cimt

Service · Selection

Selecting your data platform

You choose a data management platform for ten years. The choice is a four to six month decision process, with requirements, a shortlist, an RFP, a proof of concept and a decision the board can sign. cimt guides that process from frame to contract.

A platform selection is the process in which you decide on which platform your organisation will integrate, store, monitor and analyse data for the coming years. It follows your data strategy and target architecture, and precedes the implementation. cimt orders the requirements along the knowledge areas of DAMA DMBoK, so integration, quality, metadata, security and analytics each have their place and no requirement is forgotten.

The work is in gathering and prioritising requirements with business, IT, security and privacy, in demos that prove something, in a proof of concept on your own data and in a TCO that looks beyond the licence. Comparing brochures is the smallest part.

The approach

Seven phases, one decision

From frame to contract in four to six months, depending on the size of the proof of concept. Each phase produces a document that carries the next.

  1. 01 2 to 3 weeks

    Frame and scope

    Why now, which pain, which parts of the platform. Business case, scope boundaries, steering group and the decision criteria fixed up front.

  2. 02 3 to 5 weeks

    Gather requirements

    Workshops with business, IT, security and privacy. The current landscape mapped. A requirements catalogue with priorities and concrete use cases.

  3. 03 2 weeks

    Longlist to shortlist

    Market scan, a test against the hard requirements, a weighting model approved by the steering group beforehand. Three to four vendors remain.

  4. 04 4 to 6 weeks

    RFP and scripted demos

    One request for all vendors. Demos on your own use cases instead of the vendor’s demo data. References called.

  5. 05 3 to 6 weeks

    Proof of concept

    One or two finalists build a bounded case on your own data. Acceptance criteria are fixed in advance, the evaluation report follows the weighting model.

  6. 06 2 to 4 weeks

    Decision and contract

    Five-year TCO, licence model, exit terms, security and privacy review. An advisory report the steering group can sign.

  7. 07 ongoing

    Handover to implementation

    Roadmap, operating model and migration plan. The requirements and the weighting model carry over as the acceptance framework for the implementation.

What you hold at the end

Six documents that carry the decision

Per phase a document the steering group uses to decide. Together they form the file with which you can still justify the decision years later.

Business case and scope document

Why, what is in and out, and who decides.

Requirements catalogue

Functional, technical and non-functional, ordered along the DAMA knowledge areas, with must/should/could.

Weighting model and shortlist

Criteria, weights and scores, approved by the steering group beforehand.

RFP and demo script

One request and one script for all vendors, on your own use cases.

PoC plan and evaluation report

Acceptance criteria before, findings after, per finalist.

Advisory report with TCO

Licences, implementation, operations and exit over five years, with the recommendation.

Decide first

Four choices you make before you talk to vendors

Leave these open and the vendor makes them for you. We put them on the table in the first phase.

Build or buy

Build on open components or buy a platform. The deciding factor is the team that runs it in three years.

One suite or best-of-breed

Integration, quality, catalogue and analytics from one vendor, or the best tool per component and managing the connections yourself. Some components, such as API management, are by nature a separate component in the architecture.

Cloud, hybrid or on-premises

Data residency in the EU, latency to your source systems and what your security policy allows.

Who decides

A steering group with a mandate, a weighting model fixed in advance and a clear role for procurement.

Where it goes wrong

  • · Selecting on a feature list instead of on your own use cases.
  • · Demos on the vendor’s demo data: everything always works.
  • · TCO on licence cost only, without implementation, operations and exit.
  • · Forgetting the non-functional requirements: lineage, access control, data residency, data portability.
  • · Fixing the weighting model only once the favourite is already known.

Our partners, on the table

We work with Qlik, Snowflake and erwin. We say so up front.

cimt is a Qlik Elite Partner and implements Snowflake and erwin by Quest. That is knowledge you use in the selection: we know what these platforms can and cannot do and where the licence costs land in practice. If they fit your requirements, we say so. If they do not, we say that too. The weighting model and the acceptance criteria are fixed before any vendor is at the table, and the steering group scores alongside us. The decision stays yours.

More on our technology: Qlik Talend Cloud, Snowflake and erwin by Quest.

Free selection guide

The data platform selection guide (PDF)

Seven phases from frame to contract, with per phase the duration, the documents and the decision. The four choices to make first, the pitfalls and a timeline. As appendices a selection checklist and a requirements template along DAMA DMBoK, completed for a technical wholesaler and blank.

  • · Management summary: why, how and what
  • · Seven phases with duration, documents and decision
  • · The four choices to make first and the five pitfalls
  • · Appendix: selection checklist and requirements template, completed and blank

15 pages, PDF

No spam. The download link arrives by email and is valid for seven days.

Your point of contact

Talk to the consultants behind this service

A substantive half-hour conversation about your situation, without a sales pitch. Afterwards you know whether and how we can help.

Book an introduction
  • Taco van het Reve

    Taco van het Reve

    Managing Director

  • Mirco Kriesten

    Mirco Kriesten

    Managing Consultant

Start with the frame

Book a selection call

In a short session we map your trigger, your current landscape and the four choices to make first. Afterwards you know how large the process is for you and where to start.

Frequently asked questions

Platform selection: what data leaders ask us

How long does a data platform selection take?

Count on four to six months from frame to contract. The largest variable is the proof of concept: one finalist on one use case takes three weeks, two finalists on your own data six. A selection without a proof of concept is faster, but then you sign on the basis of demos.

Can cimt do only the requirements phase?

Yes. Many organisations run the market scan and the negotiation themselves and mainly want the requirements to be complete, prioritised and testable. The requirements catalogue and the weighting model are available as a separate engagement.

How independent is cimt as a Qlik, Snowflake and erwin partner?

Not neutral, but transparent. We name our partnerships up front and fix the weighting model and the acceptance criteria before any vendor is at the table. The steering group scores alongside us. You use our product knowledge to test more sharply, including with vendors we have no relationship with.

What does a proof of concept cost and who pays?

The vendor usually provides the environment and part of the hours; your own time goes into supplying data, use cases and evaluators. We guide the setup, the acceptance criteria and the evaluation. Agree in advance what remains of the PoC environment if you do not continue with that vendor.

We already hold licences. Is a selection still worthwhile?

Often it is. Existing licences are a criterion in the weighting model, not a conclusion in advance. A short test of the current platform against the requirements shows whether extending, complementing or replacing is the best route.