Which knowledge areas a specialized real estate AI can cover

The actual added value of AI tailored to the real estate industry lies not in generic language understanding, but in the depth and currency of the underlying specialist data. For professional users, four areas of knowledge are particularly relevant.

Standard land value and official valuation methods

Standard land values are to be determined and published comprehensively by the valuation committees of the municipalities and states according to § 196 of the Building Code – the exact legal text can be found under gesetze-im-internet.de. In practice, the real challenge is not finding a single value, but rather its interpretation: How has the standard land value of a location developed over the past years, how does it differ from the standard land value of the neighboring area, and how is it incorporated into a cost or income approach? Those who want to delve deeper into the methodological basics of valuation will find a detailed comparison of the common methods in the article Valuation Methods for Real Estate: 5 Approaches to Determining Value as well as a practical interpretation in the article on the Standard Land Value Map.

Rent law and current Federal Court of Justice case law

Rental law is one of the most judgment-rich areas of German law. Whether it’s rent increases, operating cost statements, termination due to personal need, or beauty repair clauses – hardly a quarter goes by without relevant decisions from the Federal Court of Justice. For property management companies that oversee hundreds of rental agreements, the timely integration of new rulings into their own contract practices is a tangible competitive factor. A specialized AI can systematically process legal precedents here – but it explicitly does not replace legal individual case analysis, instead providing the preliminary research that would have previously taken hours of manual database work.

Funding programs from the federal government, states, and municipalities

The funding landscape for energy-efficient renovation, age-appropriate renovation, and new construction is notoriously complex: KfW programs at the federal level, supplemented by state development banks and sometimes municipal grants, which constantly change in terms, deadlines, and combinability. For developers and investors planning a renovation strategy across a larger portfolio, knowledge of the currently valid programs directly determines the calculation of a project.

Market and Return Data

Purchase price factors, rental returns, vacancy rates, and price developments differ not only from city to city, but often from street to street. Investors making a location decision need reliable, up-to-date comparable values instead of press articles from two years ago.

Building regulations and heritage protection: the underestimated regional factor

Alongside standard land values, rental law, and funding programs, there is a fifth area of knowledge that is regularly underestimated in practice: the state building regulations and local heritage protection. Unlike the BGB or the DSGVO, the state building code is not uniformly regulated across the country – setbacks, parking space requirements, and energy retrofitting obligations can differ significantly between Bavaria, North Rhine-Westphalia, or Berlin. For developers operating across regions, this means: A calculation that works in one federal state may fail in another due to differing setback regulations or an additional approval requirement for protected building structures. A specialized real estate AI based on current, state-specific expertise can provide an initial orientation before the costly detailed review by an architect or the building authority takes place.

Practical example: An investor examines an apartment building

A concrete scenario makes the difference tangible. An investor receives the brochure of a multi-family house with twelve residential units in a medium-sized western German city. The asking price is 24 times the annual net cold rent. Before making a purchase decision, he wants to gain a first solid assessment within a few hours – not days – before commissioning an expert for a full inspection.

In the classical approach, this means: calling the local valuation committee or submitting an application via the state’s standard land value portal, reviewing the city’s rental index, researching comparable sales via real estate agent networks or paid market databases, checking whether energy efficiency renovation obligations under the Building Energy Act apply, and assessing whether funding is available for a planned modernization. Depending on the availability of contacts, this process can take several days.

With a specialized, GDPR-compliant real estate AI, the same preliminary check can be initiated in a condensed form: The investor provides location, year of construction, living area, and asking price, and receives an assessment of the price-to-value ratio in comparison to the development of the standard land value in the location, a summary of the rent range for comparable apartments, a note on relevant energy requirements, and a first overview of possible funding programs for a future renovation. The result does not replace a full inspection by an expert – but it provides a reliable basis within a short time to decide whether the effort of a more detailed inspection is even worthwhile. That is where the real efficiency gain lies: not in replacing professional expertise, but in pre-filtering the cases that are actually worth this expertise.

Second example: Operating cost review for a property management

A second, just as everyday scenario concerns a property management company with around 40 managed condominium associations. At the end of the year, owners increasingly ask about operating cost items that have notably increased compared to the previous year – for example, insurance premiums or caretaker costs. Traditionally, a clerk would have to manually go through the billing history for each inquiry, check the relevant BGH case law on the allocability of individual cost items, and formulate the response individually – a time-consuming process that can take weeks for several hundred units.

