Business Intelligence Data Strategy Power BI

How to Choose a BI Migration Partner | Numlytics

How to Choose a BI Migration Partner | Numlytics
Business Intelligence

Choosing a BI Migration Partner: Questions Every Data Leader Should Ask

⏱️ 10 min read
Business Intelligence · Data Strategy
Choosing a BI migration consulting partner - an evaluation scorecard covering technical proof, Microsoft Solutions Partner designations, pricing models, red flags, and reference checks for data leaders

The partner you hire determines your migration's outcome more than the platform you migrate to - this scorecard covers the questions that separate a credible BI migration partner from a confident pitch deck.

Finding the right BI migration consulting partner matters more to a project's outcome than almost any technology decision that follows it. Two organisations can migrate to the identical Power BI or Microsoft Fabric architecture with the identical licence spend, and one will finish on time with adoption climbing while the other stalls at 60% complete with a legacy platform still limping along in parallel - and the difference is very rarely the technology. It is who was in the room designing and delivering the migration.

This guide is written for CDOs, IT Directors, and Data and Analytics Managers building an RFP or shortlist for a BI migration partner. It covers the technical capability questions that separate proof from a pitch deck, what Microsoft's partner designation framework actually signals in 2026, the answers that should end a vendor conversation, how migration pricing models differ, and the reference questions that reveal whether a past project actually achieved adoption - not just a technical go-live.

Why the Partner Matters More Than the Platform

Most partner evaluation checklists ask a version of the same three questions: how many certifications does the team hold, is the firm a Microsoft Solutions Partner, and can they provide references. Those questions are necessary but nowhere near sufficient, particularly for a platform like Microsoft Fabric that is still maturing rapidly - a firm with fifty certified Power BI developers and five production Fabric deployments is a fundamentally different risk profile than one with a smaller team and twenty enterprise Fabric go-lives behind it. Capability on paper and capability under production conditions are not the same thing, and the gap between them is exactly where migrations go over budget and over time.

The evaluation framework below separates signal from credential count across six dimensions: technical proof, partner designations, red-flag answers, pricing structure, named staffing, and verifiable references. No single dimension should be decisive on its own - a partner who scores well on certifications but cannot name the specific engineer assigned to your project, or who quotes glowing case studies but cannot produce a client reference who will speak candidly, is a partner whose strongest evidence is the parts of the story they chose to tell you.

Six Questions About Technical Capability

How do you handle our existing BI estate?
A partner who proposes to evaluate and integrate your existing semantic models, workspace governance, and report library - rather than defaulting to a clean rebuild - understands enterprise reporting governance. A partner optimising for a greenfield build is optimising for their own convenience.
Reveals: inventory-first vs. rebuild-first instinct
What is your methodology, in named artefacts?
Ask for the specific deliverables at each phase - a usage audit report, a data model design document, a UAT sign-off template - not a slide describing "a comprehensive solution." Named artefacts mean the methodology has been run before and refined.
Reveals: repeatable process vs. improvised delivery
How do you handle row-level security and governance?
A migration partner should have a concrete answer for how existing authorisation logic - SAP row-level security, Active Directory groups, source-system permissions - carries through to the new platform, not a generic assurance that "governance is important to us."
Reveals: production experience vs. sales-stage confidence
What is your metric reconciliation process?
Verifying that the new platform produces the same numbers as the old one is the least glamorous and most failure-prone part of any migration. A partner with a defined reconciliation checklist and sign-off process has done this enough times to know where the discrepancies hide.
Reveals: attention to the detail that erodes trust if missed
What does your parallel-run and cutover plan look like?
A credible partner defines an explicit exit criterion for the dual-run period - named reports at parity, named business owners signing off - rather than an open-ended "until people are comfortable" timeline that quietly inflates cost.
Reveals: cost discipline vs. scope drift risk
How do you measure success after go-live?
Ask whether success is defined as a technical cutover date or as adoption metrics - active users, self-service report creation, reduction in ad-hoc requests to the data team - measured 60 and 90 days after go-live.
Reveals: deployment focus vs. adoption focus

What Microsoft Solutions Partner Designations Actually Signal

Microsoft retired the legacy Gold and Silver competency tiers in 2022 in favour of six Solutions Partner designations, and any firm still marketing "Gold Partner" status in 2026 has not renewed under the current framework - worth noting as a small but telling signal on its own. For a Power BI or Microsoft Fabric migration, the relevant designation is Solutions Partner for Data & AI (Azure), which covers Fabric, Synapse, Azure AI, and related data platform capability.

The designation is earned across three scored pillars - Performance, Skilling, and Customer Success - each requiring a minimum score, re-qualified annually against Microsoft's own billing telemetry and customer survey data rather than a one-time exam. That annual re-qualification is what makes the designation meaningful: it reflects a partner's current delivery track record, not a certification earned once and never revisited. Above the base designation sit Advanced Specializations - for example, "Analytics on Microsoft Azure" - which require passing customer case-study audits and technical assessments beyond the base designation, and are a stronger signal still for a data-platform-specific engagement.

What to Ask

Don't just ask whether a firm holds the Data & AI designation - ask how many individually certified professionals hold current Power BI Data Analyst Associate (PL-300), Fabric Analytics Engineer (DP-600), or Fabric Data Engineer (DP-700) credentials, and how many of those individuals will actually be staffed on your engagement, versus how many are counted toward the firm's designation but working on other accounts.

