I want to do something a little unusual for this essay.
Most pieces about AI and the changing cost of software are abstract. They talk about productivity gains and 10x developers and waves of hand. That is fine for thought leadership, but it is useless if you are an operating executive trying to figure out whether to actually move on a real project.
So instead of waving my hands, I'm going to walk you through a real project. A B2B portal we scoped in 2023, with the actual budget at the time, and what that exact same project would cost today. Line by line.
If you have a side bet sitting in a doc, I want you to read this with your own project in mind. Watch what happens to the numbers.
The 2023 version of the project
The brief was straightforward. A regional manufacturer wanted to give their dealer network a private portal where dealers could see live inventory, place orders, track shipments, and access marketing assets. Standard mid-market B2B work.
Here is what the engagement looked like in 2023.
Discovery and strategy: $40,000 / 6 weeks. Stakeholder interviews, dealer interviews, requirements gathering, technical scoping, written strategy doc, prioritized feature list. A senior strategist plus a project manager.
Brand and design: $60,000 / 8 weeks. UI/UX design, design system, component library, full prototype. Two designers, one design lead.
Development: $280,000 / 24 weeks. Three full-stack developers, one DevOps engineer, one QA engineer. Front end, back end, integrations to the manufacturer's ERP, authentication, deployment infrastructure.
Project management overhead: $40,000. Half-time PM across the engagement.
Launch and stabilization: $50,000 / 6 weeks. Customer migration, training materials, support documentation, post-launch fixes.
Total: $470,000. Total calendar time: 11 months from kickoff to live.
This was, by 2023 standards, a reasonable price for the scope. Not cheap. Not gouging. Market rate.
The CFO took two months to approve it. The CEO took three quarters to greenlight the actual launch. By the time the portal went live, the original strategy was already 14 months old, and the dealer landscape had shifted enough that some of the assumptions needed to be revisited.
This is the project that, in your head, you are still mentally pricing your own side bet against. Whether you realize it or not.
The 2026 version of the same project
Here is what the same scope looks like with current economics. I'm not going to soften this. I'm going to give you the actual numbers I would quote for this exact project today.
Discovery and strategy: $8,000 / 2 weeks. Same stakeholder interviews, same dealer interviews, same scoping. Faster because synthesis tools compress what used to take a strategist weeks into days. Same quality of output. One senior operator.
Brand and design: $10,000 / 2 weeks. Same UI/UX work, same design system, same prototype. Modern AI-assisted design tools produce the iteration count that used to require a team of two designers. One designer.
Development: $35,000 / 8 weeks. One senior developer using current AI-assisted coding tools, with strong architectural judgment and the right framework choices. Same scope. Same quality. Same security posture. The output that required three developers in 2023 is now achievable with one.
Project management overhead: $0. The entire project runs through a single operator. No PM layer needed when the team is one or two people.
Launch and stabilization: $7,000 / 2 weeks. Faster because deployment tooling has matured and AI-assisted documentation eliminates most of what used to be manual training-material work.
Total: $60,000. Total calendar time: 14 weeks from kickoff to live.
Same project. Same scope. Same end product. Roughly 13 percent of the original cost. Roughly 30 percent of the original calendar time.
If you are reading this and your gut reaction is "those numbers can't be right," I understand. They sound impossible because they would have been impossible in 2022.
They are not impossible now. This is just what current operators are quoting.
What this means for your specific situation
Look at the doc in your drawer. The one you've been adding to for two or three years.
When you first scoped what it would cost to build, you almost certainly used numbers closer to the 2023 column. That is what your gut and your network and your CFO and your prior experience all told you to expect. That is what you mentally priced the bet against when you decided to wait.
The numbers in the 2026 column are a different decision. A genuinely different decision.
A bet that was a $500,000 commitment with an 11-month timeline is now a $60,000 experiment with a 14-week timeline.
A bet that required pulling internal headcount or hiring a CDO to run is now a bet that runs entirely through one outside operator with no organizational disruption.
A bet that needed to be staged carefully across multiple board meetings is now a bet you could run as a discretionary spend out of an existing budget line item.
This is the conversation I want you to have with yourself. Or with me, if it would help.
The thing I want you to walk away with
The numbers in this essay are not theoretical. They are not aspirational. They are the actual cost of building, today, the kind of project that you've been quietly assuming is too expensive to start.
Your side bet is probably worth more than you think, because it costs less to test than you think.
Run the new numbers. See what the decision becomes.
Same thesis, different doorway: the number · the reframe.
If you want help re-pricing your specific project at current economics, book a call. Thirty minutes. No pitch. Just real numbers on a real idea. If you want the full 90-day program framing, we'll cover that too.