What is your low-code strategy?

2 years' of digital transformation in 2 months. The pace of digital transformation demanded by today's market (further exacerbated by the current pandemic) is accelerating faster than ever. If Digital Transformation is to be represented by a giant purple Alien hell-bent on resource optimisation, it would be Thanos (who else?). Pop-culture reference forced in just so that I can quote *I am inevitable.

However, businesses can only transform as fast as the available capacity of their limited resources can afford to - often within IT and often in partnership with outsourced technology partners. The secret weapon, then, is Low-Code. This is the multiplier effect you can have - if your digital transformation depends on a team of 10 IT superheroes, low-code is your cheat-code to transform the business as if you have a team of 10 IT superheroes plus 500 non-IT superheroes.

Low-code is not a new paradigm. One needs to recognise that low-code already exist in every organisation. The big difference is, if an organisation did not have a low-code strategy, business users will impose their own low-code to the organisation. Without the strategy. (Hint: Excel, Access, that dodgy "freeware" Bob installed in his Word).

But why low-code? Or, better still - why digital transformation? Why it matters if my current process requires 3 spreadsheet templates, 5 emails minimum, 2 separate system logins vs a modern app that automates the entire process?

Let's start with the first stage of digital transformation - capturing data digitally. This is about providing users with a digital interface to enter data with the least friction in their experience. The harder this task is, the less likely you will get good and timely data. Therefore experience is King here. Low code means your business users can help shape the app's UX - down to the colour of a button and positioning of a drop-down. But don't just restrict the UX to the classic forms-over-data design. Why not interact with real-world objects? Why not have an app that uses your device's location services and capture the location of the task without asking the user to enter their current location? Why not use the camera on the device and capture an image of an incident, and then detect objects and tag it accordingly? How about overlaying a 3d model into real physical image stream from the camera as a mixed reality experience? Don't stop at just capturing the user name and timestamp. Your Excel can already do that ages ago. The 80s called and it want its forms-over-data app back.

What do you do with the data your app have captured? Put it to work. Funnel it through your business process. And by business process, we are not talking about printing out the data and getting the manager to physically sign off on it. We want to automate the heck out of it, of course. Automation can be low-code as well, not just the app. And by automation, we meant the entire automation spectrum - from your run-of-the-mill approvals to sophisticated data orchestration across multiple APIs to Robotic Process Automation for that one VB6 app that no one dares to touch anymore for fear of angering the COM+ gods. Even if you have one business process that spans the entire spectrum - starts an approval, then runs some queries, invoke some APIs, and finally if everything checks out, runs a robotic process over that legacy app - you can low-code the entire process on one automation platform.

At the end of this process, you should get good, clean, validated data. What happens next will be critical. Most organisations will land the data into a data store and stop there. Digital transformation: Done. Well, no - not yet. What often happens is that the data lands in a latent store. Sure, you get value out of digitising and automating the process - increased efficiency, increased productivity, reduction of errors are the usual benefits (faster, more, better). This is where it is important to have a good data strategy as part of your digital transformation narrative. Simply storing the data is not a good strategy. What needs to happen here is to have an ability to describe the relationships between different sets of data. Ultimately you want to be able to prepare your data for intelligence. Imagine you have two apps, App #1 is a work-safety app, where users will use to capture workplace site incidents, and App #2 is a time-sheet app. App #1 emits data on workplace incidents. Incidents can be related to project sites. Sites can be related to workers' attendance/shifts. Shifts can be related to timesheets. Timesheet data is emitted by App #2. Now, at their intersection, we can now paint a picture of what type of incidents will occur if resources are overloaded for a specific type of projects. How about that? To derive this type of intelligence, we need good data models that supports sophisticated relationships, and can be used by a myriad of apps. You can have a low-code platform for this too - build your relational data models that can be easily made available for analytics, with security and controls in place. Plus, you can ingest, transform and shape data from other sources into your data models.

The more apps you put into this platform, the more data you get, and ultimately the more intelligence you can extract out of it. You can now start applying business reasoning over your nice little data estate - you are no longer viewing your data as a report on how many incidents occurred in the past 12 months. You want to know the "why". Why there are more incidents of a specific category for this type of project? Ask not what you can do to your data. Ask what your data can do for you. This is all about getting insights, about knowing the why. Why there is more churn for this demography? Why this marketing campaign is more effective for new graduates?

Digital transformation should not stop here. When you start to get intelligence out of your data, you would want to consume it. This is the holy grail for most organisations. This is about how organisations can transform from being re-active, to being pro-active. This is about taking the insights you have now gained, and feeding that into machine learning models, using artificial intelligence to predict, recommend, detect anomalies. You now have good insights to guide you on how your data points are influencing each other, time to put this to good use by creating AI models so that you can predict the future.

(Credits to Tim Madden for the visuals)

(Credits to Mark Weimann for this nugget of wisdom on reporting/analytics/AI)

Why low-code matters? It matters because it greatly accelerates this path to intelligence. It frees up your IT resources to focus on creating data-driven intelligence for the organisation, and not spending time on getting the right colour for the button because Finance wants it to be exactly RGBA(128,128,128,0) or they will hate the app forever. Low-code matters because you need lots of data from lots of apps to get to meaningful intelligence for your organisation. To get lots of data from lots of apps, you need Citizen Developers.

Low-code enables true digital transformation by opening the pathway to intelligence. With great data, comes great Intelligence. And with great apps, comes great data.

Here's my closing "Expanding Mind" meme :-)

So what's your low-code strategy?

Find out more about what Microsoft Power Platform can do for your organisation.