Data Consulting vs Software Vendors: Why You Usually Need Both

Understanding the False Choice

Most data and analytics leaders face a persistent question: should we build internal capabilities, hire consultants, or invest in software? The framing of this question itself is the problem. It assumes these are competing options when, in reality, they’re complementary parts of a cohesive strategy.

The tension exists because each approach has genuine strengths. Software vendors like Looker, Tableau, and Power BI offer scalable infrastructure, pre-built connectors, and governance frameworks that would take years to build internally. Consultants bring domain expertise, change management acumen, and the ability to ask hard questions about what you actually need versus what you think you need. Neither alone is sufficient for most organizations at scale.

This article explores when to lean on each, how to avoid the common pitfall of choosing one at the expense of the other, and how to structure an engagement that maximizes ROI. We’ll ground this in real scenarios—from startups embedding analytics into their product to enterprise portfolio companies standardizing KPI reporting.

The Core Problem: Vendors Solve Infrastructure, Not Strategy

Let’s start with what software vendors actually do. A managed Apache Superset platform like D23, or enterprise tools like Looker and Tableau, provides:

• Infrastructure and hosting: You don’t manage servers, backups, or scaling.
• Connectors and data integration: Pre-built adapters to your data warehouse, APIs, and databases.
• UI and visualization layers: Dashboards, charts, and drill-down capabilities.
• Governance and security: Role-based access control, audit logs, and compliance frameworks.
• API-first architecture: The ability to embed analytics into applications or trigger actions based on data.

What they do not do is tell you:

• Which metrics actually matter to your business.
• How to structure your data model so dashboards remain maintainable at scale.
• Whether your current KPI definitions align with how you’re actually running the business.
• How to build a self-serve analytics culture without creating chaos.
• What technical debt you’re carrying in your data pipeline.

A vendor will give you a powerful shovel. They won’t tell you where to dig.

This is where the gap emerges. Many organizations buy a BI platform, invest in implementation, and then watch adoption plateau because the dashboards don’t answer the questions people actually care about. The platform isn’t the problem—the strategy is.

What Consultants Actually Bring to the Table

Data consultants operate in a different layer. Their role is to bridge the gap between business problems and technical solutions. A competent data consultant will:

• Conduct discovery: Spend time understanding your business model, competitive dynamics, and the decisions that matter most.
• Audit your data: Assess data quality, lineage, and whether your current infrastructure can support your analytics ambitions.
• Design the analytics strategy: Define the right metrics, ownership models, and governance frameworks before you buy or build anything.
• Guide tool selection: Help you evaluate whether a managed platform like D23’s Apache Superset hosting is right, or whether you need something heavier like Looker.
• Implement with context: Build dashboards and data models that reflect your actual business logic, not generic templates.
• Build internal capability: Train your team, document processes, and hand off ownership so you’re not dependent on external help forever.

Consultants are expensive precisely because they carry institutional knowledge about what works and what doesn’t. They’ve seen dozens of data platform implementations and know which failure modes to avoid.

However, consultants also have limitations. They’re not scalable in the way software is. Once they finish an engagement, you own the outcome. If your data infrastructure needs to handle 10x growth in query volume, no consultant can magically make that happen—you need better infrastructure, which is what vendors provide.

The Real Reason You Need Both

According to research on digital transformation and consulting models, the tension between external expertise and platform solutions is increasingly resolved through integration rather than substitution. Organizations that succeed at scale do three things:

1. They invest in software infrastructure first (or early). This gives you the foundation to scale without rebuilding every time your data volume grows. Whether that’s a managed Superset instance, Looker, or another platform depends on your use case, but you need something that handles the operational burden.

2. They hire consultants to design the strategy and initial implementation. The consultant’s job is to ensure your tool investment solves a real problem, not just automate confusion. They also help you avoid common pitfalls—like building dashboards nobody uses because they don’t answer the right questions.

3. They build internal expertise to sustain the system. Once the consultant leaves, your team owns the analytics platform. The software vendor provides the infrastructure; your team provides the judgment about what to measure and how to interpret it.

This is not a sequential process (consultant, then software, then internal team). It’s overlapping. In a well-structured engagement, the consultant helps you select and configure the software while simultaneously training your team to maintain it.