A DSGVO-compliant real estate AI can support this here without the need to transfer names or individual contract data of the owners into an external system: The inquiry can be reduced to a purely factual question – for example, “Is a 30 percent increase in building insurance compared to the previous year generally assessable as a charge, and which legal precedents are relevant?” – and provides a structured professional basis for the administrator’s individual response. The personal assignment to the specific community of owners remains entirely in the hands of the property management and is not even entered into the AI system. This principle – separating factual questions from personal data wherever possible – is one of the most effective data protection measures and is simultaneously applicable regardless of which specific AI provider is used.

Classic Research vs. AI-Supported Research Compared

The difference between classic Google research and specialized real estate AI rarely lies in the general availability of information – most data is somewhere publicly accessible. The difference lies in effort, bundling, and contextual understanding.

The table clearly shows that this is not a competition between “human versus machine,” but rather a shift in effort: AI takes over the time-consuming bundling and pre-structuring, while humans retain the professional final decision and responsibility for the final evaluation.

Strategic Outlook: How does the role of real estate agent, appraiser, and advisor change?

The obvious concern of many market participants is that AI could make real estate agents, appraisers, or advisors obsolete. This concern is too short-sighted. A more realistic scenario is a shift in roles: from pure information provider to organizer and decision companion.

From Information Provider to Organizer

A real estate agent who today mainly advertises having access to standard land values or comparable properties loses their differentiation as soon as this data becomes more widely and quickly available. What remains valuable is the person who can develop a reliable negotiation strategy, a realistic price assessment taking into account local peculiarities, or a solid exit strategy for an investor from these data. The same applies to appraisers: The pure data collection for a valuation report can be accelerated, but the professional plausibilization, on-site inspection, and liability assumption remain bound to the individual.

The new trust question: Who is liable for AI misjudgments?

With increasing use of AI, the discussion on liability is also shifting. An AI-supported preliminary assessment does not replace a valuation by a publicly appointed and sworn expert and must not be communicated as such. Anyone in property management, real estate agency, or advisory roles who forwards AI-generated outputs to third parties without review still assumes professional responsibility for them. Therefore, reputable providers explicitly label AI-supported assessments as guidance rather than legal or valuation opinions. This clarification is not a legal footnote, but a prerequisite for AI tools to be trusted within the industry at all.

New Task Distribution in the Team

In larger property management companies and construction developers, a division of tasks is becoming apparent, where standard inquiries – such as regarding subsidies or common rental law issues – are increasingly prepared using AI, while experienced employees focus on complex individual cases, negotiations, and communication with property owners. This changes job profiles rather than reducing positions – the demand for real estate expertise does not decrease due to AI, but shifts towards competencies in assessment and advisory services.

Macro Perspective: AI as a Competitive Factor on the German Real Estate Market 2026+

From a market perspective, access to faster, more reliable, and legally secure data analysis is increasingly becoming a competitive factor. In a market where purchasing decisions are often made under time pressure, the person who receives a reliable initial assessment not after several days but within hours gains an advantage. This applies equally to investors in bidding competitions, to developers calculating land purchases, and to property management companies, which are increasingly judged on their digital advisory quality in the competition for management mandates.

Regulatory Double Beat: EU AI Act Meets GDPR

Parallel to the technological development, the regulatory framework is becoming more stringent. With the EU AI Act, a second regulatory framework is emerging alongside the GDPR, which particularly addresses risk-prone AI applications. For the real estate industry, it is relevant that creditworthiness checks or automated decision proposals in financing may be subject to stricter transparency and documentation obligations in the future. Pure knowledge and research support, as described in the context of this contribution, typically does not fall into the highest risk category of the AI Act – but it also benefits if providers provide transparency about training data, limitations of applicability, and server location upfront, rather than disclosing them only upon request. Those who already rely today on a GDPR-compliant, Germany-based AI infrastructure thereby gain a regulatory advantage over competitors relying on opaque cloud solutions from third countries. A current assessment of data protection authorities regarding AI applications can also be found on the pages of the Data Protection Conference, the body of German supervisory authorities.

