Author: Géza Nagy, Head of Sales Technical Analysts at Attrecto
In custom software development, we have reached a singularity of sorts. The industry playbooks followed for years – or even decades – have vanished, and the new rule is that there are no set rules.
Everyone in tech talks about how AI transforms engineering workflows, but there is one area where people stay surprisingly quiet despite it being where the fiercest battles are fought: scoping, estimation, and pricing.
Over the past year, since we embedded AI-native planning and engineering workflows deep into our daily operations, we have faced an intriguing paradox:
- The relieved: Clients who previously shelved business-critical digital initiatives because traditional agency quotes were simply out of reach. Seeing that a senior team backed by modern tooling can deliver their roadmap within a realistic budget brings immediate relief and genuine partnership.
- The skeptics: When estimates come in significantly leaner than conventional market benchmarks, natural hesitation creeps in. It’s not that they don’t appreciate lower costs. Legacy market habits make them wonder: “Do they truly grasp the complexity? Can this level of quality be sustained long-term?”
The Presales Shift: Collaborative Design from Day Zero
This dynamic has fundamentally rewritten our presales process. The classic formula of “send us a brief, and we’ll send back an arbitrary quote” is dead.
Presales is no longer an administrative or sales step. It is the very first phase of active collaboration:
- Joint exploration, system design, and scoping start immediately after the initial conversation.
- We don’t just put numbers on the table. We demonstrate deep architectural competence and concrete technical answers as early as possible, to show why the estimate is realistic.
- We break down the methodology with complete transparency: where we rely on the judgment of seasoned senior engineers, where AI accelerates boilerplate delivery, and why that balance protects long-term stability.
Market Chaos: Which game are you playing?
The rules of engagement are hazy right now, and market players are operating in entirely different realities:
- Those in denial: “Legacy” agencies pretending the massive technological leaps of recent years never happened. They still bill standard, boilerplate phases at bloated man-day rates, for work an experienced engineer delivers in hours using modern toolchains.
- The cowboy approach: Freelancers and micro-studios shipping raw, uncontrolled AI-generated code. They offer rock-bottom prices, but leave clients stranded without system architecture, security standards, or maintainability.
- Client uncertainty: Decision-makers are stuck in the middle. Developers and PMs hold widely divergent views on what efficiency gains are realistic, and clients often have to weigh bare-bones freelancer pricing against the security of an agency with a 16-year track record.
No One-Size-Fits-All: Every Project Demands a Custom Workflow
AI is an exceptional tool, but without deep domain expertise and senior oversight, it doesn’t create real efficiency, it just simply generates technical debt faster.
That’s why our scoping is always tailored to each client’s actual business reality:
- Tailored workflows: An early-stage startup validating an MVP needs a completely different approach and risk profile than an enterprise platform burdened with complex legacy integrations.
- Exactly what is needed: We avoid over-engineering simpler challenges, but never cut corners on architecture when resilience is critical.
- Passing the savings forward: We don’t pocket the hours saved by modern tooling as inflated margins. We pass those gains directly to our clients, because in an unpredictable, shifting market, transparency is what builds lasting trust.
The software landscape has transformed, and the old benchmarks aren’t coming back. Credibility is no longer measured by how many billable hours an agency can invoice, but by how transparently and effectively it co-creates value with clients from day zero.
How are you seeing this in your own domain? Have the decision-makers around you changed how they evaluate proposals and tech budgets over the past year?