When to Hire a Consultant First

You should prioritize consulting engagement if:

You’re Starting from Scratch or Recovering from Failed Implementations

If you’ve never had a coherent analytics strategy, or if you’ve implemented Tableau or Power BI and nobody uses it, a consultant can diagnose why. Often, the problem isn’t the tool—it’s that the metrics are wrong, the data model is confusing, or the governance model creates friction.

A consultant will help you understand what went wrong and what to fix before you spend more money on tools or infrastructure.

Your Data Infrastructure Is Unclear or Fragmented

If you have data in multiple systems (Salesforce, your data warehouse, Google Analytics, operational databases), a consultant can help you understand what you actually have, what quality issues exist, and what consolidation or integration work is needed before any BI tool can be effective.

You’re Evaluating Whether to Build or Buy

Many engineering-heavy organizations ask: should we build our own analytics platform, or use something off-the-shelf? This is a complex decision that depends on your product, your customers, your team’s capacity, and your timeline. A consultant with experience in both embedded analytics and vendor platforms can help you think through the tradeoffs.

For example, if you’re building a SaaS product and need to embed dashboards for your customers, you might use a platform like D23’s embedded analytics capabilities rather than building from scratch. A consultant can help you evaluate whether that’s the right call and how to integrate it into your product architecture.

You’re Standardizing Analytics Across Multiple Teams or Companies

If you’re a private equity firm rolling out KPI reporting across portfolio companies, or a large enterprise with dozens of business units, a consultant can help you design a governance model and metrics framework that works across all of them. This is beyond what any vendor tool alone can do.

When to Prioritize Software Investment

You should focus on tool selection and implementation if:

You Have Clear Analytics Requirements and a Capable Internal Team

If you already know what dashboards you need, your data is clean and accessible, and you have engineers or analysts who can configure a tool, then you might be able to skip heavy consulting and go straight to a platform. Many fast-moving startups operate this way—they pick a tool (often based on what their team knows) and build from there.

However, even in this scenario, a brief consulting engagement focused on data architecture or governance can prevent costly mistakes.

You Need to Scale Query Performance or Concurrent Users

If you’re hitting limits with your current setup—queries are slow, dashboards are timing out, or you can’t support more concurrent users—the problem is likely infrastructure, not strategy. You need a vendor solution that can handle your scale. This is where managed platforms become critical.

A platform like D23 with Apache Superset’s API-first architecture can handle embedded analytics at scale, supporting both internal dashboards and customer-facing analytics without the overhead of managing Looker’s licensing or Tableau’s infrastructure costs.

You’re Embedding Analytics Into Your Product

If you’re building self-serve BI or dashboards into your application for customers, you need a tool designed for that. Embedded analytics requires API-first architecture, white-labeling, and row-level security—capabilities that are built into purpose-designed platforms.

Consultants can help you design the analytics strategy (what metrics to expose, how to organize them), but the software is non-negotiable for execution.

You Have Limited Budget or Timeline

Consulting is expensive and slow. If you need dashboards in weeks, not months, and your budget is constrained, you might start with a platform and bring in consultants later for optimization. This is a reasonable path, though it risks building dashboards that don’t solve the right problems.

How to Structure an Effective Engagement

If you decide to use both consultants and software, here’s how to make it work:

Phase 1: Discovery and Strategy (Consultant-Led, 4-8 Weeks)

The consultant’s first job is to understand your business, your data, and your analytics maturity. This phase includes:

• Business interviews: Understand the key decisions your organization makes, who makes them, and what data they need.
• Data audit: Map your current data infrastructure, identify quality issues, and assess what consolidation or cleanup is needed.
• Tool evaluation: Based on your requirements, recommend a platform. This might be a managed Apache Superset instance, Looker, Tableau, or something else depending on your use case, budget, and technical constraints.
• Governance framework: Define who owns metrics, how dashboards are approved, and how you’ll manage self-serve analytics without creating chaos.

The output is a strategy document and a tool recommendation with a clear rationale.

Phase 2: Tool Selection and Initial Configuration (Vendor + Consultant, 2-4 Weeks)

Once you’ve chosen a platform, the consultant works with the vendor to configure it for your environment. This includes:

• Data source setup: Connect your data warehouse, databases, and APIs.
• Initial data modeling: Build the semantic layer (dimensions, measures, relationships) that analysts and business users will query against.
• Governance implementation: Set up roles, permissions, and approval workflows.

The consultant ensures the vendor configuration aligns with your strategy, not just their defaults.

