Products · Prism
Before you build it, buy it, or bet on AI. Know what matters most.
You are about to spend serious money building something, buying a platform that nearly fits, or betting on AI. Prism tells you which of those is right for your business, and what to do first, before you commit to any of it.
Two weeks. Fixed scope. We will tell you not to proceed if that is the answer.
What Prism is
You are about to spend real money building something, buying a platform, or betting on AI, and you are not certain it is the right call. Prism does the homework across your product, your numbers and your setup, then gives you a straight answer in two weeks: build it, buy it, or not yet, and what to do first.
An independent two week go or no go on the technology call you are about to make. Typically 10 to 13 business days from the point we start.
Not your dev shop
They tell you whether they can build it. They are paid to build it. They will not tell you whether you should.
Not a gut check call
An hour on a call reframes the question. It does no homework inside your actual business.
Not a consultancy deck
A large firm takes weeks and hands you a presentation. You need a decision, not a document.
Does this sound like you?
- You are weighing whether to build something yourself or buy a platform that nearly fits.
- Everyone is telling you to invest in AI and you cannot tell which part is real.
- Your engineers say they can build it. You are not sure that answers the question.
- You are about to replace a system that mostly works, and the cost keeps growing.
- The commitment is large and hard to walk back once it starts.
If two or more of those are true, the decision is worth two weeks before you commit.
What you get
The answer comes first. Everything else is there to support it.
The answer, first
Build, buy, or not yet, stated plainly on the first page, with how confident we are and what would change it.
The case behind it
What you believed going in, what the evidence showed, where those differed, and the trade offs made explicit rather than hidden.
What to do first
The first moves that matter and what they unlock, what is valid but should wait, and what looks attractive and should not take resources now.
The numbers per move
Cost, dependencies, downside and cost of delay, where the evidence supports an estimate and marked clearly where it does not.
A live session with your leadership
Sixty to ninety minutes walking your team through the decision and defending it, recorded so you can share it.
Everything packaged
The full record, the working files and an editable tracker, in your hands within 24 to 48 hours of the session.
What we need from you
A few hours of your time in total. The work is ours to carry.
- One short kickoff session with you and the people who hold the relevant parts.
- A few focused conversations across the two weeks.
- Access only to what the decision actually touches, nothing more.
- Ninety minutes at the end for the live session.
What we will tell you
- We will tell you not to build it, if that is what the evidence says.
- We will tell you the evidence is not enough to decide, rather than inventing certainty.
- We will tell you when a cheaper specialist is the right answer.
- We do not sell you the build afterwards. Nothing we recommend pays us more.
When Prism is the right call
- The decision commits meaningfully more than the engagement does.
- It depends on more than one part of the business, so no single specialist can settle it.
- It is genuinely still open, and expensive to reverse once it starts.
- You will act on what the evidence says, including being told to wait.
When it is not
- A decision you have already made and want confirmed.
- A narrow question one expert could answer on their own.
- A general review of the whole business with no specific move at stake.
- The building itself. Prism ends at the decision.
If one of those describes you, tell us anyway and we will point you somewhere better.
Questions
- How much does it cost to build custom software?
- Published build ranges run from roughly $25,000 for a narrow MVP to $400,000 or more for an enterprise platform, and the average custom software project ran about $132,000 and took 13 months. Those numbers are close to useless on their own, because the same brief is routinely quoted at $40,000 by one firm and $400,000 by another. The spread is driven by who is actually writing the code, how tightly the scope is defined, and whether testing, security and integration work are inside the quote or waiting outside it. The useful question is not the headline number, it is what the quote excludes.
- What does it really cost once it is running?
- This is the number that gets missed. Annual maintenance commonly runs 15 to 25 percent of the original build cost, every year, and one analysis puts the initial build at under 35 percent of five year total cost. Integration work commonly runs 35 to 50 percent of initial development cost and is the most frequent source of overrun. Anything with AI in it carries inference and retraining costs that scale with usage rather than with team size. Ask any vendor for the three year number, not the build number.
- How long does it take to build custom software?
- A focused MVP takes roughly 8 to 14 weeks, a full web application 3 to 6 months, and enterprise work 6 to 18 months. Build projects run about 38 percent over on cost and 42 percent over on timeline, so treat any quoted date as the optimistic case unless real discovery sits behind it.
- Should we build, buy, or partner?
- Partnering is the option most comparisons leave out, and it is often the right one: buy a platform with a real API and build only the part that is genuinely yours. The honest test is whether the thing in question is your competitive advantage or a utility. Build what differentiates you. Buy what does not. Partner where the answer is some of each. Most teams get this wrong by treating a utility as a differentiator because building it feels more satisfying.
- How do you actually decide?
- Price all of it, not just the build: the five year running cost on both sides, the integration surface, what your team stops doing while they build, and the cost of being wrong. Then ask who gave you each number and what they earn from it. One figure is worth knowing: Projects that did real discovery before committing land within about 10 to 15 percent of budget, while projects that skipped it run 30 to 60 percent over. The homework is not overhead. It is the cheapest part of the decision.
- Can I trust the cost comparisons I find online?
- Read who wrote them. Almost every page ranking for build versus buy is published by a software agency, which means the party explaining your decision is also the party paid for one of the answers. That does not make their numbers wrong, but it does mean nobody in that search result is neutral. We are not selling you the build either way, which is the only reason our answer here is worth more than theirs.
- What if the answer is not to proceed?
- That is a common and entirely legitimate outcome, and it is usually where the money is saved. We would rather tell you to stop than be comfortable.
- How much of my time does this take?
- A few hours in total: one kickoff session, a handful of focused conversations with the people who hold the relevant parts, and the live session at the end. The work is ours to carry.
- Is this right for every decision?
- No. If one specialist can answer your question, a specialist is cheaper and better, and we will say so. This is for the call that depends on several parts of the business at once.
Talk through the decision you are facing
Book a call if you would rather talk it through, or send a few lines and we will reply. Either is fine, and neither commits you to anything.