By Mina Choi · Reepa Solutions
The most common question in our initial conversations is not "How does it work?" but "What will it cost?" And the most honest answer we can give starts with a counter-question — not as a sales tactic, but because two projects with the same working title can differ by a factor of ten. Nevertheless, the desire for a ballpark figure is entirely reasonable: management and CFOs need an order of magnitude before they can think further. This article provides exactly that — realistic hourly rates by region and role, contract models with their strengths and weaknesses, project cost ranges for the most common types of work, an honest breakdown of hidden costs, and a look at the AI effect and current funding programmes. Anyone who finishes this article with a solid ballpark for their own project has achieved the goal. For context on the bigger picture, see our Software Development Guide for the Mid-Market.
Why the question is rarely answered directly
Anyone researching online finds endless cost calculators but hardly any concrete numbers. There are three reasons for this, and all three are understandable. First, the term "software" is enormously elastic — a five-page landing page and an industrial IoT platform both fall under it, yet the latter can cost a hundred to a thousand times more. Second, the price depends heavily on non-functional requirements that are almost never mentioned in the first conversation: availability, scalability, multi-language support, accessibility, compliance, audit trails, data protection impact assessments. These requirements double the budget in many projects without changing the visible features at all. Third, the hourly rate is only half the story — a good senior developer often completes the same task in half the time of a cheaper junior, making the end result less expensive despite the higher rate.
Reputable vendors do not sidestep this dilemma; they address it openly: they quote hourly rates and typical project ranges, propose a brief discovery phase, and only then provide a concrete range with documented assumptions. Anyone quoting a fixed price without a scope is either building in a risk buffer of 30 to 50 percent, or will cut corners on the outcome later — usually on tests, documentation, and hardening. Both outcomes are expensive for the client; the bill just arrives after launch.
From our experience: in roughly 70 percent of enquiries, a range with a factor of two can be named after a two-hour initial conversation. After a discovery phase, that range typically narrows to a factor of 1.3 and can be converted into a fixed price or a cap-and-collar model.
Hourly rates in 2026 by region
The global hourly rate landscape has shifted since 2020. Nearshore providers from Central and Eastern Europe have become more expensive. DACH consultancies are holding their rates steady because the shortage of skilled professionals offsets the efficiency gains from AI tools. Offshore providers from Asia have remained affordable at the lower end, while tier-1 providers have moved closer to nearshore rates.
| Region / Model | Hourly rate (Mid–Senior) | Advantages | Risks |
|---|---|---|---|
| DACH in-house (fully loaded) | 90–160 € | Domain knowledge, direct availability, cultural proximity | Difficult recruiting, high fixed costs, limited upward scalability |
| DACH consultancy / agency | 130–200 € | Rapid scaling, specialist pool, delivery accountability | Highest price level, staff turnover within the team possible |
| Nearshore Europe (PL, RO, PT, ES) | 50–90 € | Compatible time zone, good English, EU law, strong senior quality | Language barriers in specialist domains, travel required for workshops |
| Offshore (IN, VN, PH) | 25–45 € | Lowest price, large talent pools, 24/7 models possible | Time zone differences, high coordination overhead, wide quality variance |
The table shows gross figures that clients actually pay — not developer salaries. Fully loaded in-house costs include gross salary, social security contributions, workspace, tooling, training, and a realistic utilisation factor of 70 to 80 percent. A company hiring a senior at a gross salary of 100,000 euros pays roughly 130,000 euros per year as an employer — at 1,500 productive hours that works out to 87 euros per hour. For many mid-market companies the real figure is closer to 110 to 130 euros due to meetings, leave, and recruiting overhead.
A note on the offshore segment: the hourly rates are tempting, but total cost of ownership often is not. We have seen multiple projects where offshore code had to be replaced after 18 months. This is not a categorical statement, but a reminder that the hourly rate differential alone is not a sound basis for a decision. For a deeper look at the selection model, see our article on Nearshore vs Offshore vs In-House.