Phase 3: Dashboard and Metric Development (Consultant-Led, 6-12 Weeks)

This is where the actual analytics work happens. The consultant builds the dashboards and reports that address the key business questions identified in Phase 1. This includes:

• KPI dashboards: High-level metrics for executives and business leaders.
• Operational dashboards: Detailed metrics for functional teams (sales, marketing, product, finance).
• Exploratory environments: Self-serve areas where analysts can dig deeper.

During this phase, the consultant is also training your internal team on how to maintain and extend the dashboards.

Phase 4: Handoff and Capability Building (Consultant + Internal Team, 4-8 Weeks)

As the engagement winds down, the focus shifts to building internal capability. The consultant documents everything—data models, dashboard logic, governance processes—and trains your team to maintain and evolve the analytics platform.

This is critical. If the consultant is the only person who understands how the system works, you’re dependent on them forever. The goal is to make them redundant.

Real-World Scenarios

Scenario 1: Scale-Up Embedding Analytics Into Their Product

A B2B SaaS company with $10M ARR needs to embed analytics dashboards into their product so customers can see their own performance data. Their engineering team is capable but has never built analytics infrastructure.

The consultant’s role: Help them understand what analytics their customers actually want (discovery), recommend a platform designed for embedded analytics, and design the initial data model and dashboard suite.

The software vendor’s role: Provide the infrastructure to embed dashboards at scale, handle authentication and row-level security, and expose APIs so the product team can integrate analytics into their workflows.

The outcome: Within 3 months, customers have access to embedded dashboards. The engineering team owns the configuration and can add new dashboards as customer needs evolve. The consultant is no longer needed, but the platform scales as the company grows.

For a company in this position, D23’s embedded analytics capabilities on Apache Superset provide the right balance—managed infrastructure without the licensing overhead of Looker, with API-first architecture designed for product integration.

Scenario 2: Private Equity Firm Standardizing KPI Reporting Across Portfolio

A PE firm with 15 portfolio companies wants to standardize how they track KPIs and fund performance. Each portfolio company uses different tools and metrics, making it hard for the PE firm to compare performance or identify issues.

The consultant’s role: Define a standard set of KPIs that apply across all portfolio companies (while allowing for business-specific metrics). Design a governance model for how metrics are defined and updated. Help each portfolio company map their data to the standard framework.

The software vendor’s role: Provide a centralized platform where all portfolio companies can report their metrics. Handle data ingestion from each company’s systems, manage security and data isolation, and provide dashboards for the PE firm’s investment team.

The outcome: The PE firm has a unified view of portfolio performance within 6 months. Each portfolio company maintains their own analytics, but feeds standardized metrics to the central platform. The consultant helps with ongoing metric evolution; the platform handles the operational burden.

In this scenario, a managed Apache Superset platform like D23 with strong API capabilities and data consulting support is ideal—you get the flexibility to define custom metrics without the licensing costs of enterprise platforms.

Scenario 3: Enterprise Recovering from Failed Tableau Implementation

A large enterprise spent $500K on Tableau licenses and implementation, but adoption is low. Dashboards exist, but people don’t use them because they don’t answer the questions people care about. Leadership is frustrated and considering a different tool.

The consultant’s role: Diagnose why adoption failed (probably the dashboards don’t reflect how the business actually operates). Conduct interviews to understand what metrics people actually need. Recommend whether to fix the Tableau implementation or switch to a different platform.

The software vendor’s role: If the recommendation is to stay with Tableau, help reconfigure it based on the consultant’s strategy. If it’s to switch to something lighter and more flexible, migrate the useful dashboards and retire the rest.

The outcome: Within 3 months, people are using the dashboards because they answer real questions. Adoption increases from 20% to 70%. The organization has learned what went wrong and built processes to prevent it next time.

This scenario illustrates why consulting is often a better investment than tool-switching. The problem usually isn’t the tool—it’s the strategy.

Avoiding the Consultant Dependency Trap

One risk with consulting is that you become dependent on the consultant. They leave, and suddenly nobody knows how the system works. Here’s how to avoid it:

Make Documentation Non-Negotiable

Every dashboard, data model, and process should be documented. Not in a consultant’s head—in your systems. Use tools like internal wikis or data catalogs to keep metadata and logic accessible.

Insist on Knowledge Transfer

The consultant’s job includes training your team. This should be structured and ongoing, not ad-hoc. Your analysts and engineers should be able to build new dashboards and modify existing ones without the consultant.

