TL;DR
- ✓Public case studies are useful evidence, but they are also a self-selected marketing record. In the current public directory cohort, we found 1,616 distinct case-study rows across 314 partner profiles. Those profiles represent 314 of 1,507 eligible partners (20.8%).
- ✓The strongest pattern is representation, not a verdict on who gets the “best” result. Manufacturing appears in 365 distinct case studies (22.6% of the full row denominator). That means buyers have more public stories to inspect in that category; it does not establish a higher implementation-success rate.
- ✓The cohort denominator is every public, active, non-hidden Dynamics VAR or mixed-model partner profile at generation time. The source table is case_study_analyses. Every total below counts distinct case-study id values; repeated array elements and several aliases from the same row cannot increase the case-study denominator.
Public case studies are useful evidence, but they are also a self-selected marketing record. In the current public directory cohort, we found 1,616 distinct case-study rows across 314 partner profiles. Those profiles represent 314 of 1,507 eligible partners (20.8%).
The strongest pattern is representation, not a verdict on who gets the “best” result. Manufacturing appears in 365 distinct case studies (22.6% of the full row denominator). That means buyers have more public stories to inspect in that category; it does not establish a higher implementation-success rate.
The analyzed cohort and field coverage
The cohort denominator is every public, active, non-hidden Dynamics VAR or mixed-model partner profile at generation time. The source table is case_study_analyses. Every total below counts distinct case-study id values; repeated array elements and several aliases from the same row cannot increase the case-study denominator.
| Field | Rows with a value | Coverage of 1,616 case studies |
|---|---|---|
| Customer industry | 1,615 of 1,616 | 99.9% |
| Project type | 1,616 of 1,616 | 100.0% |
| Mapped product family | 1,161 of 1,616 | 71.8% |
| Challenges solved | 1,563 of 1,616 | 96.7% |
| Customer size | 1,616 of 1,616 | 100.0% |
| Reported outcomes | 1,565 of 1,616 | 96.8% |
| Key metrics | 1,260 of 1,616 | 78.0% |
Industry representation: the public-story leaderboard
| Industry | Distinct case studies | Share of all 1,616 rows |
|---|---|---|
| Manufacturing | 365 | 22.6% |
| Retail | 141 | 8.7% |
| Healthcare | 105 | 6.5% |
| Financial Services | 76 | 4.7% |
| Non-profit | 67 | 4.1% |
| Professional Services | 62 | 3.8% |
| Government | 45 | 2.8% |
| Construction | 33 | 2.0% |
| Education | 32 | 2.0% |
| Food & Beverage | 32 | 2.0% |
| Agriculture | 25 | 1.5% |
| Automotive | 24 | 1.5% |
| Hospitality | 23 | 1.4% |
| Logistics | 20 | 1.2% |
| Transportation | 18 | 1.1% |
A thicker public record can make due diligence easier, while a thin record should prompt better questions rather than a negative conclusion. Use the partner directory to identify relevant profiles, then ask for references matching your own operating context.
Primary project types in the published record
| Project type | Distinct case studies | Share of 1,616 classified rows |
|---|---|---|
| Implementation | 1,270 | 78.6% |
| Migration | 169 | 10.5% |
| Optimization | 69 | 4.3% |
| Integration | 61 | 3.8% |
| Support | 41 | 2.5% |
| Other | 6 | 0.4% |
One row receives one primary project-type label even where the underlying work was broader. These shares describe how case studies are framed, not the mix of Dynamics projects delivered across the market.
Products featured versus products listed
Case-study product names and canonical profile products are mapped through the versioned product-family alias map (tdp-product-family-aliases schema version 1). A family is counted at most once per case-study row and once per eligible partner profile. This is especially important for Azure: a row containing several Azure aliases still contributes exactly one Azure case study.
| Product family | Distinct case studies | Share of 1,616 case studies | Eligible profiles listing family | Share of 1,507 profiles | Representation gap |
|---|---|---|---|---|---|
| Business Central | 675 | 41.8% | 918 | 60.9% | -19.1 percentage points |
| Power BI | 253 | 15.7% | 884 | 58.7% | -43.0 percentage points |
| Azure | 189 | 11.7% | 1,061 | 70.4% | -58.7 percentage points |
| Finance & Operations | 156 | 9.7% | 331 | 22.0% | -12.3 percentage points |
| Microsoft 365 | 153 | 9.5% | 1,035 | 68.7% | -59.2 percentage points |
| Sales | 77 | 4.8% | 523 | 34.7% | -29.9 percentage points |
| Supply Chain Management | 72 | 4.5% | 289 | 19.2% | -14.7 percentage points |
| Power Platform | 59 | 3.7% | 609 | 40.4% | -36.8 percentage points |
| Power Apps | 49 | 3.0% | 268 | 17.8% | -14.8 percentage points |
| Power Automate | 38 | 2.4% | 246 | 16.3% | -14.0 percentage points |
A gap is not proof of weak delivery or unmet demand. It can reflect confidential work, editorial choices, alias limitations, or incomplete profiles. It is most useful as a prompt: “Which recent projects in this product family can you describe, and what was your role?”
