We Killed Power BI, You Should Too.

Death to Power BI

My team and I are embracing A.I. to build bespoke solutions for our stakeholders at an extremely rapid pace. We were in the process of decommissioning a Business Intelligence (BI) tool and moving to Power BI (PBI) so that there was a unified solution and to save costs. We then began considering al options and of course A.l. was in the discussion.

Why Power BI?

Our first thought as reasonable engineers was,

“Why are we moving to Power BI ?”

In our case, we get all our data from an internal data platform, where, the data should already have undergone the ETL process already and even if we then needed to enhance the data further, in Power Bl, we would need to use M-code, which is just a farce. The alternative was to write the code ourselves using python for the ETL (which has terrific data modelling libraries) and then use Javascript for the frontend. The advantage here being that we have far greater flexibility (we’re not limited b y Power Bl’s handholding) and maintainability - if done well, whilst still maintaining the user friendliness of desktop and mobile experiences, natively. Of course, Power Bl offers these, but there are a lot of nuances it doesn’t support. For example, a stakeholder of ours was very disappointed that you could not have custom, dynaminc colour text because, in the curent tool, they used it as a quick way to visualy filter between things like green, for good, and red, for bad. From a technical POV, given our needs, there was no reason why should stay with Power Bl; but, there are more reasons than technical, and user, to consider and in this case it was cost and time.

Everything Everywhere All at Once

Microsoft’s M365 includes SO much for the £80 per user, per year. Yes, we could successfully decommission Power Bl, but we weren’t going to get rid of Teams, Outlook, OneDrive, Powerpoint, Sharepoint, Excel, Word and the many, many others. So what was the point? And what even is the point of this article if I used a clickbait title? I want to explore the idea of the growth of bespoke solutions in the near-mid future and the death of software product companies in the long term. I think companies will demand bespoke solutions as opposed to one-size-fits-all SaaS or SaaP solutions. I postulate that they are dead. Why would I pay these subscriptions when I can hire an internal team of ~5, enabled by A.l., to productise all my needs. Now, of course, this is not the death of all SaaS solutions, but I think the market might become too squeezed. I think startups will more so and more so utilise open-source tooling like Obsidian for documentation and LibreOffice for all the office suite apps, where they can enhance them to suit their needs. This will keep costs down and provide the tool exactly as they need it. The issue then, is, employees leaving and new features not being added, but now with A.l. able to index and get context of entire code bases, new starters will be able to get caught up to speed very quickly.

5% Feature Usage

Furthermore, I invite you to consider how many of these apps features you really use? Take outlook or example. It’s unbelievably feature rich and it is as old as time itself - so you’ve had plenty of time to use them all, yet; I guarantee you’ve never ventured far outside of ‘new mail’, ‘attachment’ & ‘text formatting’. The problem the arises when there is a niche element that you want really bad but it does not have it… That’s where add-ons to apps like Outlook spawned from, allowing people to add the desired functionality. The caveat being the limitations of what Outlook exposes to you and allows you to do. Whereas, you could quickly add the feature yourself in your bespoke solution.

Conclussions & Actions

I say this as I think that consultancies are actually well positioned to do well in the bespoke market because that’s pretty much what they do already. What wil be interesting wil be the monetisation of this for the consultancies. I think that big companies wil become even more of a “try before you buy”-type business model. As for SaaS enterprises, I think they wil struggle and I think the room for software only companies will reduce. They wil need to pivot to be a technology company once more and their purpose wil need to be working on frontier technologies like Augmented Reality (AR), energy solutions, low gravity manufacturing and so on.