Clients come to you with visions. Sometimes those visions are achievable within the budget and timeline. Sometimes they'd take three times the money and four times the time. The skill is in honoring the vision while guiding toward something that's actually buildable, without crushing their excitement in the process.
Why this matters
Dismissing a client's vision as unrealistic damages the relationship and makes them feel stupid for having ambition. Blindly agreeing to an impossible vision leads to a failed project and broken trust. The middle path, which is where you want to be, is showing them a realistic route to something they're excited about.
The principles
Validate the vision first. Before you talk about constraints, show that you understand and appreciate what they're trying to do. "I love that idea, personalized experiences drive real engagement" lands very differently from "that's going to be expensive."
Explore the "why." What are they really trying to achieve? The specific feature request might be unrealistic, but the underlying goal is usually achievable through a different approach.
Propose phases. "We can't build all of this by October, but we can launch Phase 1 with the core of what you want, then build toward the full vision." Phasing turns an impossible project into a series of possible ones.
Focus on outcomes, not features. The client doesn't actually want "AI-powered recommendations." They want higher engagement and more sales. There might be a simpler path to those outcomes.
What this looks like
Client: "We want an AI-powered recommendation engine that learns from user behavior in real-time and personalizes everything."
"I love that vision. Personalized experiences drive real engagement, and this could be a serious differentiator. Let me help us think about how to get there.
The full vision, real-time AI learning with comprehensive personalization, would take 6+ months and a significant budget to build well.
Here's what I'd propose:
Phase 1 (doable in our timeline): rule-based recommendations using user preferences and popular items. This gives you personalization on day one and starts collecting the behavior data we'd need for smarter algorithms.
Phase 2 (3 months post-launch): introduce machine learning trained on the real user data we've been collecting.
Phase 3: the full real-time adaptive AI you described.
You launch with personalization immediately, and each phase builds toward the full vision using real data instead of guesses. Does that path work?"
Why It Works
Honors the vision. Shows the realistic path. Offers immediate value. Each phase builds logically toward their goal.
Tips
- Always validate the vision before discussing constraints
- Ask "why" to understand the goal behind the feature request
- Offer phased approaches as your default response to ambitious visions
- Show creative alternatives that achieve similar outcomes differently
- Use "yes, and" thinking: build on their ideas rather than shutting them down
- Focus on user outcomes they care about, not technical feasibility
How this connects
This combines handling unrealistic requests (redirecting diplomatically), explaining constraints (why certain things take time), managing expectations (what's achievable when), and showing enthusiasm (staying excited about their vision while being honest about the path).
Things to try
- Next time a vision exceeds what's feasible: validate first, then propose phases.
- Practice: "I love that vision. Here's how we get there..."
- Build a sense of what things typically cost and how long they take, so you can set realistic expectations in the moment.
- Focus conversations on outcomes: "What result are you hoping for?" rather than features.