Challenges described in case studies
Challenge values are free-text extraction labels and can overlap within one case study. The table reports the most repeated normalized labels; each label counts a given case-study ID no more than once.
| Challenge | Distinct case studies | Share of all rows |
|---|---|---|
| Inefficient inventory management | 10 | 0.6% |
| High operational costs | 9 | 0.6% |
| Inefficient manual processes | 8 | 0.5% |
| Operational inefficiencies | 7 | 0.4% |
| Inefficient production processes | 5 | 0.3% |
| Manual data entry errors | 5 | 0.3% |
| Time-consuming manual processes | 5 | 0.3% |
| Inefficient operational processes | 4 | 0.2% |
| Inefficient supply chain processes | 4 | 0.2% |
| Lack of process visibility | 4 | 0.2% |
| Outdated accounting system | 4 | 0.2% |
| Outdated legacy systems | 4 | 0.2% |
| Disconnected systems | 3 | 0.2% |
| High fulfillment costs | 3 | 0.2% |
| Inefficient manual reporting processes | 3 | 0.2% |
Estimated customer size
Customer size is an extracted estimate, not a verified employee or revenue band. Missing size values remain outside the classified denominator and are disclosed in the coverage table.
| Customer size | Distinct case studies | Share of all rows |
|---|---|---|
| Mid-Market | 962 | 59.5% |
| Enterprise | 462 | 28.6% |
| SMB | 107 | 6.6% |
| Unknown | 85 | 5.3% |
Outcomes partners choose to report
Outcome values preserve normalized free-text labels from public stories. They are partner-published claims, not independently audited measures, and several outcome labels can occur in one row.
| Reported outcome | Distinct case studies | Share of all rows |
|---|---|---|
| Improved operational efficiency | 46 | 2.8% |
| Increased operational efficiency | 36 | 2.2% |
| Enhanced operational efficiency | 33 | 2.0% |
| Streamlined operations | 18 | 1.1% |
| Improved customer experience | 10 | 0.6% |
| Enhanced customer satisfaction | 9 | 0.6% |
| Improved customer service | 9 | 0.6% |
| Streamlined financial processes | 9 | 0.6% |
| Enhanced decision-making capabilities | 8 | 0.5% |
| Enhanced data-driven decision making | 7 | 0.4% |
| Improved inventory management | 7 | 0.4% |
| Streamlined business processes | 7 | 0.4% |
| Streamlined processes | 7 | 0.4% |
| Enhanced inventory management | 6 | 0.4% |
| Increased customer satisfaction | 6 | 0.4% |
How to use this evidence in a shortlist
- Start with relevant industry and product evidence, but do not stop at the logo wall.
- Ask what was measured, over what period, and whether the partner can provide a customer reference.
- Compare implementation scope, integrations, data migration, governance, and change-management work with your own requirements.
- Use our partner-selection guides to structure a comparable reference-check process.
Methodology and limitations
This snapshot’s latest source extraction timestamp is April 2, 2026 at 10:15 AM UTC (2026-04-02T10:15:12.199Z). The generator selects the current public cohort where status = 'active', website_hidden_at IS NULL, and dynamics_partner_type IN ('var', 'mixed'), then retains case_study_analyses rows belonging to those profiles. Its stable pagination orders rows by primary key, de-duplicates by case-study ID, and uses deterministic count-descending/label-ascending ranking.
Coverage is 314 of 1,507 eligible profiles (20.8%). The analysis uses AI-assisted structured extraction from public pages and can inherit source-page omissions, label ambiguity, duplication before URL-level upsert, and extraction error. Product mapping is limited to aliases in schema version 1; unmapped labels are excluded from the product-family breakdown and disclosed through its field coverage.
Most importantly, this is a self-selected success-story sample. Partners choose which customers to feature, customers may decline publication, and unsuccessful or routine projects are unlikely to appear. Reported outcomes and metrics are not independently verified. See our methodology and how we collect data for more context.
Share this analysis
LinkedIn: We analyzed 1,616 distinct public Dynamics case studies across 314 partner profiles. The useful signal is representation—what partners choose to publish—not a shortcut to judging implementation success. →
X: 1,616 public Dynamics case studies, counted by distinct row and mapped through a versioned product-family taxonomy. Here is what the published record covers—and what it cannot prove. →