Hourly rates by role
Within a region, hourly rates vary considerably by role and experience. The following figures apply to DACH consultancies with delivery accountability — nearshore and offshore rates are typically 30 to 60 percent lower; fully loaded in-house costs are 20 to 40 percent lower.
| Role | DACH consultancy hourly rate | Typical responsibilities |
|---|---|---|
| Junior developer (0–2 years) | 80–110 € | Implementing well-defined tickets under guidance, tests, smaller features |
| Mid developer (2–5 years) | 110–150 € | Independent feature delivery, code reviews, smaller architectural contributions |
| Senior developer (5–10 years) | 140–200 € | Module ownership, architectural decisions, mentoring, stakeholder communication |
| Tech Lead | 170–230 € | Team leadership, technical roadmap, interface between product and management |
| Software Architect | 190–260 € | System architecture, technology selection, risk analysis, cross-cutting standards |
| UX/UI Designer | 110–170 € | User research, wireframes, design system, interaction concepts |
| DevOps / SRE | 130–190 € | CI/CD, infrastructure, observability, incident response |
| Project Manager | 120–180 € | Scope, budget, risk, stakeholders, reporting |
An observation from numerous audits: teams composed exclusively of senior profiles are not necessarily the most productive. A healthy team mix of 1 senior to 2–3 mid-level developers and the occasional junior delivers the best balance of quality, speed, and hourly cost in almost every project. If you receive a proposal in which 80 percent of hours are allocated to senior roles, it is worth asking whether the tasks genuinely require that.
Contract models: fixed price vs T&M vs cap-and-collar
The choice of contract model determines who bears which risk. Three models dominate the market; each has clear strengths and clear weaknesses.
- Fixed priceThe vendor bears the delivery risk; the client bears the requirements risk. Suitable for clearly scoped engagements — marketing sites, standard migrations, integrations with known systems. A detailed specification is a prerequisite; without one the vendor will price in a buffer of 30 to 50 percent, or change requests during the project will erode the model entirely.
- Time-and-Material (T&M)The client bears the full risk; in exchange there is no buffer. Suitable for exploratory work — MVPs, new products, complex integrations, R&D. Weekly reporting, burn-down charts, clear backlog management, and a stop mechanism for budget overruns are essential. Without this discipline, T&M contracts spiral out of control in almost every project.
- Cap-and-CollarA blended model: T&M billing with a bilateral cap. If effort exceeds the cap, the vendor absorbs the extra work at their own cost. If effort falls below the collar, the client benefits. Suitable for projects with a moderate share of unknowns. A shared discovery workshop is required to make the range reliable — without it, both parties end up in a conflict.
From our consulting practice, the most common and sustainable structure for the mid-market is a two-stage model: first, a priced discovery phase of one to three weeks with a clear deliverable (requirements document, architecture sketch, risk register, cost estimate); second, a cap-and-collar contract for delivery with shared skin-in-the-game on both sides. Clients who insist on a fixed price from the outset either pay a buffer premium or receive lower quality. Clients who accept T&M from the outset risk budget overruns. The blended model addresses both risks.
Project costs by type
The following ranges are 2026 market rates for engagements with a DACH consultancy or a combined DACH/nearshore model. They cover discovery, design, development, testing, and initial launch — maintenance and further development are calculated separately.
| Project type | Cost range | Typical duration | Team size |
|---|---|---|---|
| Landing page with CMS | 5,000–15,000 € | 2–6 weeks | 1–2 people |
| Marketing site (multilingual, strong SEO) | 15,000–30,000 € | 6–10 weeks | 2–3 people |
| MVP of a web application | 30,000–80,000 € | 2–4 months | 2–4 people |
| Mid-size web app / customer portal | 80,000–250,000 € | 4–9 months | 3–6 people |
| SaaS product to launch | 150,000–500,000 € | 6–14 months | 4–8 people |
| Enterprise software / B2B platform | 300,000–2,000,000 €+ | 9–24 months | 6–15 people |
The ranges are wide, and that is not hedging — within each category, real projects genuinely differ by a factor of two to four. An MVP of a web application with standard auth, three models, and a simple interface lands between 30,000 and 50,000 euros. The same MVP with multi-tenancy, audit trail, real-time features, and initial AI components lands between 60,000 and 80,000 euros — with the same consultancy, the same language, and the same people. The difference is in scope, not in the vendor. Understanding the range logic allows you to actively manage it: through clear prioritisation, pragmatic requirements, and a shared understanding of "good enough for phase 1". For the specific mechanics of MVP projects, see our article on MVP Development in 90 Days.
Realistic cost estimate for your project
Do you have an idea or a concrete project? In a 30-minute initial conversation we will place your scope within a range, identify the biggest cost drivers, and propose a suitable contract model — free of charge and without obligation.
