The opportunity,
and the trap
More companies want to do the same thing: put the data they already hold to work. Not for its own sake, but against a clear commercial objective. Done well, a data product does one or more of three things. It generates revenue, it improves efficiency, or it reduces exposure to a level the business can accept. An offering they can sell, a feature that lifts retention, an automation that strips out cost, a model that keeps them the right side of a regulator. The prize is real, and the businesses that reach it treat data the way they treat any other product. It has customers, economics, a lifecycle, and a standard it has to meet.
The trap is treating the work as an engineering problem. Most leaders picture data strategy as pipelines and schemas: move the data, model it, store it, and the value follows. It does not. The data initiatives that stall rarely stall on the plumbing. They stall because nobody proved there was a paying customer, because the business did not have the rights to use the data that way, because no one trusted the numbers coming out, because the model that looked good in a demo fell over in production, or because a security review sank the deal at the last minute.
Building a data product takes different disciplines working together. Do it with just engineers, or work it out as you go, and the cost does not show up in the build. It shows up later as rework, delays, and outcomes that get missed.
This guide sets out the different lenses that make up a good data strategy, the roadmap that takes an idea to a product, and the places where bringing in the right expertise pays for itself several times over.
Friday's Position
Protect your data and turn it into commercial value at the same time. Most providers make you choose between the two. Treating protection and value as separate decisions is exactly what makes data products expensive, slow, and risky.
Every data product serves a commercial objective
A good data strategy is not a plan to “do more with data”. It is a plan to move a specific commercial number. There are only three commercial outcomes that ultimately matter, and every data product worth building maps to at least one of them.
Revenue
What it means
Data generates revenue, directly or indirectly.
What it looks like
Selling data products, insights, or data-as-a-service; personalisation, better customer service, and new features that lift conversion, retention, and lifetime value.
Efficiency
What it means
Data cuts the cost of doing business.
What it looks like
Automating manual work, allocating people and resources better, reducing waste, forecasting demand, and streamlining operations.
Exposure
What it means
Data reduces exposure, and its financial cost, to within tolerance.
What it looks like
Cutting regulatory, reputational, and competitive exposure: meeting obligations, preventing fraud and loss, protecting reputation, keeping pace with competitors, and avoiding fines, breaches, and failed audits. Every form of exposure carries a financial cost.
The first test
of a strategy
If you can’t say which of these three objectives a data product serves, and roughly how much it will move, it is not a strategy, it is an activity. Clarity of objective is what separates a data product from an expensive experiment.
This objective is the thread that runs through everything that follows. It is set in the Commercial & Value lens, but every other lens exists to deliver it safely, reliably, and at a cost that keeps it worthwhile. A product can serve more than one objective, but it must serve at least one, deliberately and measurably.
What data strategy actually is
A data strategy exists to answer the questions that decide whether a data product is worth building and whether it survives contact with real customers, real regulators, and real scale:
-
Which commercial objective does it serve (revenue, efficiency, or exposure), and is it worth building?
-
Are we allowed to use this data in this way?
-
Can we trust what comes out of it, and can we prove it?
-
Does the analysis or model actually work, and hold up over time?
-
Can we build, run, and scale it reliably and affordably?
-
Is it secure, and will customers and regulators trust it?
Each question belongs to a discipline. Together they make up a data strategy, and a data product needs to answer all six.
Commercial & Value
The question it answers
Is there a market, a margin, and a buyer?
What breaks without it
You build something technically impressive that nobody pays for.
Rights & Requirements
The question it answers
Are we legally and contractually allowed to use this data this way?
What breaks without it
A launch blocked or unwound late, fines, or a product you cannot monetise.
Governance & Management
The question it answers
Who owns the data, what does it mean, and is it fit for purpose?
What breaks without it
Nobody trusts the output; you fail a customer’s due diligence or an audit.
Analytics & AI
The question it answers
Does the analysis or model actually work, and can we defend it?
What breaks without it
A model that demos well, fails in production, and produces answers you cannot stand behind.
Technology & Infrastructure
The question it answers
Can we build, run, and scale it reliably and affordably?
What breaks without it
It works for ten customers and collapses at a thousand, or the cloud bill eats the margin.
Security & Trust
The question it answers
Is it safe, and can customers and regulators trust it?
What breaks without it
A breach, or a failed security review that kills an enterprise deal.
Notice how rights, governance, and security appear at every stage, not only at the end. That is deliberate. The cost of fixing a rights, governance, or security problem rises sharply the later it is found. A constraint spotted at Strategy is a design decision. The same constraint found at Deploy is rework, delay, legal cost, and sometimes a product that cannot launch. Budgets rarely balloon on the build. They balloon on the rework the build made necessary, and on the outcomes that slip because no one owned the objective in the first place.
The two most expensive words in a data product are “later.” As in, “we’ll sort the rights later,” or “security can review it later.”
Later is always the most expensive time to find out.
Getting it right without hiring six functions
The honest problem for most CEOs is that you can’t justify hiring a commercial data lead, a data lawyer, a governance manager, a data scientist, a platform team, and a security specialist to build one product. The people who can cover all six lenses well are rare and expensive, and stitching the coverage together from separate contractors and agencies is where the coordination cost and the overruns actually live. Everyone owns their piece; nobody owns the join.
This is the gap Friday is built to close. We embed the cross-workstream expertise across the whole roadmap, so every lens has an owner and the joins between them are held by one team. Strategic data leadership that spans commercial, rights, governance, analytics, technology, and security, backed by our platform, Antonio, which makes governance live in your systems rather than in a document. One retainer, embedded, covering the disciplines that would otherwise be six hires or six suppliers.
On AI
AI is the most urgent reason to get your data right. Because good data means good AI outputs. Skip this, and AI simply makes bad data fail faster and more publicly.
Where to start
The cheapest stage to get right is the first one. A short Strategy & Opportunity assessment tests whether the value hypothesis holds, whether you have the rights, and where the real risk sits, before the build budget is committed. It is the single highest-return conversation you can have about a data product, and it is where we would suggest starting.
That Friday feeling
Strategy and governance so good that data becomes easy. Build the product once, on foundations that hold, and turn the data you already own into something you can sell, safely.
Want hot takes
on the latest data news, and a snapshot of what we’ve been up to at Friday HQ?
Please only provide your work email address. You can unsubscribe from our marketing communications at any time. For more information, please review our Privacy Policy.
+44 07384417395 | +61 0452 679 796
Studio 3, Rose Lipman Building
43 De Beauvoir Road
London, N1 5SF
18/49 Head Street
Brighton, Victoria
3186
© 2026 by Alex Gomez for Friday Initiatives.
Built on Wix Studio™