Your Project Is Only as Good as the Data Behind It

Your Project Is Only as Good as the Data Behind It

Why delivery leads need to question the data behind their decisions, not just track milestones
Why delivery leads need to question the data behind their decisions, not just track milestones

Ai

Violetta Luri

,

Portfolio Manager

A portrait


A tech project can be delivered on time, within budget and still fail to achieve what the customer actually needs. Sometimes the warning signs were in the data long before launch.

We spend a lot of time tracking milestones, managing scope and reporting progress. But how often do we question the data behind the decisions we're making?

I started in UX/UI design, moved into Product Manager and Project Manager roles, and now work as a Tech Portfolio Manager. Through my experience working with multiple websites, e-commerce applications, various CMS platforms and martech solutions, I've learned that understanding and using data effectively is just as important as managing budgets, timelines and delivery teams.

Technology keeps changing, especially with AI becoming part of both the products we build and the tools we use every day.

But one thing remains consistent. Good delivery depends on how well we understand and use the information behind our decisions.

Data needs attention throughout delivery

When we talk about data, the conversation often turns to analytics, dashboards and reporting. But its role goes much further.

In UX, data helps us understand customer behaviour and where users struggle. In product management, it helps us validate assumptions and decide what is worth building. In technology delivery, it helps identify dependencies, assess risks and make informed decisions about scope and investment.

The challenge is that data problems often become visible when delivery is already underway.

Take a CMS migration. It's common to uncover outdated pages, duplicated information, missing metadata or content with no clear owner. Add customer profiles and consent records scattered across martech platforms, and a seemingly straightforward project becomes much more complicated.

Now introduce AI-powered search, personalisation or a customer assistant. If the underlying content is outdated or inconsistent, even a well-built solution can produce unreliable results.

This is something I believe is becoming increasingly important. Content is data too. Product descriptions, FAQs, policies and website content are no longer just information for customers to read. They can also become source material for AI-powered experiences.

If nobody owns that information or checks its quality, the problem won't disappear when we introduce better technology.

Data should influence the plan, not just report progress

For me, using data effectively means considering it throughout the delivery lifecycle.

During discovery, we need to understand what data exists, whether it's reliable and whether we can use it. Finding gaps early helps avoid building estimates and delivery plans around assumptions.

During development, real data should help us validate functionality, challenge requirements and identify problems before they become expensive changes.

After launch, it should tell us whether the product is delivering the expected value.

A website redesign might improve traffic while conversion rates decline. A new integration might work technically but still require significant manual intervention. A project can meet its delivery milestones without meeting its business objectives.

These are the situations where data should influence our next decision, rather than simply appear in another report.

I also found the approach shared by Netflix in its article on data as a product particularly relevant.

The principle is simple. Data needs clear ownership, an understanding of who uses it, and ongoing attention to quality. Much like any product we deliver.

For PMs and delivery leads, this means understanding which data issues could affect the product and making sure they are reflected in priorities, risks and delivery decisions.

Why this matters even more when using AI

I'm not saying every PM needs to become a data analyst, engineer or AI specialist.

But I do believe we need to understand data well enough to ask better questions, challenge assumptions and make informed decisions.

This becomes especially relevant when AI is part of our everyday work.

Many of us already use AI to analyse requirements, review project risks, prepare reports or explore delivery scenarios.

But the value we get depends on more than writing a good prompt.

Asking AI to identify project risks is one thing. Providing an up-to-date delivery plan, outstanding dependencies and a risk register, then asking it to assess potential impacts, is something quite different.

Of course, that information must be shared through approved tools, and the results still need to be checked.

Knowing which data is relevant, whether it's trustworthy and how to apply it helps us write better prompts and recognise when AI-generated recommendations don't reflect the reality of the project.

AI can help us process information faster, but we still need the knowledge and judgment to use it correctly.

This is also why I've been spending more time developing my understanding of data, AI governance and security. Not because I want to move away from product and technology delivery, but because these areas are becoming increasingly connected.

The bigger opportunity

The value comes from knowing what questions to ask, bringing the right specialists into the conversation and understanding how the answers affect the product, the budget and the delivery plan.

That means recognising risks earlier, challenging assumptions, prioritising the work that creates value and being prepared to change direction when the evidence tells us something different.

We often talk about AI helping teams deliver faster.

I think the bigger opportunity is using data to make better delivery decisions in the first place.

Because successful technology delivery isn't simply about getting something live. It's about making sure we've built the right thing, that it works with real information and that it delivers the value we promised.

GENERATIVE SEO

Want to ensure your website doesn't get left behind in the future of SEO?

GENERATIVE SEO

Want to ensure your website doesn't get left behind in the future of SEO?

GENERATIVE SEO

Want to ensure your website doesn't get left behind in the future of SEO?

Talk to us about your challenges, dreams, and ambitions

X social media icon

Codehouse acknowledges the Traditional Owners of Country throughout Australia. We pay our respects to Elders past and present.

©

2026

All rights reserved, Codehouse

Talk to us about your challenges, dreams, and ambitions

X social media icon

Codehouse acknowledges the Traditional Owners of Country throughout Australia. We pay our respects to Elders past and present.

©

2026

All rights reserved, Codehouse

Talk to us about your challenges, dreams, and ambitions

X social media icon

Codehouse acknowledges the Traditional Owners of Country throughout Australia. We pay our respects to Elders past and present.

©

2026

All rights reserved, Codehouse