Define Success as Consultant Redundancy

If the consultant is still essential 6 months after they leave, the engagement failed. The goal is to make them unnecessary while maintaining the value they created.

Build Internal Analytics Leadership

You need someone on your team who owns analytics strategy and governance. This person works with the consultant during the engagement and takes over once they leave. They’re not a data analyst—they’re a data leader who understands both business and technology.

The Economics: When Does Each Make Sense?

Let’s talk money. Consulting is expensive—typically $150K-$500K+ for a meaningful engagement. Software platforms vary widely: managed Superset might be $5K-$20K/month, Looker is $5K-$50K+/month depending on usage, Tableau is similar.

Here’s the rough math:

Scenario A: Consulting Only

• Cost: $300K upfront
• Benefit: Strategy, initial dashboards, trained team
• Risk: No scalable infrastructure; you own all maintenance and upgrades
• Best for: Organizations with engineering resources and limited scale

Scenario B: Software Only

• Cost: $10K-$20K/month ($120K-$240K/year)
• Benefit: Scalable infrastructure, governance, support
• Risk: Dashboards might not address real business problems; adoption suffers
• Best for: Organizations with clear requirements and internal capability

Scenario C: Consulting + Software (Recommended)

• Cost: $300K consulting + $10K-$20K/month software = $420K-$540K first year
• Benefit: Strategy + infrastructure + trained team + scalability
• ROI: Faster time-to-value, higher adoption, lower total cost of ownership over 3+ years
• Best for: Most organizations at scale

The combined approach costs more upfront but delivers better outcomes and lower total cost of ownership. A failed platform implementation costs far more than a good consultant.

Evaluating Consultants and Vendors

Not all consultants are equal, and neither are vendors. Here’s what to look for:

In a Data Consultant

• Domain expertise: They’ve worked in your industry or with similar data challenges.
• Technical depth: They understand data warehousing, SQL, data modeling, and the platforms they recommend.
• Change management skills: They can help your organization adopt new processes, not just build dashboards.
• Vendor agnosticism: They recommend tools based on your needs, not their relationships.
• Clear communication: They explain technical concepts in business terms and document everything.

In a Software Vendor

• Scalability: Can the platform handle your growth in data volume and users?
• API-first design: Can you embed analytics, integrate with other tools, and automate workflows?
• Governance and security: Can you enforce data access policies and audit usage?
• Support and community: Is there responsive support and an active user community?
• Total cost of ownership: What’s the real cost when you include implementation, training, and ongoing support?

For organizations evaluating managed Apache Superset, D23 provides a managed alternative to Preset (the commercial Superset vendor) with a focus on data consulting and AI-powered analytics. The combination of managed infrastructure and consulting support addresses both the software and strategy needs.

The Future: Consulting and Software Integration

The boundary between consulting and software is blurring. According to recent research on AI-powered consulting models, enterprises are increasingly seeking vendors that combine platform capabilities with consulting expertise.

This makes sense. A BI platform with built-in AI features like text-to-SQL (natural language query generation) and AI-assisted metric design can reduce the need for some consulting work. But it increases the need for other consulting—helping you think through which metrics to automate, how to govern AI-generated insights, and how to build a data culture that uses these tools effectively.

Platforms like D23 with AI-powered analytics and MCP server integration represent this evolution—combining managed infrastructure with AI capabilities that traditionally required consulting expertise, while still providing access to data consulting for strategy and implementation.

Key Takeaways

1. Vendors solve infrastructure; consultants solve strategy. You need both for sustainable success at scale.

2. Start with consulting if you’re uncertain about your requirements or have failed implementations to recover from. Start with software if you have clear needs and internal capability.

3. The best engagements are phased: discovery and strategy, tool selection, implementation, and handoff. This ensures the consultant and vendor work together toward the same outcome.

4. Make consultant redundancy a success metric. If they’re still essential after they leave, something went wrong.

5. Evaluate total cost of ownership over 3+ years, not just upfront spend. Consulting + software often costs less than either alone due to faster time-to-value and higher adoption.

6. Document everything and build internal leadership. Your organization owns the outcome; the consultant and vendor are enablers.

The question isn’t whether to hire consultants or buy software. It’s how to combine them effectively to build sustainable analytics capability that drives decisions and grows with your organization.

For organizations building on Apache Superset, whether through D23’s managed platform or open-source deployment, the principle remains: invest in both strategy and infrastructure, and ensure they’re aligned from day one.