What this means for pricing on the market

Efficiency gains from AI-supported research will also affect pricing on the real estate market in the medium term. The faster and more reliably market participants can assess standard land values, comparable rents, and subsidy conditions, the narrower the information gap becomes between professional large investors and smaller private investors, who have previously often only had access to comparable depth through paid market databases or expensive appraisals. This should lead to a certain alignment of the information base in the coming years – without the actual evaluation and negotiation skills losing their importance. On the contrary: If basic information becomes available to all market participants more quickly, the competitive advantage will shift even more towards the ability to draw the right strategic conclusions from this information.

Who builds the lead now will defend it in 2030

Technological advances in the real estate industry rarely emerge overnight, but rather cumulatively: Those who start today to bundle evaluation, legal, and subsidy knowledge with AI-supported tools build up a data set and a workflow over several years that is difficult to catch up with later. Conversely, market participants who completely forego AI due to data privacy concerns risk a long-term efficiency disadvantage compared to competitors who have found a legally secure solution. The actual strategic decision is therefore not “AI yes or no,” but “which AI infrastructure meets the compliance requirements of my industry.”

Exactly at this point, Lukinski’s own AI tool comes into play, accessible at lukinski.de/ai/. It is designed as a real estate knowledge assistant that answers questions about standard land values, valuation methods, rental law, funding programs, and market data – compliant with the GDPR and operated on German servers. For investors, property management companies, and developers, this means that the preliminary check described in the above practical example can be initiated without sensitive inputs being sent to an uncontrolled US cloud. The tool does not replace an expert, legal advice, or a notarial review – but it does shorten exactly the pre-research phase, which in the classical approach takes up the most time.

Looking ahead to the coming years, it can be anticipated that GDPR-compliant real estate AI will become a standard expectation, rather than a differentiating feature. Those who actively shape this transition, rather than waiting for it, secure an advantage that is hardly catchable in a data-driven market.

An eviction for personal use occurs when a landlord terminates the lease because they need the apartment for themselves, family members, or members of their household. The legal basis is § 573 Abs. 2 Nr. 2 BGB. Personal use is the most common reason for termination in German rental law.

Requirements: When is personal use effective?

For an eviction for personal use to be legally valid, all of the following points must be met:

• Authorized person: The landlord themselves, their spouse, life partner, children, parents, siblings, nieces, nephews, in-laws, or people from the landlord’s household (e.g. a caregiver).
• Legitimate need: The need must be specific and verifiable – not a “stockpiling need”. Examples: Moving for age-related reasons, separation, new job in the city.
• Proportionality: The apartment size must match the registered need. A 200 m² apartment for a single person without a special justification can be invalid.
• Font style: The termination must be in writing and clearly state the reason of personal need.

Notice periods for personal need

The periods depend on the duration of the rental agreement:

The notice must arrive by the third working day of the last month before the notice period expires.

Who has special protection against termination?

Certain tenant groups have increased protection – the court may reject a personal need termination even if it is valid:

• Tenants over 70 years old with a long rental duration
• Pregnant women and families with young children
• Tenants with severe disabilities
• Tenants with a long rental duration (20+ years), if no comparable alternative is available

The court then weighs the interests of the landlord and tenant against each other. A termination can still be enforced – but it takes longer.

What must the termination include?

The notice of termination for personal use must include the following details:

• Name and address of the landlord
• Name and address of the tenant
• Designation of the apartment
• Exact termination date (date at the end of the month)
• Specific reason for personal use: Who is moving in (name, family relationship) and why this particular apartment is needed
• Signature of the landlord

Missing or vague reasons lead to the termination being invalid.

Tenant’s objection

The tenant can object to the termination in writing up to two months before the end of the lease if the termination would cause an unreasonable hardship for them or their family. The court then decides whether the hardship grounds are present.

Fabricated personal use: What consequences does this pose for the landlord?

Anyone who fabricates a need for personal use risks significant consequences:

• Compensation for damages: The tenant can claim moving costs, higher rent in the new apartment, and other damages.
• Criminal report: In severe cases, fraud (§ 263 StGB) may be considered.
• Duty to review: Landlords are obliged to offer the apartment to the tenant if the need for self-occupation ceases.

Alternatives to termination for self-occupation

• Termination agreement: Mutual termination with a severance payment – often the faster and cheaper solution.
• Rental agreement with time-limited tenancy: When reletting, agree on a fixed-term rental agreement with a self-occupation clause.
• Modernization termination: If comprehensive renovation is planned – own section (§ 573 Abs. 2 No. 3 BGB).