Skip to main content

9 ways self-service BI solutions fall short

Recently, Tableau put out a presentation entitled “10 Ways PowerBI Falls Short”. The aggressive approach geared at pinpointing Microsoft’s weak points opened some eyes and maybe ruffled some feathers.

Truth be told, it’s easy to pull out a handful of items you may offer (or excel in) that others don’t. So we thought we’d take this a step further and point out, not a list of lacking features, but look at the overall problems that are fraught with most of the self-service BI tools in general.

To be fair, it’s a market seemingly full of solutions that somehow still appears unaccommodated because these stalwarts that have either built their solutions based on somewhat faulted/inaccurate ideas of self-service, or have been trying to migrate their existing traditional BI solutions to the cloud without renovating the legacy technical infrastructure.

So, to take over where Tableau has left off (or, to carry on what they have started), we offer you 10 ways that most of the self-service BI solutions fall short:

1. Hidden Complexity

Just because a vendor claims ease of use, doesn’t mean that it is true. Some are better than others in different stages from the beginning (preparing data) to the end (building analysis and acting upon the findings). It may be easy for the business professionals to use the dashboards built by IT experts, but someone with the necessary technical expertise has to build them first. Building analysis should require no programming.

Furthermore, building multi-dimensional analysis should not require a pre-built data warehouse or data model (i.e. OLAP cube) residing in the data source. Most “self-service” BI tools out there either choose to neglect data processing, requiring source data to be as clean as possible and with data models built beforehand, or are bundled with ETL tools which inevitably get IT involved and fall back to the IT centric BI approach.

2. Web Access vs. Cloud-Native

When offering a dedicated Windows application for creating dashboards, one of the most important steps – building analysis – has to be done on the ground. Although users may be given the option of a sharing analysis solution or a publicly-hosted service, essentially that is just providing web access. In addition, the application will have to be installed and maintained on each computer where users need to build analysis. The desktop application, server, and public cloud service are licensed separately which further complicates customer’s adoption process. This is the model used by the leading self service BI vendors. 

Why would these solutions require a ridiculously complex and time consuming method? Because it’s expensive and it requires their help. Using a cloud-native platform, a user only needs an account and a device that can browse the Internet. Everything from building analysis, generating reports, sharing information, to collaborating can all be done in the cloud – with one single product.

3. Collaboration

BI tools put lots of emphasis on producing flashy charts and reports, abandoning the users at the most critical juncture. They have no mechanisms in place to facilitate users taking actions upon the analysed results.

Users should not have to sift through emails or to check on the status. All comments, feedback, and correspondence as well as detailed history for tracking and auditing need to be made available in one accessible place. It should also allow parties to upload other related documents to share with one another and keep those resources related to the analysis.

4. Integration

Ground-based or browser-based BI tools cannot easily compete with a cloud native infrastructure when it comes to its inherent ability to seamlessly integrate outside data sources including spreadsheets, relational databases, web services, 3rd party cloud-based services, networking devices, and social media. This model also enables the BI solution to be seamlessly integrated with other cloud service providers.

5. Customisation

When there is the implicit reliance on data modeling, programming, and/or ETL process, and business users want to make changes to their analysis, they will need to contact IT (and hope they understand what is needed). True self-service allows analysis to be customised by the business users who know their data and analysis the best.

Tools need to allow users to create custom home pages as well as customise the look-and-feel of the entire experience. So, even though users are technically using the same platform, their analysis can be customised and packaged as if it is a tailor-made BI solution just for them. Most importantly, all the customisation should be easy enough to do by the users themselves with no technical skills needed.

6. Kick-Start

BI vendors are typically tasked with providing a proof of concept when trying to sell their products to new customers. And it shouldn’t be a surprise that they are mainly focused on big money accounts. But the best self-service BI will provide businesses of all sizes substantial ROI.

BI solutions with a cloud-native infrastructure are able to easily offer prospects sandbox accounts for them to further experience the product right after the initial demonstration meeting or offer a proof of concept with a typical turn-around cycle, from requirements gathering to final demonstration, of about a week. BI vendors can and should help customers grow their businesses by using their solutions, not select customers based on their ability to pay.

7. Cost & Scalability

The licensing models of the premise-based BI platforms require a substantial capital investment so an organisation needs to weigh the investment with ROI as they roll out a solution in large scale, plus the ongoing maintenance cost as applications and servers are installed on-premise. Capital investment also imposes vendor lock down – since a significant chunk of budget has already been spent, organisations may have no choice but to keep using what they’ve purchased even if later on their objectives have shifted or better solutions have emerged.

The cloud-native solutions can offer a simple yet flexible subscription-based pricing model. There is no capital investment, no minimums, easy to get started, and easy to scale up as the organisation’s analysis demand increases.

8. Continuous Delivery

Being cloud-native allows BI to offer simple and flexible pricing and engineering teams to frequently enrich products and subsequently the user experience. Enhancements can be constantly rolled out under a fast-paced release cycle. Customers will quickly start to appreciate the value-added service – they keep benefiting from the continuous improvements of the platform without extra charges.

While for typical self-service BI vendors using traditional client-server architecture and a traditional license-based pricing, upgrades are only for a limited time. After which any upgrade would become another round of capital investment.

9. Live Support

When your BI features 100 per cent cloud flexibility, facilitates ad-hoc, dynamic analysis and accelerates real-time information sharing and collaboration, why stop there?

Customer support can also be live. Should users have any questions regarding how to use the system, live chat and voice calls can easily be supported.

Ben Tai, Founder and CEO of DrivenBI (opens in new tab)

Image Credit: Shutterstock/Sergey Nivens