Practitioner-Led Data Consulting — Fractional CDO Model

Most large system integrators hand the work to juniors after the SOW signs. The partner who presented in the discovery meeting moves to the next pursuit. The graduate who just completed their Microsoft certification is now building your data architecture.

This is not a secret. It is a structural feature of the SI model — and for mid-market industrials, it is expensive.

Large SI engagements are priced on blended rates. A senior architect at the front of the engagement validates the approach, the commercial team closes the deal, and the delivery team — assembled from whoever is available at the right billing tier — executes. The economics require this. There is no other way to make the margins work at scale.

For an enterprise with a dedicated IT governance function, a programme management office, and internal architects who can hold the delivery team accountable, this model functions adequately. The client has the internal capacity to manage the vendor.

Mid-market industrials — manufacturers running SAP ByDesign or SAP S/4HANA, FMCG operations on Microsoft Dynamics 365, 3PL companies holding their data together with a combination of ERP and spreadsheets — typically do not have that internal capacity. The IT Head is also the infrastructure manager, the ERP administrator, and the person who fields the Power BI support calls from finance. When a junior consultant from the SI makes an architectural decision that creates a two-year problem, there is no one internally positioned to catch it in time.

The result: a data project that was supposed to deliver a Microsoft Fabric lakehouse and a Power BI reporting layer in six months is still running at month ten. The scope has drifted. The original business case is no longer connected to what is being built. The CFO is asking whether the investment will ever return.

The Fractional CDO model — which is how MyData Insights structures engagements — is a different proposition. Not a different pitch. A structurally different delivery model.

Amit is the person in the discovery conversation. Amit is the person who designs the architecture. Amit is the person who builds the Microsoft Fabric environment, configures the Azure Data Factory pipelines, and validates that the Power BI semantic model is answering the questions the Operations Director actually needs answered — not the questions that were easiest to model.

There is no junior behind the scenes. There is no handoff between pre-sales and delivery. The person who scoped the project is the person executing it.

For a mid-market industrial, this matters for reasons that go beyond personnel preference. It means the commercial terms can be fixed-scope. The risk of scope drift is carried by the practitioner, not the client. It means the six-week prototype — the working proof that the approach is sound — is built by the person who designed it, not delegated to a resource who read the design document. It means honest conversations happen earlier, because there is no commercial incentive to defer bad news to protect a renewal.

MyData Insights runs every engagement through four phases, and the sequencing is non-negotiable.

Discover first. Before any code is written, before any Fabric workspace is provisioned, before any data pipeline is configured — a structured assessment of the data landscape, the operational problems that data is supposed to solve, and the gap between the two. This is where most engagements reveal their real constraint: not the technology, but the quality and accessibility of the source data from SAP ByDesign, Microsoft Dynamics 365, or whatever ERP the client is running.

Prototype second. A working proof of the core use case — not a slide, not a demo on cleaned sample data, but a functional prototype connected to actual client data — delivered within six weeks. If the prototype does not produce value, the engagement stops. The client has spent six weeks and a defined fixed fee, not twelve months and a six-figure commitment.

Deploy third. The prototype becomes production. Governance is established. Users are trained. The Power BI reports are in front of the people who need to make decisions, not sitting in a workspace that only the IT Head can access.

Expand last. Once the first use case is working and trusted, the next one is scoped. Not before. The single most common failure mode in data projects is expanding scope before the foundation is stable. Unify the data. Predict with AI. Act with automation. In that order — always.

The decision-makers at mid-market industrial businesses — CEOs, COOs, GMs — are not buying technology. They are buying a working outcome within a defined time and cost envelope. They have been burned by projects that promised transformation and delivered a dashboard that nobody looks at.

The Fractional CDO model aligns the practitioner's incentives with the client's outcomes in a way the SI pyramid model structurally cannot. A fixed-scope, six-week prototype means the practitioner absorbs the risk of overrun. An honest conversation about what the data environment can and cannot support in the first Discover session means the client is not twelve months into an engagement before discovering that their SAP ByD master data is in a state that makes reliable reporting impossible.

It also means the engagement can be right-sized. A 3PL operation in Dubai needing a Microsoft Fabric lakehouse connected to their WMS and a Power BI OTIF dashboard does not need the same overhead as a global manufacturing group rolling out Fabric across eight plants. The former needs a practitioner who can start building in week one. The latter needs a programme team. These are different things, and conflating them serves neither client.

If your organisation needs a multi-year, multi-system transformation programme with a team of twenty across three countries, the Fractional CDO model is not the right vehicle. That kind of programme requires a delivery organisation with depth across multiple specialisms, dedicated programme management, and the capacity to sustain a long-running engagement.

If your legal entity requires a vendor of a certain size, or your procurement policy requires three competing bids from organisations above a revenue threshold, the Fractional CDO model may not meet your compliance requirements.

And if what you need is a strategy document — a roadmap for a board presentation, with no expectation of hands-on execution — this is also not the right fit. The value here is in building, not advising.

The Discover phase — a structured assessment of your data environment, your operational reporting gaps, and the highest-value use case to prototype — takes two to three weeks. The prototype phase produces a working Microsoft Fabric dataset or Power BI report connected to your actual source data within six weeks of the Discover engagement completing. Not a 50-slide roadmap. A working output.