Request a free cost estimateHidden costs — the items most often overlooked
In most enquiries we receive, the pure development effort is estimated realistically — the surrounding hidden items are not. These five categories are missing from almost every budget draft and end up costing 20 to 50 percent more than expected:
- Discovery and requirements clarificationWorkshops, user interviews, competitive analysis, architectural decisions before the first line of code. Realistically 5 to 15 percent of the build budget, and up to 20 percent in complex domains. Cutting corners here costs significantly more in the build phase.
- Design and user experienceWireframes, clickable prototypes, design system, micro-interactions, accessibility. Realistically 10 to 20 percent of the build budget. For customer-facing products often 25 percent, because usability quality has a direct impact on business outcomes.
- Testing and quality assuranceUnit tests, integration tests, end-to-end tests, manual QA, penetration testing before launch. Realistically 20 to 30 percent of the build budget. Cutting this to 5 percent produces a product with a high defect rate and expensive maintenance.
- Deployment, infrastructure, and observabilityCI/CD, cloud setup, monitoring, logging, backups, disaster recovery, compliance configuration. Realistically 10 to 20 percent of the build budget, plus ongoing cloud costs of 200 to 5,000 euros per month depending on load profile.
- Refactoring and technical debt reductionTechnical debt inevitably accumulates during development and must be addressed in the first 6 to 12 months after launch. Planning for this prevents the classic pattern where feature velocity drops to 30 percent after 18 months. Realistically 10 to 15 percent of the build budget in the first year of maintenance.
The sum of these five items amounts to 55 to 100 percent of the pure development effort — in other words, the total project budget is typically 1.5 to 2 times what appears as "development" in an initial estimate. Ignoring this rule of thumb means under-budgeting the project from the outset.
Annual maintenance costs — a realistic view
Software is not a piece of furniture. Security updates, dependency management, new requirements, and compliance changes generate ongoing costs. The benchmark that has held for decades: 15 to 25 percent of the original build cost recurs annually for maintenance and further development.
The lower end (15 percent) applies to products with a low rate of change — internal tools, marketing sites, classic migration products. The upper end (25 to 30 percent) applies to customer-facing SaaS products that continuously require new features, undergo regular security audits, and run on a shifting platform landscape. A SaaS product with 300,000 euros in build costs typically generates 60,000 to 90,000 euros of maintenance and development costs per year — this is not a penalty, but the prerequisite for a product that remains competitive.
A common trap: clients budget maintenance low for the first two years because the product is "brand new". Reality: the most refactoring and feature catch-up work happens in the first two years, because actual usage patterns only become visible after launch.
The AI coding effect — are prices falling or not?
Since 2024, AI-assisted programming has become embedded in productive workflows. Copilot, Claude Code, Cursor, and comparable tools are standard in most consultancies. The legitimate question from clients: does this show up in prices?
Studies from 2025 show productivity gains of 15 to 30 percent on routine tasks — boilerplate, tests, documentation, simple refactorings. For architectural decisions, domain logic, and performance optimisation, the effect is below 10 percent. Hourly rates have barely moved: efficiency gains are reinvested in higher quality, AI licences cost 30 to 100 euros per developer per month, and consultancies absorb the effect in better margins because the skills shortage keeps prices stable.
For clients this means in practice: the same price, more output. Anyone commissioning a project in 2026 typically gets better test coverage, more complete documentation, and shorter delivery times than in 2022 at a comparable hourly rate.
Funding — Digital-Jetzt, go-digital, innovation vouchers
For German mid-market companies, several active funding programmes are available in 2026 that support software projects with grants of between 30 and 70 percent. The most important ones:
- Digital-Jetzt (BMWK) — Grant of up to 50,000 euros, in special cases up to 100,000 euros, for digitalisation projects in SMEs. Funding rate 30 to 50 percent depending on company size and location.
- go-digital (BMWK) — Advisory grant for SMEs, covers up to 50 percent of consulting day rates, maximum 16,500 euros per project. Suitable for digitalisation strategy, IT security, and digital market development.
- Innovation vouchers — State programmes in almost all German federal states, funding technological consulting and development with amounts between 2,500 and 35,000 euros depending on the state and programme tier.
- ZIM (Central Innovation Programme for SMEs) — Funding for research and development projects, grants between 25 and 55 percent, typical project volumes of 100,000 to 550,000 euros.