Six Red Flags That Should End the Conversation

Flag 01
A platform recommendation before any diagnosis
If a proposal recommends a specific platform migration in the first meeting, before an audit of your current reports, data sources, and usage patterns, the recommendation is coming from the vendor's product incentives, not your organisation's actual needs.
Flag 02
Deliverables described only as "a comprehensive solution"
Vague deliverable language is a substitute for a methodology the firm hasn't actually run before. Ask for named artefacts and a sample from a prior, anonymised engagement.
Flag 03
Marketing language with no third-party verification
"AI-powered," "best-in-class," and "next-generation" describe a vendor's marketing ambition, not a verified capability. Every credible claim should come with a named customer, a specific metric, or a published methodology behind it.
Flag 04
Case studies with no named client and no specific numbers
A curated success story without a company name, an industry, or a specific before-and-after metric is not evidence - it's copy. Ask to speak to the client directly, not to read a one-page summary the vendor wrote.
Flag 05
No answer for existing governance and security carry-over
A partner who cannot describe, specifically, how your current row-level security or source-system authorisation model carries through to the new platform has not run a migration with governance requirements as demanding as yours.
Flag 06
Reluctance to name the assigned team before signing
If a firm won't name who specifically will be staffed on the engagement until after the contract is signed, assume the strongest people in the sales pitch will not be the people doing the work.
"Every credible BI migration partner should answer all of the technical and staffing questions above without hedging. If a firm dodges even one, that tells you something worth investigating before you sign - not after."

Pricing Models: What Each One Is Really Buying You

BI consulting pricing varies enormously by engagement shape, and understanding what each shape typically costs helps calibrate whether a quote is reasonable before negotiation even starts.

Engagement typeTypical durationList price range (senior-led US firms)
Diagnostic / usage audit3–6 weeks$30,000–$80,000
Cleanup / rationalisation6–16 weeks$120,000–$400,000
New build / platform migration12–36 weeks$200,000–$1,200,000
Managed services retainerOngoing, monthly$8,000–$30,000/month

A quote well below these ranges for an equivalent scope is not automatically good value - it often signals a fixed-price engagement with tightly bounded scope that will generate change orders the moment your actual data model complexity becomes apparent. A quote well above these ranges should come with a correspondingly clear justification: a larger named team, deeper specialisation, or a materially more complex data estate than the standard range assumes. Most mid-market organisations end up spending $300,000–$500,000 in their first year across one audit, one build or cleanup, and the start of an ongoing retainer - treat any single-engagement quote against that fuller-year context, not in isolation.

Ask About the Named Team, Not the Credential Count

A firm's total certification count is a company-level statistic; it tells you almost nothing about who will actually be on your project. Before signing, ask for the names, certifications, and prior project history of the specific individuals who will be staffed on your engagement - not the firm's broader bench. Ask what happens if a named team member leaves mid-engagement, and what the firm's actual bench depth looks like beyond the two or three people in the sales pitch. A firm that hesitates to commit specific names to a statement of work is telling you, indirectly, that the people who sold you the engagement are not the people who will deliver it.

References That Prove Adoption, Not Just Deployment

The most common mistake in reference checks is asking only whether a project went live on time and on budget. A technically successful cutover with low user adoption is not a successful migration - it is a technically successful failure. When speaking with a reference, ask specifically for adoption metrics before and after: the number of active users, the number of business users creating their own reports rather than requesting them from the data team, and the change in ad-hoc report request volume in the 90 days following go-live. A partner whose references can speak fluently to these numbers has actually been held accountable for adoption in past engagements, not just for a deployment date.

Key Takeaways
  • The partner determines a BI migration's outcome more than the platform choice does - evaluate the delivery team and methodology as rigorously as the technology architecture.
  • A platform recommendation offered before any diagnosis, and deliverables described only as "a comprehensive solution," are the two most common red flags in vendor proposals.
  • Microsoft's Solutions Partner for Data & AI (Azure) designation is scored annually across Performance, Skilling, and Customer Success - a more meaningful signal than a one-time certification, but still only a starting filter, not a substitute for named-team questions.
  • Realistic pricing benchmarks: $30K–$80K for a diagnostic audit, $120K–$400K for a cleanup, $200K–$1.2M for a full platform migration, and $8K–$30K/month for an ongoing managed retainer.
  • Reference checks should ask about adoption metrics - active users, self-service report creation, reduced ad-hoc requests - not just on-time, on-budget delivery.

Building Your Evaluation Scorecard

The strongest partner evaluations score every shortlisted firm against the same written scorecard - technical proof, designation and named certifications, pricing structure and transparency, named staffing commitment, and verified reference outcomes - rather than relying on the impression left by the best presentation. Building that scorecard before the first vendor call, and scoring every firm against it consistently, is what prevents the strongest pitch deck from winning over the strongest track record.

Numlytics is a certified Power BI consulting practice built around named-team delivery, not a rotating bench - every engagement is scoped and staffed by consultants you can meet and reference-check before signing. Explore our Power BI consulting practice and our approach to data governance consulting, or speak with a certified consultant to see exactly who would be staffed on your migration before you sign anything.

BI Migration Consulting
Meet your named consulting team before you sign - not after
Numlytics scopes every engagement with the specific consultants who will do the work, references you can call directly, and pricing tied to your real BI estate - not a generic quote.
Speak With Your Named Consulting Team →