Skip to content

Part 2: Selecting a BI Solution

The Guide to Building a Dashboard for your Organization

The following article (Selecting a BI Solution) is part of our Free Guide to BI.


Once a company has decided to get started with business intelligence (BI), they must choose a tool or platform on which to build their dashboards. There are many vendors that offer solutions in this space, including SaaS (software-as-a-service) tools like Domo, Tableau, Grow, Looker, Qlik, and Power BI. But, in the process of selecting a BI solution, companies need to achieve a balance across multiple criteria rather than looking at just a single factor (e.g. cost, functionality, integration with existing systems). 

Below is a list of criteria that are key for companies in choosing a BI solution. Each is presented with a Decision Weight (DW) that indicates its impact on the final selection – this is where participation from multiple departments is needed to determine the best solution.

1. Functionality (DW of 40)

  • Underlying Data: This includes:
  • Connector Availability: Does the BI solution easily integrate with key third-party data and databases through pre-built connectors across all marketing/HR/Sales/Finance business functions (e.g. Amazon Redshift, Amazon S3, Dropbox, Salesforce, Google Analytics, Google Ads, Facebook Ads, Netsuite, Flurry, Jira, Mailchimp, Microsoft Excel)?
  • API Connectivity: Does the BI solution easily integrate with any application program interface (API)?
  • Data Limitations: Does the BI solution import and store all the data that it has been connected to on its platform? This would mean that any data cleaning, combining and transforming would happen on the platform. If this is the case, what are the limits of that storage (in terms of gigabytes or rows of data)? Alternatively, does the BI solution create a “live” connection to the data hosted by the publisher, or its service providers, when displaying a dashboard or card/graph? This would mean that only simple data cleaning, combining and transforming could be performed by the platform.
  • Data Manipulation: Does the BI solution offer a visual tool, usable by a business/marketing analyst, to support cleaning, combining, and transforming the company’s data? Does the BI solution support SQL and R for developers?
  • Dashboard and Cards/Graphs: The mechanics around building and using dashboards and cards/graphs. This includes:
  • User Experience (UX): What is the technical level (e.g. coding skills) needed to build cards/graphs and dashboards? Does the building and editing of every card/graph need the participation of a developer?
  • Filtering and User-Defined Cohorts: Is there dashboard-level filtering? Are these user-defined cohorts persistent as users click onto cards/graphs?
  • White Labeling: Can the dashboard and/or individual cards/graphs be customized to show the publisher’s logo and branding or embedded into the publisher’s site? Can the publisher add custom messaging?
  • Mobile: Does the vendor offer mobile access that allows card/graph or dashboard creation and editing?
  • Digital Signage: Does the vendor offer a digital signage solution to show reformatted dashboards on public space displays?
  • Alerts: User-defined alerts can be triggered and sent (e.g. email, text) based on important changes or thresholds crossed at either the dataset level, the metric or the card/graph level.
  • Usage Reporting: The BI system provides access to either detailed reports or raw data on how users are interacting with built dashboards and cards/graphs.
  • Security: This includes:
    • User Authentication: Support for single sign-on (SSO)
    • Access rights: Can the company define roles within the BI solution to grant users specific rights to access and/or edit selected datasets, cards/graphs, and dashboards? Does the BI solution support row-level security within a dataset? 

2. Technology and Infrastructure (DW of 25)

  • SaaS vs On-Prem: Is the BI solution offered as SaaS, as a locally installed software (on-prem), or a hybrid combination of the two?
  • Backup:  Data backup and redundancy plan, frequency (hourly, daily), and onsite/offsite setup.
  • Updates: Regular updates in functionality (e.g. data connectors, card/graph types, dashboard design), and underlying technology (e.g. dataset size limitations, redesign speed). 
  • Security: Handling of company’s data, including:
    • Theft/Copying Countermeasures
    • Competing Client Data Separation
    • Employee Monitoring (e.g. cameras, PC logs, supervision)
    • Security of Premises: 24-hour guard, alarm, and personnel entrusted with keys.

3. Service and Support (DW of 15)

  • Support:  Proper communication and timely response, including:
    • Contact Methods (e.g. phone, IM, online system, email)
    • Location: Is there a US client support team?
    • Hours: Overlap with publisher working hours and ability to extend them for deadlines.
    • Service Level Agreements (SLA): Does vendor offer an agreement that guarantees strict delivery terms, including penalties?
  • Training: Program to train company in vendor’s systems, procedures, tools and best practices.
  • Consulting: Does vendor offer help in further tailoring its solution beyond the standard package offered?

4. Vendor Reliability (DW of 20)

  • Experience: Age of the business, management experience, and existing/past clients (their type, depth and length of relationship, frequency of repeat business, and opinion of the vendor’s products and services). 
  • Strategy: Vendor’s proactive vision and strategy to expand functional coverage, upgrade technology and infrastructure, serve clients more efficiently, and improve market position.
  • Financial Confidence: Life expectancy of the vendor. This includes:
    • Financial Backing: How is the company financed?
    • Employees: How many employees are currently employed?
    • Financial Results: Is the vendor profitable or showing a solid trend towards profitability?
    • Market Share: Vendor’s share of the relevant market.
    • Dependence: How much of vendor’s total business is from specific client?

It is particularly important that an extensive proof of concept (PoC) be conducted with the selected solution (or, if possible, the shortlisted solutions) to ascertain whether the various departments involved are satisfied with the solution’s actual functionality in a real environment.

Furthermore, a company should only look at pricing after evaluating and selecting a shortlist of possible BI vendors according to the four criteria groups above. Though various pricing models exist in the BI space that depend on whether the solution is SaaS or on-prem, the company should focus primarily on comparing:

      • Per user costs over three years (including support and maintenance)
      • Training costs
      • Proof of concept (PoC) costs
      • Any consulting and third-party integration/implementation costs

While the long-term success of a BI solution is heavily dependent on how important data and the intelligence deriving from it is to an organization, selecting the wrong vendor can hamper even the best plans.