Funding does not come automatically — application, business case, proof of use, and fund drawdowns consume time that is not covered by the grant itself. Realistically, 5 to 10 percent of the grant volume represents additional effort for the application and billing process. Even so, the effort pays off for most of our mid-market clients — a 50 percent grant on a 100,000-euro project represents a significant saving even after deducting application overhead.
The Reepa approach — discovery before fixed price
Our standard model for mid-market clients is two-stage. Step one is a one-to-three-week discovery phase at a fixed price (typically 4,500 to 18,000 euros depending on depth). The deliverable includes a requirements document, an architecture sketch, a risk register, and a cost estimate with a factor-1.3 range. Step two is delivery under a cap-and-collar contract with shared skin-in-the-game. If effort exceeds the cap, we absorb the extra work. If it falls below the collar, we share the benefit with the client.
This model has produced a hit rate of over 90 percent over the past 36 months. For clients it means planning certainty without paying the 30- to 50-percent buffer of a classic fixed-price model.
Anyone considering starting a software project — whether a marketing site, MVP, customer portal, or SaaS product — should choose a vendor who is willing to work in this order: first clarify scope, then set the price. Vendors who name a price before the scope are optimising for their own risk, not for the outcome. For more on the choice between custom and standard solutions, see our article on Custom Software vs Off-the-Shelf Software.
Frequently asked questions
Why do so few vendors quote a direct price for software development?
Because without a defined scope, no honest price can be given. Two projects with the same working title — say, "customer portal" — can differ by a factor of ten depending on the data model, integrations, compliance requirements, and non-functional properties such as scalability or multi-language support. Reputable vendors quote hourly rates and project ranges, but only provide a concrete figure after a brief discovery phase. Anyone quoting a fixed price without a scope is either building in a risk buffer of 30 to 50 percent, or will cut corners on the deliverable later.
What is a realistic hourly rate for a senior developer in Germany in 2026?
Senior developers at specialist consultancies in the DACH region are priced between 140 and 200 euros per hour in 2026. In-house senior developers cost employers roughly 90 to 130 euros per productive hour on a fully-loaded basis — assuming a gross salary of 85,000 to 110,000 euros plus ancillary costs and a utilisation factor. Nearshore seniors from Poland, Romania, or Portugal are priced at 65 to 95 euros per hour; offshore from India or Vietnam at 30 to 50 euros — with significantly higher coordination overhead.
Is a fixed price or time-and-material the better choice?
Fixed price makes sense for clearly scoped engagements with low unknowns — marketing sites, standard integrations, classic migrations. Time-and-material is appropriate for exploratory projects where requirements sharpen during development — MVPs, new products, complex integrations. For mid-market clients, a blended model is often ideal: a clearly priced discovery phase, followed by a cap-and-collar contract with capped hours and shared risk on both sides.
What does a typical SaaS product cost from scratch to launch?
A market-ready SaaS product with multi-tenancy, billing, authentication, and meaningful feature depth sits between 150,000 and 500,000 euros to production launch in 2026. The range depends heavily on maturity — a public beta with a modest feature set is achievable at the lower end; a product with compliance modules, audit trail, multi-language UI, and native mobile apps lands at the top. After launch, typically 20 to 30 percent of build costs recur annually for maintenance and further development.
Does AI-assisted coding noticeably reduce development costs?
Partly, but not as dramatically as marketing claims suggest. Studies from 2025 and real-world experience in 2026 show productivity gains of 15 to 30 percent for routine code, tests, and documentation. For architectural decisions, integrations, and complex domain logic, the effect is well under 10 percent. Hourly rates have barely moved — consultancies reinvest efficiency gains into higher quality, more tests, and faster delivery rather than price reductions. For clients this means: the same price, more output — not a lower price for the same output.
Ready to estimate your software project realistically?
Let's talk for 30 minutes with no obligation. We will place your scope within a range, identify the biggest cost drivers, propose a suitable contract model, and check whether funding programmes such as Digital-Jetzt or go-digital apply to your project.
Schedule a 30-minute conversation
IT architect and software engineer with over ten years of experience delivering mid-market software projects — from MVPs to enterprise platforms. Writes regularly on software architecture, development economics, and contract models for clients.
Reviewed: 22 May 2026 · More about Mina