Forms Management

Power Apps Polysemous Design

Power Apps Polysemous Design

I don’t know if it was intentional or just an artifact of picking a particular word that just happened to have dual meanings related to the subject matter.  My initial thought was the use of Express related to the something traveling at high-speed definition of the word, but then the Figma/ideation part of it speaks to definitions of one expressing themselves through creative impulses…e.g., design.  Hoping it was intentional, but it’s ok if it wasn’t…either way good name.

I’ll be honest, when I first started seeing all the posts on LinkedIn and Twitter about the Power Apps Express Design Preview, I had a moment of anxiety and discomfort, similar to my StoragePoint days when Microsoft released the RBS FILESTREAM Provider.  Back then those feelings quickly turned into excitement bordering on elation when I realized, as delivered, it didn’t replace what we were doing, it amplified the need for the types of solutions ISVs were building/providing.  In other words, it provided us with an opportunity to have a conversation with a much larger audience than was readily accessible to us at that time.  In this case, Express Design amplifies the need for automated processes and associated tooling to accelerate conversion of legacy forms and ideation work product to modern digital experiences.  I especially appreciate the second part of that…the ideation part…because it’s for some reason not obvious to most folks that ideation (net-new work products) is a very important part of this conversation.  There’s a future-proofing aspect to this as well…how do we stop making the same mistake of building these forms and apps in platforms where they live for some period and then die there because we didn’t start with a design or mock-up or model that preserves it in a form that can be easily transformed into a new, more modern form/app platform in the future.

I honestly love everything about Express Design, except the sketching a form out on a piece of paper…I totally see this on a whiteboard, digital or otherwise, but please don’t use paper.  Save the trees!  And how would you iterate?  Wite-out™?  It’s cute, demos nice, looks good on paper (sorry, had to), but not a real-world use case in my opinion.  It could also be because I have terrible handwriting (the worst) and am left-handed, so the flashbacks to the ink or graphite-smudged side of my left make sketching stuff with paper and pen a non-starter for me.

On to the tech… it is obviously in its early days and needs a lot of refinement on the unstructured side (images, static PDFs, etc.), and while I have no doubt it will improve, I can’t imagine it will ever go as far as ISV solutions like Kudzu will…which is fine. Express Design is an OOB solution that’s part of your existing license (…or can easily be added to your subscription) that has these real or artificial upper limits, with partner solutions that fill in the gaps and/or provide advanced capabilities.  This model has worked for Microsoft for decades and it’s hard to argue with the results.  With that in mind, let’s look at how Kudzu stacks up against this early offering from Microsoft.

As a test, we ran several customer-provided forms as well as some from our test library, and the results should demonstrate clearly where the value proposition lies for the respective offerings.

In one case, Express Design was only able to recognize 20% (5 of 25) of the inputs on a very simple, single page form from a large insurance company.  Not surprising given the “hey, try this type of form, not this type of form” you’re presented with in the first step of the Express Design dialog.  You can see the results were not great.  It had a wide range of issues, including Inconsistent recognition of boxed inputs (arguably the easiest inputs to detect because they are fully framed and contain labeling information), not able to detect lined inputs where they are segmented by labels, not able to detect multi-line inputs, incorrectly identified inputs where labels were below the line, and for some reason missed two obvious checkboxes.  I think it just simply struggled with the variations in how inputs were presented (boxed, unboxed, segmented, label to the left, labels below, etc.) in the form along with the form’s density.  I’m sure thresholds and confidence levels could be tweaked under the covers to improve the result somewhat…maybe some of that “tuning” will be surfaced to end users in future releases.

In sharp contrast, Kudzu correctly identified all 25 input fields, including the multi-line text fields associated with Questions 1 & 2 along with the checkboxes…I have a guess as to why we picked them up and Express Design didn’t but won’t give away ingredients to the secret sauce.  We also correctly associated labels with all the inputs, rendered the form top to bottom left to right, and populated available Accessibility attributes to bring the form into compliance with Section 508 and WCAG regulations…note the difference between the Data Field and Accessible Label properties of the selected input…all of this is done automatically.  We also perform contrast detection and recommendation/remediation, so if there were elements of this form that didn’t meet a 4.5:1 minimum contrast ratio requirement we would flag them, recommend the minimum color change needed to achieve a 4.5:1 contrast ratio, and give the option to automatically repair or manually repair by a user in our Uniform model tooling.

At this point, without any need to make any changes, this form is ready to be authored as a Power App.  Kudzu’s Power Apps writer would generate the MSAPP package along with creating the necessary Dataverse entities (in this case one, but multiples if we were dealing with any repeating sections/tables) to back it…support for new or existing SharePoint lists is in an upcoming release.

For another example, we took the standard Volunteer Signup Word template that was used in another video we watched, and while it performed better, there was still a lot of work to do.  It missed several inputs, and for some reason, the output had obvious layout and alignment issues.  Again, I have no doubt it will get better.

Here’s what it looked like in Kudzu…again, ready to go without any fiddling.  Ready to be authored as a Power App.

And to be clear, converting these types of forms barely scratches the surface of what the Kudzu platform is capable of.  Here’s page 12 of an 18-page PDF provided to us by another prospective customer.  It has over 700 inputs, is incredibly input dense, and has tables with different types of inputs anchored with compound column and row headers.

You’ll note that the field name for the selected input is a combination of the group header (FREQUENTLY), it’s child (Carry), and the row (21-50lbs.) …with symbology conversion (the dash) and abbreviation (lbs.) expansion…and then we sentence case that for the Accessible label.  Would you convert this 18-page monster to a Power App?  Never.  Maybe convert to our runtime hosted in SharePoint or maybe as a Nintex K2 SmartForm.  But you could definitely convert it to a WCAG-compliant fillable PDF.

You can watch the video (linked below) to see for yourself.  In the video, you’ll see me convert a PDF (the aforementioned Volunteer Signup), a multi-view Balsamiq mock-up and an InfoPath form to Power Apps.  The full list of readers and writers can be found on our website at with more coming later this year, including Figma.

BTW, we’re taking a very different approach to our Figma integration…no kits.  I appreciate what Microsoft is trying to do but am not convinced serious designers want to be put in the position to build out platform-specific designs.  I personally know a lot of design professionals that would likely never use this because a design should be easily transformable by an engineer or citizen developer into their platform of choice…not laden with platform-specific elements and metadata.  Rarely have I heard a designer ask web or desktop?  Razor, Blazor, React or Angular?  Power App or Nintex K2 SmartForm?  They always ask what form factor(s)…and that’s about it.  Tell them “Here’s a kit you have to use in your design because we’re building Power Apps” and prepare yourself for the blank stares and a thorough explanation on design works.  That’s our take anyway.

Want to see a short demo of our Power Apps capabilities watch this: Kudzu for Power Apps – YouTube