- The recommendation was not to proceed, because the integration would create more debt than value.
- Two complex systems both owning the same data means a two-way sync problem you maintain forever.
- Aim for a single source of truth: one system owns each piece of information, others defer.
- A paid, system-agnostic discovery phase de-risks the decision and can credit toward a later build.
Last week, the principal of an engineering and architecture firm called, weighing a big new project-operations platform. They'd already made up their minds; they just wanted us to wire it into the systems they already ran, so everything would stay in sync, forever. It sounded simple. It was going to hand them a second full-time job.
The word hiding all the trouble
The whole vision rested on one phrase: "so everything stays in sync." The trouble is what that requires once two substantial systems both insist on owning the same information. The new platform wanted to own projects, costs, and customers, and so did the systems they already ran. We weren't being asked to move data once. We were being asked to make both hold the same truth forever.
Two systems that must agree about the same number, forever, is one of the most expensive promises in software.
A project manager bumps a budget in the new platform on Tuesday, and minutes later the finance system updates the same project's cost. Multiply that disagreement by every field, user, and odd bit of timing, and you're no longer maintaining software. You're refereeing an endless argument.
When the two systems disagree about a project's budget, which one is right? Every answer we could build cost real money to maintain, and it'd keep costing through every future update to either system.
The software had real value, but the integration to keep the two honest would've been large and fragile, and the tax to keep it in sync was higher than the value, due every month.
So we told them not to buy it
This is uncomfortable for a firm that makes its living doing implementations, but we told them the truth: don't buy this, not in this shape and not yet. The integration would behave for a few months, then drift, and reconciliation would eat someone's days.
A firm that will tell you not to buy something is the firm you can believe when it tells you to.
Trust is the only thing we really sell. If we take your money for a project we believe is a mistake, our advice is worth nothing the day you most need it.
The principle underneath it
One idea does all the work: for any given piece of information, you want a single source of truth. One system owns a project's budget; others can read it, but only one owns it. The fix here wasn't a smarter sync engine. It was to let one system own the projects while the rest simply listened.
What we told them to do instead
The other half of the answer is a discovery phase, sometimes called a needs assessment: a paid, deliberately system-agnostic engagement with one job, to understand the problem well enough to make a sound decision. You map how the business runs and who owns what, before anyone signs a large statement of work. The plan tells you whether to build at all, and what to build, for a fraction of the cost of getting it wrong.
Spending in the low tens of thousands to truly understand a problem is a bargain if it keeps you from spending several hundred thousand on the wrong solution. Discovery is cheap insurance against an expensive mistake.
It's system-agnostic, so the people running it aren't trying to sell you a platform and can tell you the truth, even when it's "not the project you asked for." And a well-structured discovery usually credits its cost toward the implementation if you proceed, so you keep the option to walk away.
The lesson for your next big decision
The lesson isn't "never integrate" and it isn't "never build," it's to do the thinking before the spending: ask where your single source of truth should live, be honest about whether you're signing up for a two-way sync problem you'll maintain forever, and buy the diligence first.
Let's talk it through
If you're weighing a big systems decision and you want a straight, system-agnostic read on how to proceed, that's what a discovery phase is for. Reach out, and we'll tell you honestly, even if the answer is no.