We look at your data before we quote you, and here is why that matters.

The usual sales process is: you demo the system, you impress the client, you send a proposal based on what you showed them. The client reads the proposal and says yes or asks for amendments. By the time a contract is signed, both sides have made a promise they think they understand.

Then the client sends you the actual data, and that promise becomes impossible to keep.

The demo was real. Everything worked. The client saw the AI reading their data, extracting figures, building reports. But the demo ran on either synthetic data or a carefully curated sample. The real data, when you finally see it, is different in every material way.

The demo is not the reality

A demo is a controlled environment. It showcases the system’s best case. The data in a demo is clean, consistent, and complete. Columns align. Date formats match. Numbers reconcile. The AI reads it perfectly and the system performs exactly as promised.

But almost no real portfolio has data that looks like a demo. The spreadsheets that actually exist in a firm have been inherited, modified, cobbled together from systems that no longer exist. They have inconsistencies. They have gaps. They have notes in cells where a number ought to be.

“The quote changes when you see the real data. The question is whether you quote before you look.”

Teddy James, Tercero Analytics

This is not a failing of the demos or the AI systems. It is a fact of how real data lives in real organisations. And it is the reason why we do not scope engagements on the basis of demonstrations.

The data review is where you find the actual scope

We ask every potential client to share their data before we quote. Not polished data. Not a sample. Whatever they actually have. We do a short review, usually a few hours of work. We look at formats, we look at completeness, we look at where the meaningful inconsistencies are.

This review tells us two things. First, it tells us what the system actually needs to handle. Second, and more importantly, it tells us what the real scope of work is. The demo said we could extract an occupancy rate from column D. The data review says column D has six different formats and uses three different calculation methods depending on which asset and which era of the portfolio we are looking at.

Now we know what engineering is actually needed. Not what the demo suggested was needed, but what the data itself requires.

Why this matters for both sides

For us, the data review is insurance. It prevents us from signing a contract that sounds reasonable but becomes impossible once we see what we are actually working with. We cannot promise to extract clean numbers from a system that has none. We can promise to extract numbers and show you exactly what inconsistencies we had to resolve to do it. But that second promise is a different engagement than the first one.

For you, the data review is a gift. You get to see what is actually possible with the data you have, not the data you wish you had. You get a clear picture of where the AI system will be confident and where it will need human review. You get a quote that is built on reality instead of optimism.

We have never had a client who regretted asking for the data review before the quote. We have had plenty who regret not asking earlier, after they have already committed to a price that was built on a demo.

The honest conversation starts with your data

This is not how a lot of technology is sold. Normally you get the pitch, then you buy the box, then you find out whether it actually works with your stuff. The framing is that the box is the problem if it does not fit. Your data is messy so the vendor promises to make it less messy, or to build you a cleaner layer on top, or to handle the messiness with enough training and tuning that you never notice.

We think that is backwards. Your data is not the problem. Your data is the starting point. The system that is going to work for you is the one that is built knowing exactly what your data looks like and what it needs.

That is why we review the data first. It is why we quote second. And it is why the quote you get is honest about what you are actually buying instead of being a promise that depends on your data being different from what it is.